Unit 1

Capital Markets, Consumption, and Investment

INTRODUCTION

Essentially, capital is wealth, usually in the form of money or property. Capital markets exist when two groups interact with each other viz.; those who are seeking capital and those who provide capital. The capital seekers are the businesses and governments who want to finance their projects and enterprises by borrowing or selling equity stakes. The capital providers are the people and institutions who are willing to lend or buy, expecting to realize a profit.

The central purpose of a corporation is to provide value for the shareholders Or in simple terms, it is to make money. The discipline that studies money, markets and their operation with an eye to practical application is called finance. Thus, financial theory has a significant impact on the decisions made by the leaders of corporations.

Accountants have been described as historians. They record how much is spent, earned and invested. They sort, classify and organize. However, what these actions mean, is not their primary concern. Economists, on the other hand, are concerned with the more general questions of how markets operate, why they operate thus and what this means for individuals, businesses and nations. Financial professionals stand in the middle ground. While they must stay in touch with the accounting and have a keen understanding of the real-world operations of the firm, they also attempt to understand the esoteric aspects of economic theory.

Though finance requires certain decisions to be made, as opposed to the decisions' impact being recorded, predictions and estimations must be made. However, unlike economics, which addresses such general questions, finance professionals use theories, models and statistical tools to answer specific questions. An economist may ask the question, "Is this industry expanding its operations, and if so, at what rate?" The financial professional may use some of the same tools, but is more likely to ask, "Should our company expand operations, and if so, when?"

In case of distinction made between finance and economics , then there is a possibility of some overlap between the disciplines and indeed, some of the most esteemed scholars in financial theory are great economists. John Maynard Keynes, William Sharpe and Oskar Morgenstern were responsible for innovations like the capital asset pricing model and modern portfolio theory. However, these scholars are almost universally referred to as economists, not financial theorists.

While financial theory, like all theory, is imperfect, it does provide rational guidance for dealing with uncertainty. For example, analysts may regard the company's stock price as too low. This may lead the directors of the corporation to purchase shares of the stock from shareholders, allowing the firm to retain more of its profits as it would not have to pay dividends on shares that it repurchased. Another possibility would be theorists predicting significant changes in the interest rates demanded by credit markets. The corporation may then hasten, or delay, obtaining needed credit lines.

The greatest challenge to financial theory is the reliance on reason. While objective assessment of measurable facts and application of proven formulas may seem unassailable, as Publius Syrus said "Everything is worth what its purchaser will pay for it." Pricing models, market statistics and economic data may be invaluable for providing predictions, but sometimes a particular product, company or brand name may fall into fashion or out of favor for reasons that may be apparent only in retrospect or forever remain inscrutable.

Let us first see what exactly are the financial markets and institutions.

A financial market is a market in which financial assets are traded. In addition to enabling exchange of previously issued financial assets, financial markets facilitate borrowing and lending by facilitating the sale by newly issued financial assets. Examples of financial markets include the Dar es Salam Stock Exchange (DSE), (resale of previously issued stock shares), the Tanzanian government bond market (resale of previously issued bonds), and the Treasury bills auction (sales of newly issued T-bills). A financial institution is an institution whose primary source of profits is through financial asset transactions. Examples of such financial institutions include discount brokers (e.g., Charles Schwab and Associates), banks, insurance companies, and complex multi-function financial institutions such as Merrill Lynch.

Financial markets serve six basic functions. These functions are briefly listed below:

Borrowing and Lending: Financial markets permit the transfer of funds (purchasing power) from one agent to another for either investment or consumption purposes.

Price Determination: Financial markets provide vehicles by which prices are set both for newly issued financial assets and for the existing stock of financial assets.

Information Aggregation and Coordination: Financial markets act as collectors and aggregators of information about financial asset values and the flow of funds from lenders to borrowers.

Risk Sharing: Financial markets allow a transfer of risk from those who undertake investments to those who provide funds for those investments.

Liquidity: Financial markets provide the holders of financial assets with a chance to resell or liquidate these assets.

Efficiency: Financial markets reduce transaction costs and information costs.

In attempting to characterize the way financial markets operate, one must consider both the various types of financial institutions that participate in such markets and the various ways in which these markets are structured.

Types of financial markets

These are the types of financial markets which are as under,

Capital market: Capital market is a market where companies and government raise their capital for its operations and investment.

Stock market: Stock market is a public place where stocks and securities are traded. Bond market: Bond market is a financial place where debt securities and bonds are traded. Commodity market: Commodity markets are those markets where raw material and primary

products are traded. Money market: Money markets are those markets where assets and other one year maturity

securities are traded is called money market. Derivative market: Derivative markets are those markets which deals the future contracts. Future market: Future market deals the buying and selling of commodities on a specified future

date. Insurance market: Insurance market deals the risk of loss in future or uncertain period. Foreign exchange market: Foreign exchange market is the world wide market which deals the

different currencies.

STOCK BROKER

Stock SellerStock Buyer

(Passed Through)

PaymentPayment

Stock SharesStock Shares

BOND Bond SellerBond Buyer

(Bond Inventory

PaymentPayment

BondsBonds

Major Players in Financial MarketsBy definition, financial institutions are institutions that participate in financial markets, i.e., in the creation and/or exchange of financial assets. At present in the United States, financial institutions can be roughly classified into the following four categories: "brokers;" "dealers;" "investment bankers;" and "financial intermediaries."

Brokers: A broker is a commissioned agent of a buyer (or seller) who facilitates trade by locating a seller (or buyer) to complete the desired transaction. A broker does not take a position in the assets he or she trades -- that is, the broker does not maintain inventories in these assets. The profits of brokers are determined by the commissions they charge to the users of their services (either the buyers, the sellers, or both). Examples of brokers include real estate brokers and stock brokers.

Diagrammatic Illustration of a Stock Broker:

Dealers:

Like brokers, dealers facilitate trade by matching buyers with sellers of assets; they do not engage in asset transformation. Unlike brokers, however, a dealer can and does "take positions" (i.e., maintain inventories) in the assets he or she trades that permit the dealer to sell out of inventory rather than always having to locate sellers to match every offer to buy. Also, unlike brokers, dealers do not receive sales commissions. Rather, dealers make profits by buying assets at relatively low prices and reselling them at relatively high prices (buy low - sell high). The price at which a dealer offers to sell an asset (the "asked price") minus the price at which a dealer offers to buy an asset (the "bid price") is called the bid-ask spread and represents the dealer's profit margin on the asset exchange. Real-world examples of dealers include car dealers, dealers in U.S. government bonds, and Nasdaq stock dealers.

Diagrammatic Illustration of a Bond Dealer:

Investment Banks:

B H

Deposited fundsFunds

Loan contracts Deposit accounts

Lending by B Borrowing by B

Loan contracts issued by F to B are liabilities of F and assets of BDeposit accounts issued by B to H are liabilities of B and assets of H

F

An investment bank assists in the initial sale of newly issued securities (i.e., in IPOs = Initial Public Offerings) by engaging in a number of different activities:

Advice: Advising corporations on whether they should issue bonds or stock, and, for bond issues, on the particular types of payment schedules these securities should offer;

Underwriting: Guaranteeing corporations a price on the securities they offer, either individually or by having several different investment banks form a syndicate to underwrite the issue jointly;

Sales Assistance: Assisting in the sale of these securities to the public.

Standard Bank in Tanzania is the best example of Investment Banking firm.

Financial Intermediaries:

Unlike brokers, dealers, and investment banks, financial intermediaries are financial institutions that engage in financial asset transformation. That is, financial intermediaries purchase one kind of financial asset from borrowers -- generally some kind of long-term loan contract whose terms are adapted to the specific circumstances of the borrower (e.g., a mortgage) -- and sell a different kind of financial asset to savers, generally some kind of relatively liquid claim against the financial intermediary (e.g., a deposit account). In addition, unlike brokers and dealers, financial intermediaries typically hold financial assets as part of an investment portfolio rather than as an inventory for resale. In addition to making profits on their investment portfolios, financial intermediaries make profits by charging relatively high interest rates to borrowers and paying relatively low interest rates to savers.

Types of financial intermediaries include: Depository Institutions (commercial banks, savings and loan associations, mutual savings banks, credit unions); Contractual Savings Institutions (life insurance companies, fire and casualty insurance companies, pension funds, government retirement funds); and Investment Intermediaries (finance companies, stock and bond mutual funds, money market mutual funds).

Diagrammatic Example of a Financial Intermediary: A Commercial Bank

NOTE: F=Firms, B=Commercial Bank, and H=Households

Important Caution: These four types of financial institutions are simplified idealized classifications, and many actual financial institutions in the fast-changing financial landscape today engage in activities that overlap two or more of these classifications, or even to some extent fall outside these classifications. A prime example is Merrill Lynch, which simultaneously acts as a broker, a dealer (taking positions in certain stocks and bonds it sells), a financial intermediary (e.g., through its provision of mutual funds and CMA checkable deposit accounts), and an investment banker.

Types of Financial Market Structures

The costs of collecting and aggregating information determine, to a large extent, the types of financial market structures that emerge. These structures take four basic forms:

Auction markets conducted through brokers; Over-the-counter (OTC) markets conducted through dealers; Organized Exchanges, such as the Dar es Salam Stock Exchange, which combine auction

and OTC market features. Specifically, organized exchanges permit buyers and sellers to trade with each other in a centralized location, like an auction. However, securities are traded on the floor of the exchange with the help of specialist traders who combine broker and dealer functions. The specialists broker trades but also stand ready to buy and sell stocks from personal inventories if buy and sell orders do not match up.

Intermediation financial markets conducted through financial intermediaries;

Financial markets taking the first three forms are generally referred to as securities markets. Some financial markets combine features from more than one of these categories, so the categories constitute only rough guidelines. Auction Markets: An auction market is some form of centralized facility (or clearing house) by which buyers and sellers, through their commissioned agents (brokers), execute trades in an open and competitive bidding process. The "centralized facility" is not necessarily a place where buyers and sellers physically meet. Rather, it is any institution that provides buyers and sellers with a centralized access to the bidding process. All of the needed information about offers to buy (bid prices) and offers to sell (asked prices) is centralized in one location which is readily accessible to all would-be buyers and sellers, e.g., through a computer network. No private exchanges between individual buyers and sellers are made outside of the centralized facility. An auction market is typically a public market in the sense that it open to all agents who wish to participate. Auction markets can either be call markets -- such as art auctions -- for which bid and asked prices are all posted at one time, or continuous markets -- such as stock exchanges and real estate markets -- for which bid and asked prices can be posted at any time the market is open and exchanges take place on a continual basis. Experimental economists have devoted a tremendous amount of attention in recent years to auction markets.

Many auction markets trade in relatively homogeneous assets (e.g., Treasury bills, notes, and bonds) to cut down on information costs. Alternatively, some auction markets (e.g., in second-hand jewelry, furniture, paintings etc.) allow would-be buyers to inspect the goods to be sold prior to the opening of the actual bidding process. This inspection can take the form of a warehouse tour, a catalog issued with pictures and descriptions of items to be sold, or (in televised auctions) a time during which assets are simply displayed one by one to viewers prior to bidding. Auction markets depend on participation for any one type of asset not being too "thin." The costs of collecting information about any one type of asset are sunk costs independent of the volume of trading in that asset. Consequently, auction markets depend on volume to spread these costs over a wide number of participants. Over-the-Counter Markets: An over-the-counter market has no centralized mechanism or facility for trading. Instead, the market is a public market consisting of a number of dealers spread across a region, a country, or indeed the world, who make the market in some type of asset. That is, the dealers themselves post bid and asked prices for this asset and then stand ready to buy or sell units of this asset with anyone who chooses to trade at these posted prices. The dealers provide customers more flexibility in trading than brokers, because dealers can offset imbalances in the demand and supply of assets by trading out of their own accounts. Intermediation Financial Markets: An intermediation financial market is a financial market in which financial intermediaries help transfer funds from savers to borrowers by issuing certain types of financial assets to savers and receiving other types of financial assets from borrowers. The financial assets issued to savers are claims against the financial intermediaries, hence liabilities of the financial intermediaries, whereas the financial assets received from borrowers are claims against the borrowers, hence assets of the financial intermediaries.

Capital Markets

Markets exist to facilitate the purchase and sale of goods and services. The financial market exists to facilitate sale and purchase of financial instruments and companies of two major markets, namely the capital market and the money market. The distinction between capital market and money market is that capital market mainly deals in medium and long-term investments (maturity more than a year) while the money market deals in short term investments (maturity upto a year). Capital markets provide avenue where companies can raise funds to expand on their businesses or establish new ones by issuing securities owned by the companies. Like businesses in the private sector, Government issue its securities to raise funds in capital markets to build electricity damn, construct new roads, bridges by issues.

Capital market can be divided into two segments viz. primary and secondary. The primary market is mainly used by issuers for raising fresh capital from the investors by making initial public offers or rights issues or offers for sale of equity or debt. The secondary market provides liquidity to these instruments, through trading and settlement on the stock exchanges.

Capital market is, thus, important for raising funds for capital formation and investments and forms a very vital link for economic development of any country. The capital market provides a means for issuers to raise capital from

investors (who have surplus money available from saving for investment). Thus, savings normally flow from household sector to business or Government sector, which normally invest more than they save. A vibrant and efficient capital market is the most important parameter for evaluating health of any economy.

In a market economy, the role of the capital market is very important. The good functioning of the capital market is vital in the contemporary economy, in order to achieve an efficient transfer of monetary resources from those who save money toward those who need capital and who succeed to offer it a superior utilisation; the capital market can influence significantly the quality of investment decisions. The gathering of temporary capitals that are available in the economy, the reallocation of those that are insufficiently or inefficiently used at a certain moment and even the favouring of some sectorial reorganisations, outline the capital market’s place in the economy of many countries. Among the mechanisms which are specific for the capital market, the mechanism of financing the economy is the most important one, although, due to the fact that an ensemble of interconnected mechanisms contribute to the good functioning of the capital market, the way in which each of them function determines the quality of the ensemble. Thus, for instance, the efficient allocation of resources, which is achieved by the help of the capital market, relies on the information obtained on the basis of the (secondary) market prices. The inadequate functioning of the stock market mechanism – as a component part of the ensemble mechanism that is specific for the capital market – and eventually the lack of relevance of the stock market prices, can generate errors inside the system and can alter its finality. In this case, the efficient allocation of resources will fail to appear, and this is a pithy specific trace of the capital market. The contact between agents with deficit of money and the ones with monetary surplus can take place in a direct way (direct financing), but also by the means of any financial intermediation form (indirect financing), situation in which specific operators realise the connection between the real economy and the financial market. In this case, the financial intermediaries could be banks, investment funds, pension funds, insurance companies or other non-bank financial institutions. Even if, traditionally, the companies appeared only as agents with deficit of money, in the last two or three decades it could be noticed a change in the financial behaviour of the modern firms: these are not considering anymore the financial market (both the capital and the monetary market) only as sources for rising funds (as issuers of financial assets), but appears more often as buyers of financial assets. The capital market fulfils the transfer function of current purchasing power, in monetary form, from companies which have a surplus of funds to those which have a deficit, in exchange for reimbursing a greater purchasing power in the future; in this way the capital market makes possible to separate the saving act from the investment one.In addition, the capital market mechanism allows not only an efficient allocation of the financial resources available at a certain moment in an economy – from the market’s point of view – but also permits to allot funds according the return and the risk – from the investor’s point of view – offering a large variety of financial instruments with different profitableness-risk characteristics, suitable for saving or risk covering. Nowadays, the protection against financial risks becomes a necessity, imposed by the transformations in the global economy, by the accented instability and the financial crisis that affects without discrimination both developed and emerging stock markets. Covering the risk, that could be realised by the help of different operations, market orders or derivatives, defines the function of insurance against risks, specific function of the capital markets. The capital market allows risk dispersion between investors (of the diversifiable risk), exactly in the same measure in which each of them is willing to assume it, too. From the issuers’ point of view, the money which is necessary for the development or the unfolding of their activity can be mobilised by the help of the capital market at accessible costs, theoretically

speaking smaller than those possibly obtained by the help of the banking market or by other financial intermediaries. From the investors’ point of view, the advantages of the mobilisation process (the advantages of the financial assets), in fact those of the organisation in the form of a joint-stock companies, is under the propriety transfer, which can be rapidly achieved, in a simple and safe way, by the help of the stock exchange, due to the securities’ negotiability. In this context, the stock market facilitates the investments in financial assets, their circulation and mobility and through the offered suppleness sustains, indirectly, the economy’s financing by the help of primary market. The good primary market functioning depends also on the credibility, attractiveness and efficiency of the secondary market. Through the secondary market, a negotiation one, the financial assets owners have the possibility of converting it in cash money. Few persons would invest in securities without the certitude that could transform them in money, at the market price, selling on the stock market. The stock exchange inspires trust to participants, confers legality to operations and trust to obtain the best price at the moment. The stock exchange credibility is sustained also by regulations and assumed deontology rules, in the last years insisting on the ethic dimension of the capital market.The capital market allows the property separation from joint-stock company management, shareholders having the option for this the delegation of a professional team, which can be permanent monitored, and whose mandate will finally be the maximisation of the market value of the company’s shares. By the means of concentrating a large number of stocks, some of the significant shareholders will have the possibility to implicate – directly or through their representatives – in the management of the company that (partially) they owned, this way contributing to the general development of the economy. In fact, the stock market, through the transparency insured and the control possibilities stimulate the management of the listed companies to a better administration. It is well funded the observation that in developing countries we should pay the same attention to the set up and development of an efficient capital market, in the same way in which there are preoccupations for the infrastructure or telecommunications’ development.This thing becomes more important in the transitional countries, taking into account the necessity of resources reorientation from the inefficient sectors towards the efficient ones, being thus ensured the increasing efficiency in economy, being supported both the economic reform process and the privatisation process. The stock exchange is labelled as a “resonance box” for the economical, political and social events, both internal and international, reflecting as a real barometer the state of the economy. It is justified that the stock market cannot be – at least not on the long run) “better” than the real economy, being in fact their reflect. If the real economy absolutely influences the evolution of the stock market, the reverse is also of interest: the flows on the secondary market are useful clues and determinates reorientation of funds to the primary market, reconsideration of options regarding the investments, in a whole, an more efficient allot of the funds available at a certain moment. The impact of the stock market on the real economy is important, because investors from the developed countries are used to watch the market’s evolution and to take into account the information thus offered regarding the economy’s future. The symbolic values, the securities, need a real basis for conferring economic viability; without this, the “virtuous circle” could transform into a “vicious circle”, the capital market is disconnected to the real economy.The importance of the capital market in an economy is given by the significant role played in the firms’ and state’s financing, by the weight of the direct financing among the financing modalities. Beside the apparent important thing – the large transaction volume on the stock market – what it really matters is the place which is taken by the (primary) capital market in the development of the joint-stock companies (direct financing), and this thing is sometimes forgotten, or appears as secondary. In the same idea comes another observation, concerning the danger of exaggerate orienting the attention toward the secondary market, not to the primary one: “[…] the markets and the financial

activity are only modalities for reaching a final objective. The bankers, the brokers and the other professionals in the field aren’t anything else that pipes through which circulate money from the saver to the final user. Often, the pipes tend to assume a greater importance than the two ends of the circuit, utilising various artifices […]”.The well functioning of the capital market is a solid foundation for the insurance of a lasting growth, on a long term, of the national economy; the financial market and first of all the capital market represents in many countries – and it also could represent in Romania, too – the engine for the economic development. For the transitional countries towards the economy market, the stock exchange mechanism was only one of the mechanisms that had to be rediscovered. Although the capital market has suffered profound transformations in the developed countries during the last decades, the modernisation was rather institutional and organisational and less from the specific mechanisms’ point of view. From some points of view, the transitional countries have made many progresses lately with a view to the capital market, especially the secondary one, approaching the level of the countries with tradition in the domain. Unfortunately, the modern infrastructures, computers, the performant telecommunication system, the adequate software, are not enough. In order to have a functioning stock market mechanism, fulfilling its primordial function, we need savings, trust in the economy’s perspectives, and an increasing production.

The capital market in Tanzania can be described as an “emerging market”. Bank financing and government subsidies have for along time been a source of finance for state corporations and companies. There is a noticeable absence of publicly owned companies (i.e., companies allowed to invite subscriptions from the public). Many companies are private, i.e., companies whose right to transfer shares is severely limited. The number of securities is limited, with government debt instruments being the only securities in the market (stocks treasury bills and bonds). Pension and provident funds are the only major collective investment schemes and there are no unit trusts.

In the transition of the country’s economy from a “planned” economy, dominated by parastatal enterprises, towards a “market” economy, where the private sector is expected to play an increasingly important role, the Capital Markets and Securities Act was enacted. As a result the first stock exchange has just started operations in Tanzania.

Foreign Exchange Markets

Tanzania introduced the Interbank Foreign Exchange Market (IFEM) in 1994, initially conducted on an open outcry basis, with authorized dealers assembled at the Bank of Tanzania. The Bank of Tanzania supervised the daily sessions by inviting offers and bids and awarding deals at the highest bid. Telephone dealing was later introduced in May 1996. Authorised dealers are now considering introducing electronic dealing.

The mode of payment is agreed as stated in the contract and indicated in the deal confirmation slip. Either a Bank of Tanzania transfer, cheque or bankers payment (which are cleared and settled like any other cheque in the clearing houses) effects payment.

In the case where transfer is through current accounts maintained at the central bank, entries are passed on due date by debiting or crediting the respective bank’s local currency accounts. Where payment also involves US dollars, the dollar accounts are debited or credited accordingly. Bureaux de change and banks cater for the retail market in which individuals and business satisfy their foreign exchange requirements.

Functions of the capital market

The major objectives of capital market are:

To mobilize resources for investments. To facilitate buying and selling of securities. To facilitate the process of efficient price discovery. To facilitate settlement of transactions in accordance with the predetermined time schedules.

Investment Instruments

Investment is a deployment of funds in one or more types of assets that will be held over a period of time. Various forms of investment are available to an investor. They cover bank deposits, term deposits, recurring deposits, company deposits, postal savings schemes, deposits with non-bank financial intermediaries, Government and corporate bonds, life insurance and provident funds, equity shares, mutual funds, tangible assets like gold, silver and jewellery, real estate and work of arts, etc.

Capital market instruments can be broadly divided into two categories namely

Debt, Equity and Hybrid instruments. Derivative Products like Futures, Options, Forward rate agreements and Swaps.

Debt: Instruments that are issued by the issuers for borrowing money from the investors with a defined tenure and mutually agreed terms and conditions for payment of interest and repayment of principal.

Debt instruments are basically obligations undertaken by the issuer of the instruments as regards certain future cash flows representing interest and principal, which the issuer would pay to the legal owner of the instrument. Debt instruments are of various types. The key terms that distinguish one debt instrument from another are as follows:

Issuer of the instrument Face value of the instrument Interest rate and payment terms Repayment terms (and therefore maturity period/tenor) Security or collateral provided by the issuer.

Different kinds of money market instruments, which represent debt, are commercial papers (CP), certificates of deposit (CD), treasury bills (T-bills) etc.

Equity: Instruments that grant the investor a specified share of ownership of assets of a company and right to proportionate part of any dividend declared. Shares issued by a company represent the equity. The shares could generally be either ordinary shares or preference shares.

Company/Issuer InvestorsSecurities

Major differences between Equity and Debt: Share represents the smallest unit of ownership of a company. If a company has issued, 1,00,000 shares, and a person owns 10 of them, he owns 0.01% of the company. A debenture or a bond represents the smallest unit of lending. The bond or debenture holder gets an assured interest only for the period of holding and repayment of principal at the expiry thereof, while the shareholder is part-owner of the issuer company and has invested in its future, with a corresponding share in its profit or loss. The loss is, however, limited to the value of the shares owned by him.

Hybrids: Instruments that include features of both debt and equity, such as bonds with equity warrants e.g., convertible debentures and bonds.

Derivatives: Derivative is defined as a contract or instrument, whose value is derived from the underlying asset, as it has no independent value. Underlying asset can be securities, commodities, bullion, currency, etc. The capital market in Tanzania is not sufficiently mature or deep enough to support derivatives on equities. Therefore, CMSA does not see, within the period covered by this Strategic Plan, any consideration being given to, for example, equity options, single stock futures, or exchange traded funds. It does see however the possibility of other derivatives such as:

• Index futures or index options;• Interest rate futures;• Currency futures or options;• Commodity futures.

During the life of the Strategic Plan, the CMSA will conduct a study of the feasibility of introducing one or more of the above derivative products. In the meantime however, the Act will be amended to ensure that it caters for all forms of derivative instruments.Futures (Index and Stock): Futures are the standardized contracts in terms of quantity, delivery time and place for settlement on a pre-determined date in future. It is a legally binding agreement between a seller and a buyer, which requires the seller to deliver to the buyer, a specified quantity of security at a specified time in the future, at a specified price. Such contracts are traded on the exchanges.

Options (Index and Stock): These are deferred delivery contracts that give the buyer the right, but not the obligation to buy or sell a specified security at a specified price on or before a specified future date.

Segments of Capital Market

The Capital Market consists of primary markets and secondary markets.

Primary Market: A market where the issuers access the prospective investors directly for funds required by them either for expansion or for meeting the working capital needs. This process is called disintermediation where the funds flow directly from investors to issuers.

The other alternative for issuers is to access the financial institutions and banks for funds. This process is called intermediation where the money flows from investors to banks/ financial institutions and then to issuers.

Primary market comprises of a market for new issues of shares and debentures, where investors apply directly to the issuer for allotment of shares/debentures and pay application money to the issuer. Primary market is one where issuers contact directly to the public at large in search of capital and is distinguished from the secondary market, where investors buy/sell listed shares/debentures on the stock exchange from/to new/existing investors.

Primary market helps public limited companies as well as Government organizations to issue their securities to the new/existing shareholders by making a public issue/rights issue. Issuer’s increase capital by expanding their capital base. This enables them to finance their growth plans or meet their working capital requirements, etc. After the public issue, the securities of the issuer are listed on a stock exchange(s) provided it complies with requirements prescribed by the stock exchange(s) in this regard. The securities, thereafter, become marketable. The issuers generally get their securities listed on one or more than one stock exchange. Listing of securities on more than one stock exchange, enhances liquidity of the securities and results in increased volume of trading.

A formal public offer consists of an invitation to the public for subscription to the equity shares, preference shares or debentures has to be made by a company highlighting the details such as future prospects, financial viability and analyze the risk factors so that an investor can take an informed decision to make an investment. For this purpose, the company issues a prospectus in case of public issue and a letter of offer in case of rights issue, which is essentially made to its existing shareholders. This document is generally known as Offer document. It has the information about business of the company, promoters and business collaboration, management, the board of directors, cost of the project and the means of finance, status of the project, business prospects and profitability, the size of the issue, listing, tax benefits if any, and the names of underwriters and managers to the issue, etc.

The issuers are, thus, required to make adequate disclosures in the offer documents to enable the investors to decide about the investment.

Making public issue of securities is fraught with risk. There is always a possibility that the issue may not attract minimum subscription stipulated in the prospectus. The risk may be high or low depending upon promoters making the issue, the track record of the company, the size of the issue, the nature of project for which the issue is being made, the general economic conditions, etc. Issuers would like to free themselves of this worry and attend to their operations wholeheartedly if they could have someone else to worry on their behalf. For this purpose the companies approach underwriters who provide this service.

Normally, whenever an existing company comes out with a further issue of securities, the existing holders have the first right to subscribe to the issue in proportion to their existing holdings. Such an issue to the existing holders is called ‘Rights issue’. The price of the security before the entitlement of rights issue is known as the cum-rights price. The price after the entitlement of rights issue is known as the ex-rights price. The difference between the two is a measure of the market value of a right entitlement. An existing holder, besides subscribing to such an issue, can let his rights lapse, or renounce his rights in favor of another person (free, or for a consideration) by signing the renunciation form.

The companies declare dividends, interim as well as final, generally from the profits after the tax. The dividend is declared on the face value or par value of a share, and not on its market price. A company may choose to capitalize part of its reserves by issuing bonus shares to existing shareholders in proportion to their holdings, to convert the reserves into equity. The management of the company may do this by transferring some amount from the reserves account to the share capital account by a mere book entry. Bonus shares are issued free of cost and the number of shareholders remains the same. Their proportionate holdings do not change. After an issue of bonus shares, the price of a company’s share drops generally in proportion to the issue.

Activities in the Primary Market1. Appointment of merchant bankers 2. Pricing of securities being issued 3. Communication/ Marketing of the issue 4. Information on credit risk 5. Making public issues 6. Collection of money 7. Minimum subscription 8. Listing on the stock exchange(s) 9. Allotment of securities in demat / physical mode 10. Record keeping

Secondary Market: In the secondary market the investors buy / sell securities through stock exchanges. Trading of securities on stock exchange results in exchange of money and securities between the investors.

Secondary market provides liquidity to the securities on the exchange(s) and this activity commences subsequent to the original issue. For example, having subscribed to the securities of a company, if one wishes to sell the same, it can be done through the secondary market. Similarly one can also buy the securities of a company from the secondary market.

A stock exchange is the only important institution in the secondary market for providing a platform to the investors for buying and selling of securities through its members. In other words, the stock exchange is the place where already issued securities of companies are bought and sold by investors. Thus, secondary market activity is different from the primary market in which the issuers issue securities directly to the investors.

Traditionally, a stock exchange has been an association of its members or stock brokers, formed for the purpose of facilitating the buying and selling of securities by the public and institutions at large and regulating its day to day operations.

Affiliated Organizations

1. Dar Es Salaam Stock Exchange (DSE)2. East African Member States Securities and Regulatory Authorities (EASRA)3. Capital Markets Development Committee (CMDC)4. The International Organization of Securities Commissions (IOSCO)

Dar Es Salaam Stock Exchange (DSE)The Dar es Salaam Stock Exchange was incorporated in September 1996 as a private company limited by guarantee and not having a share capital under the Companies Ordinance (Cap. 212). The DSE is therefore a non-profit organization created to facilitate the Government implementation of the economic reforms and in future to encourage the wider share ownership of privatized and all the companies in Tanzania and facilitate rising of medium and long-term capital.The formation of the DSE followed the enactment of the Capital Markets and Securities Act, 1994 and the establishment of the Capital Markets and Securities Authority (CMSA), the industry regulatory body charged with the mandate of promoting conditions for the development of capital markets in Tanzania and regulating the industry.

East African Member States Securities and Regulatory Authorities (EASRA)

EASRA is made upcomprised of the Capital Markets Authority Kenya (CMA (K)), Capital Markets Authority Uganda (CMA (U)) and the Capital Markets and Securities Authority in Tanzania (CMSA). It was set upestablished on 5th March 1997 after through the signing of a Memorandum of Understanding (MOU) formally constituting establishing a framework for mutual cooperation in the area of capital markets development and with a mission to facilitate the harmonization of securities laws among the East African member states and to promote information sharing and cooperation among the members. EASRA aimsintends to provide mutual assistance to its member states in; the development of capital markets institutions, exchanging information to facilitate the enforcement of their respective laws and regulations, cross border surveillance, public education awareness, co-ordination on technical assistance and promotion of regional consultancies. This is tackled done with the help of through three standing committees which are each charged with the task spelt out under the committee name. These are:

i. Legal Issues Committee ii. Disclosure and Accounting Standards Committee iii. Market Development Committee

The formationstructure of EASRA was grounded based on the premiseprinciple; • That securities markets are fundamental to the development of private enterprises, the mobilization of

savings and investments, the efficientwell-organized allocation of resources, and the promotion of economic growth.

• That international co-operation can facilitatesmooth the progress of the development of effective operation of securities markets

• That increasing internationalization of securities markets adds new and dynamic dimension to the economies of all nations.

• That maintaining effective domestic legal and regulatory structures is essential to market integrity and investor protection.

• EASRA holds meetings every quarter in the three East African countries on a rotational basis.

Capital Markets Development Committee (CMDC)The Capital Markets Development Committee (CMDC) was founded established in 2001 and functions operates under the auspices of the East African Community as one of the standing committees. CMDC membership includes comprises chief executives of the securities regulatory authorities of the member countries and the chief executives of the securities exchanges of the member countries. It phrasesformulates policy, develops and makes recommendation to the council of ministers, regulation and integration of Capital Markets of Kenya, Uganda and Tanzania. The Committee meets as it deems necessaryindispensable to carry out its mandate from time to time and to report to the Council at least once a year.

The International Organization of Securities Commissions (IOSCO)Headquarters in Madrid, Spain, The International Organization of Securities Commissions (IOSCO) has its headquarters in Madrid and has approximatelyabout 140 members. The three chief objectives upon which IOSCO’s securities regulation is grounded based are the protection of investors, ensuring that markets are fair, efficient and transparent and the reduction of systemic risk.The benefits of IOSCO membership includecomprise information exchange at an international level, mutual legal assistance through a multilateral MOU, setting regulatory standards and best practice, resource persons for training, lower registration fees to events and market recognition. Tanzania is a member of IOSCO.The member agencies in the International Organization of Securities Commissions have resolved, through its permanent structures:

• to cooperate together to promote high standards of regulation in order to maintain just, efficient and sound markets;

• to exchange information on their respective experiences in order to promote the development of domestic markets;

• to unite their efforts to establish standards and an effective surveillance of international securities transactions; and

• to provide mutual assistance to promote the integrity of the markets by a rigorous application of the standards and by effective enforcement against offenses.

Securities listed on the stock exchange(s) have the following advantages: • The stock exchange(s) provides a fair market place. • It enhances liquidity. • Their price is determined fairly. • There is continuous reporting of their prices. • Full information is available on the companies. • Rights of investors are protected.

Capital Market Regulation in Tanzania

The Capital Markets and Securities Authority (CMSA) is a Government Agency constituted established to promote and regulate securities business in the countryTanzania. It was establishedrecognized under Capital Markets and Securities Act, 1994.

The legal framework for the regulation of the securities industry is the Capital Markets and Securities Act, 1994 [Act No: 5 of 1994 as amended by Act No: 4 of 1997]. The Act is added onsupplemented by variousa variety of regulations that are promulgated by the Minister for Finance.

CMSA's Mission is “…to design and implement purposeful measures which will enable the creation and development of sustainable capital markets that are efficient, transparent, orderly, fair and equitable to all”.

Rules & Regulations Regarding Capital Markets in Tanzania• Capital Markets And Securities Act [PRINCIPAL LEGISLATION] Acts Nos. 5 of 1994: An Act to establishset

up a Capital Markets and Securities Authority for the purposes of promoting and facilitating the development of an orderly, fair and efficient capital market and securities industry in the countryTanzania, to make provisions w.r.twith respect to stock exchanges, stockbrokers and other persons dealing in securities, and for connectedassociated purposes.

• Capital Markets and Securities (Foreign Companies Public Offers Eligibility and Cross Listing Requirements) Regulations, 2003.: These Regulations may be referredcited as the Capital Markets and Securities (Foreign Companies Public Offers Eligibility and Disclosure Requirements) Regulations, 2003 and shall come into effect from 21st May, 2003

• Capital Markets and Securities (Foreign Companies Public Offers Eligibility and Cross Listing Requirements) Regulations, 2003 [as amended in 2005]: These Regulations facilitates provide for the participation in the capital markets by foreign issuers of securities, in terms of the eligibility criteria and the disclosure needs requirements for such companies to make public offers or cross list at the Dar es Salam Stock Exchange.

• Capital Markets and Securities (Foreign Investors) Regulations, 2003: These Regulations set outspecify the limit of aggregate securities to be held by foreign investors. These Regulations also prescribeset down the manner and conditions under which foreign investors will participate at the Dar es Salam Stock Exchange, the mechanism by which the Authority can monitorkeep an eye on observance of the prescribed limits by the Dar es Salam Stock Exchange SE and Central Depos

• Capital Markets and Securities (The Capitalization and Rights Issue) Regulations, 2000 : These Regulations set outspecify the disclosure requirementsnecessities that an issuer is obliged to comply with during capitalization by way of Rights Issue.

• Capital Markets and Securities (Collective Investment Schemes) Regulations, 1997: To supplementadd-on the Capital Markets and Securities Act, these Regulations make detailedthorough provisions relatingconcerning to the roles of managers, trustees, schemes, trust deeds, pricing, issue and redemption of units / shares and other relevant matters.

• Capital Markets and Securities (Conduct of Business) Regulations, 1997: The Conduct of Business list rules on conduct including comprising inducements, churning, customer right, confidentiality, charges,

execution in addition to the conductmanner of business provisions provided in the Capital Markets and Securities Act.

• Capital Markets and Securities (Advertisements) Regulations, 1997: These Regulations relate to the vetting of securities advertisements by the Authority and it provides for a number of rules and norms conditions that have to be met by advertisers in the securities business. These rules and norms comprise Conditions include the requirement for the content and presentation of the advertisement. The advertisement has to be factual, that is comparison or contrasting of investment should not be done unless it is fair, or it should not contain unfair or otherwiseor else misleading matters or it should not exaggerateoverstate the success or performance of the company.

• Capital Markets and Securities (Accounting and Financial Requirements) Regulations, 1997 : These Regulations provideendow with for the maintenance of accounting records (comprisingincluding audit trail), preparation of the annual financial statements as well of treatment of customer money in accordanceharmony with the law (i.e. in trust for the client). These Regulations were amended in 2003 to compriseinclude provisions for penalties in case of non-compliance on the part of dealers. These Regulations supplement provisions on accounts and audit, which are containedenclosed in the Capital Markets and Securities Act.

• Capital Markets and Securities (Prospectus Requirements) Regulations, 1997: These Regulations supplementadd-on the general provisions on public issues of securities which are contained in the Capital Markets and Securities Act. The prospectus is an importantsignificant document which is requirednecessary where a public offer is being made. The items requirednecessary to be comprised included in the prospectus are listed in the Regulations. These includecomprise matters to be stated in the first page of the prospectus. Others includecomprise information on the right of holders, information on bankers, capital of the issuer, debt of the issuer, any material contracts, the use of the proceedsprofits from the issue etc.

• Capital Markets and Securities (Licencing) Regulations, 1996: These Regulations set outspecify the proceduresmeasures to be complied with by the applicants for licencing for exampleinstance dealers, investment advisers or their representatives. The requisitenecessary application forms are prescribedset in the Regulations. General conditions relating to licences once obtained are also provided for, comprising including the provision that the licence shall be personal to the applicant and the requirementobligation for a licencee to inform the Authority (by written notice) of any relevant alterations.

• Capital Markets and Securities (Establishment of Stock Exchange) Regulations, 1996: These Regulations provideoffer for procedures for the establishmentinstitution of a Stock Exchange. Applications for establishmentfounding of a Stock Exchange are to be made by a corporate body to the Authority, which grants approval subject to certain conditionsstate of affairs, and will continue to regulate the stock exchange once it is approved.

• Capital Markets and Securities (Registers of Interests in Securities) Regulations, 1996 : Certain market players are requirednecessary by the Act to maintain a register in the prescribed form of the securities in which he / she has an interest. These Regulations thereforehence includecomprise the prescribed forms as well as a provision for varying of the form of register by the Authority where necessary. The registers of interest in securities enable transactions to be easily traceable by the Authority and other interested parties thus providing the requisitenecessary transparencylucidity in securities transactions.

• Capital Markets and Securities Authority Enforcement Guidelines, 2004: These Guidelines set outspecify the practices and procedures to be followed by the CMSA while en doing conducting investigations or inquiries where there is breachviolation of the law by market participants or otherwise.

• Capital Markets and Securities Scheme of Service, Staff Regulations and Code of Conduct, 2003 : In February 2003, the Authority approved the Capital Markets and Securities Authority Scheme of Service, Staff Regulations and Code of Conduct to among other things guide the affairs of its staff, carrying outexecution staff procedures and to set out criteria governing salary entry points, the mode of movement from entry point to retirement and the criteria for movement.

• Capital Markets and Securities (Conflict of Interest) Guidelines, 2002: The Guidelines aimintend at offering giving members of the Authority and employees of the CMSA a framework within which to deal with conflicts of interest and other related matters. They are also intendedanticipated to protect members of the Authority and employees of the CMSA against any suggestions that regulatory decisions

have been affected influenced by personal interests or that their investment decisions are made by using insider information. These Guidelines consist of include general principles on conflicts of interest, the policy on employees’ interests, securities transactions by employees and Authority members, treatment of gifts and consequences of default.

• Capital Markets and Securities (Corporate Governance) Guidelines, 2002: These Guidelines aimintend to improve at improving and strengthening corporate governance practices by issuers of securities through the capital markets and promote the standards of self-regulation so as to raise the level of governance in line with global international trends. The Guidelines have been issued in view of the role that good governance has in corporate performance, capital formation and maximization of shareholders value in addition to protection of investors’ rights. The Guidelines apply to public listed companies and any other issuers of securities through the capital markets comprising including issuers of debt instruments.

• Guidelines for the Issuance of Corporate Bonds and Commercial Paper, 1999 : These Guidelines set outspecify the disclosure requirementsnecessities that an issuer is obliged to complyobey with when applying for issuance of a Corporate Bond or a Commercial Paper.

• Capital Markets and Securities (Custodian Securities) Regulations, 2006: These regulations may be seen cited as the Capital Markets and Securities (Custodian of Securities) Regulations, 2006.

Activities in the Secondary Market

1. Trading of securities 2. Risk management 3. Clearing and settlement of trades 4. Delivery of securities and funds

CONSUMPTION AND INVESTMENT

Consumption fulfills individuals’ immediate needs. Sometimes the distinction between consumption and investment is not easy to make. Educational expenditures often are made to increase the quality of life and thus have a consumption element. But education also increases individuals’ ability to produce goods and services and therefore can be, and often is, considered to be production.

Individuals must consume to survive. Individuals consume a variety of products. The types of items consumed and the quantity consumed vary widely from country to country. Consumption also differs from individual to individual even within the same country.

CONSUMPTION WITH A PERFECT CAPITAL MARKET

We consider a simple 2-period world in which a single consumer must decide between consumption c0 today (in period 0) and consumption c1 tomorrow (in period 1). Our consumer is endowed with money m0 today and m1

tomorrow. Consistent with his endowment, the consumer has the opportunity to borrow or lend b0 today at interest rate r.

The equations governing the consumer’s feasible actions today and tomorrow are:

c0 = m0 + b0 (1)

c1 = m1 – (1+r)b0. (2)

It is understood that consumption in both periods must be nonnegative, which puts limits on how much the consumer may borrow or lend today, namely, –m0 ≤ b0 ≤ m1/1+r. After substituting c0 – m0 for b0 in (2), the consumer’s budget constraint is

(3)

We have written the budget constraint as an equality since we shall assume our consumer always prefers more consumption. We interpret the symbol W0 as the consumer’s present value of wealth. Note that consumer’s future value of wealth W1 would be (1+r) W0.

To determine the optimal consumption and borrowing plan, we posit a consumer utility function U(c0, c1) of the form optimal consumption and borrowing plan, we posit a consumer utility function U(c0, c1) of the form

√c 0 ,+ β√c1. As we shall see, the parameter β serves as a discount factor on future consumption. Smaller values

of β imply a larger discount factor on future consumption, which implies that our consumer prefers more consumption today.

Formally, the consumer’s optimization problem is given by:

MAX {U (c0 , c1 ): c0+c1

1+r=W 0} (4)

which may be equivalently expressed as

MAX {U (c0 ,c1 (c0 )) :0 ≤ c0≤ W 0} , (5)

where

c1 (c0 )≔ (1+r )(W 0 – c0) (6)

The form of utility implies that consumption in both periods must be positive so that the optimal choice for c0 necessarily lies strictly between 0 and W0. Accordingly, first-order optimality conditions imply that

0 = ∂ U∂ c0

+ ∂ U∂ c1

dc1

dc0

= 12√c0

−β (1+r )

2√c1

(7)

From (7) the optimal consumption plan necessarily satisfies the condition

√ c1

c0

=β (1+r ) , (8)

or, equivalently,

c1=β2(1+r )2 c0 (9)

Substituting (9) into the budget constraint (3) and solving for c0 and c1 , the optimal consumption plan is

given by:

(10)

(11)

It is important to point out here that the constants ρ0 and ρ1 are independent of present wealth W0; that is, they are known parameters strictly determined by the two discount factors β and r. Note further that the optimal utility is of the form

(12)

Example 1: Suppose m0 = 100, m1 = 990, r = 0.10 and β = 0.60. Then c0 = 0.716W0, c1 = 0.312W0 and U*(W0) = 1.182

√W 0. Here W0 = 100 + 990/1.1 = 1000. Thus c0 = 716, c1 = 312 and U*(1000) = 37.36. Due to the availability of

capital market at which to borrow or loan, the consumer has increased his utility by almost 30% above the level corresponding to the initial endowment U(100,900). To obtain the optimal consumption plan, the consumer must borrow b0 = 716 –100 = 616 today, pay back 678 tomorrow, thereby leaving him with 990–678 = 312 to consume in the final period.

CONSUMPTION AND INVESTMENT WITH A PERFECT CAPITAL MARKET

No we will take into account a world in which the consumer now has the opportunity to invest I0 today in production from which he will receive f(I0) tomorrow. The function f(I0) encapsulates the investment opportunities in production available to our consumer. For example, the consumer may wish to obtain an education while he is young, expecting a return on this investment in his working years. It is generally assumed that f(.) is strictly increasing (more investment leads to more return) but exhibits diminishing returns in that the marginal return on an incremental rise in investment declines as the total investment increases. When f(.) is differentiable, these assumptions imply the first derivative is positive and the second derivative is negative. (Such a function is called concave). To ensure at least some investment would be made (to make our subsequent calculations easier), we also assume that the derivative f’(0) is infinite.

The equations governing the consumer’s feasible actions today and tomorrow are now:

c0 + I0 = m0 + b0 (13)

c1 = m1 +f(I0) – (1+r)b0. (14)

Rearranging terms as we did before, the new budget constraint is

(15)

A close examination of (15) reveals a fundamental property: All consumers, regardless of their utility function, should first determine the optimal investment plan to increase their wealth! That is, they should select the value of I0 such that the marginal return f’(I0) = 1 + r. When the function f(I0) represents the investment opportunities for a firm in which consumers hold stock, then each consumer that holds stock would insist that firm optimize its

investment opportunity, regardless of each consumer’s different desires for consumption today versus tomorrow. Because there exists a capital market for each consumer to borrow or lend, each consumer can redistribute the increase in wealth as they desire. This principle (in various forms) is known as the Fisher Separation Theorem of Finance.

Example 2: Suppose f(I0) = 33 √ I 0 . Now f’(I0) 33/¿ and so the optimal choice for I0 =225. The additional wealth

created through investment equals 195/1.1 – 225 = 225 so that W 0=1225. From (10) and (11) the optimal

consumption plan is c0 = 877 and c1 = 382 with U*(1225) = 41.34. The utility has increased by about 10.7%, which also corresponds to 100(√1225/100 – 1). To obtain the optimal consumption plan, the consumer must borrow b0

= 877 + 225 – 100 = 1002 today, pay back 1103 tomorrow, thereby leaving him with 990 + 495 – 1102 = 382 to consume in the final period.

CONSUMPTION AND INVESTMENT WITHOUT A CAPITAL MARKET

We now consider the situation in which the consumer has investment opportunities as described in the previous section, but no longer has the opportunity to borrow or loan, i.e., b0 = 0. We shall see that without access to a capital market our consumer is far worse off.

The equations governing the consumer’s feasible actions today and tomorrow are now:

c0 + I0 = m0 (16)

c1 = m1 + f(I0) (17)

The new budget constraint may be represented as:

c1(c0) = m1 + f(m0 – c0) (18)

The first-order optimally conditions (7) imply that

√ c1

c0

=β f ' ( I 0 ) , (19)

or, equivalently, that the optimal choice for I0 must satisfy the identity

√ m1+ f (I 0)m0−I 0

=β f ' ( I 0) (20)

After substituting the specific choice for f and performing simple algebra, the optimal choice for I0 must satisfy the identity

990+33 √I 0=9801

I 0

−98.01 (21)

Since the left-hand side of (21) is an increasing function of I0 that is finite when I0 = 0 and the right-hand side of (21) is a decreasing function of I0 that is infinite when I0 = 0, a unique solution exists, which may be obtained by bisection search. (Alternatively, identity (21) is essentially a cubic equation, which has a closed-form solution.) The

optimal value for I0 is about 8.25 with a corresponding consumption plan of c0 = 91.75 and c1 = 1085 with U* = 29.34. The utility has dropped considerably to almost the level corresponding to the original endowment.

MARKETPLACES AND TRANSACTIONS COSTS

While mostthe majority neoclassical economists gave their attention focused during the early part of the twentieth century on modeling production costs in a general equilibrium framework using Cobb-Douglas production functions, Coase (1937) came up with his seminal ideas on the nature of the firm using partial equilibrium analysis. He viewed cited the significance importance of systems attributing transaction cost as one of the key concepts in economic analysis. His ideas along with together with Williamson (1985) built the foundation of what is now known as transaction cost economics with empiricalexperimental works in comparative institutional analysis and management. Transaction cost economics is also consideredwell thought-out to have extensive influence on industrial organization economics and management. The same can be said of contract theory which analyzes economic behavior of agents in resource utilization against the backdrop of tradeoffs among various transaction costs such as information, bargaining, monitoring, and agency costs. Williamson (1985) had defined a transaction as the transfer of a good or service across a technologically separable interface, and yet transaction cost economics researchers continue to ignore the significance of interfaces.An relatedassociated concept is that of modularity. Entities with defined interfaces are known as called “modules” and modularity as a structure of systems has been researched in various fields after the IT industry’s success widelyextensively attributed to the deployment of modularity in industry and product design structures. A module is commonlygenerally defined as a unit sharing multiple interfaces to interact, integrate and combine; but the term interface is referred to in very few studies of modularity. Modularity existing beyondfurther than the boundary of organization can be explained by analyzing its different various interface aspects. As a common general systems concept, modularity has a strong affinity with organization theory and has broad implications vis-à-vis outsourcing and contract manufacturing, the network organization structure in Silicon Valley and so forth. In some studies, findings show that almostapproximately all systems are, to some extentdegree, modular. Without questionNo doubt, automobile manufactures have been using a number of numerous modular components. HoweverNevertheless, various many researchers such as Clark & Fujimoto (1990) focus on their non-modular coherent structure as the most significant important success factor of these consumer durables. There is a indispensable significant requirement need for issues associated with the related to complexity to be further analyzed.In spite ofDespite its ever growing increasing popularity, neverthelesshowever, there exists no perfect accurate definition of modularity yet. Even in the most accepted definition which states that a module is a unit with elements that are “relatively” tightly and coherently coupled/connected inside and “relatively” loosely and weakly outside, that relativity is still shrouded with ambiguity and needs to be expounded further. SubstantialConsiderable transaction costs accrue as interfaces are established and used, and investigating the transaction costs mechanism necessitates requires carefulcautious study of the interfaces involved.We hereby focus on the common characteristicfeatures observed by researchers of both transaction cost economics and modularity which is interchangeabilitysubstitutability, transferability, fungibility, or redeployability, and propose to introduce the notion of interface (transaction interface) to further analyze these issues. Interface is defined as the parameters/media through which transactions are made between two entities. Our aim purpose here is to develop a model for understanding the mechanism of interfaces from the transaction cost approach. In present day’s today’s network era, growth is heavily determineddogged by the ability or capacity of an entity (any individual, organization or economy) to absorb, utilize and exchange resources through interfaces that play importantsignificant roles. We

believe that the necessity ed to study interfaces has been increasing especiallyparticularly with the proliferation of economic activities over the Internet which not only allows high-speed transmission of information but also facilitates collaboration between the entities involvedconcerned in a transaction.Transaction Cost After Coase (1937) laid downestablished a niche for the concept of transaction costs in economics, Williamson and several other scholars researched explored the concept further. “Transaction costs are arguably the most important set of prices in an economy”. In the modern world, it would be difficult to find a rich country where transaction costs sum up to less than half of national income” estimated the overall size of the transaction sector 1970 as about 45 percent of the U.S. gross national product. The concept of transaction costs is not only crucial in economics, but also is regarded considered as the backbone for the explanations of international organizations.The definition of transaction costs still varies have different meanings among different fields and there have been very few empirical studies on comparative measurement or measurement itself due to the problem. Dahlstrom and Nygaard warn that transaction cost economics “cannot survive another 30 years without empiricalpragmatic measurement of the key theoretical element, namely, “transaction costs.”Coase defines transaction costs as “the costs of using the price mechanism, which includes the costs of discovering relevant prices, and negotiating and concluding contracts” and “the costs of resources utilized for the creation, maintenance, use, and change of institutions and organizations” which includecomprise the costs of defining and measuring resources or claims, the costs of utilizing and enforcing the rights specified, and the costs of information, negotiation, and enforcement.According to Williamson (1985), transaction costs include the ex ante costs of drafting, negotiating, and safeguarding an agreement, and the ex post costs of haggling, costs of governance, and bonding costs to secure commitments. Alchian & Woodward asserted that Williamson’s focus is mainlychiefly on contractual issues such as administrating, informing, monitoring, and enforcing contracts, not on market exchange issues such as finding other people, inspecting goods, seeking agreeable terms, and writing exchange agreements. They define it as “the costs incurred in exchange transactions comprising involving the transfer of property rights and contracting transactions involving negotiating and enforcing promises about performance.”Arrow (1969) refers to transaction costs as “the costs of running the economic system.” The following is a running list of definitive statements on transaction costs in different various studies.

“the costs of processing and conveying information, coordinating, purchasing, marketing, advertising, selling, handling legal matters, shipping, and managing and supervising.” (Wallis and North, 1986).

“those costs associated with ‘greasing markets,’ including the costs of obtaining information, monitoring behaviour, compensating intermediaries, and enforcing contracts.” (Davis, 1986).

“the cost of arranging a contract ex ante and monitoring it ex post, as opposed to production costs, which are the costs of executing a contract.” (Matthews, 1986).

“all the costs which do not exist in a Robinson Crusoe economy” as it is difficult to separate one type of transaction cost from another. (Cheung, 1988).

“the costs associated with the transfer, capture, and protection of rights.” (Barzel, 1989). “the costs that arise when individuals exchange ownership rights to economic assets and

enforce their exclusive rights.” (Eggertsson, 1990). “those (the costs) involved in the transfer of goods and services from one operating unit to

another….they usually involve the transfer of property rights and are defined in contractual terms.” (Chandler and Hikino, 1990).

“the costs of defining and measuring resources or claims, plus the costs of utilizing and enforcing the rights specified” when considered in relation to existing property and contract rights; and,

quoting Coase (1960), the definition includes “costs of information, negotiation, and enforcement” when applied to the transfer of existing property rights and the establishment or transfer of contract rights between individuals (or legal entities).” (Furubotn and Richter, 1997).

“the costs of running the systems: the costs of coordinating and of motivating.” (Milgroom and Roberts, 1992)

“the sum of the costs associated with engaging in exchange and contracting activities, which are distinct from the costs of production” (Polski, 2001). The transaction costs at plant, however, are not easily separable from production costs considering that what Eggertsson (1990) designates the costs of production in the neoclassical model are not well defined.

“the costs of acquiring and handling the information about the quality of inputs, the relevant prices, the supplier’s reputation, and so on.” (Vannoni, 2002).

The list above shows indicates that the costs for contract and information are comprises included in all the definitions above, but ordering/invoicing, clearing/settlement, transportation, switching activities such as training/education, integration and management, infrastructure and media have not been clearly specified by any of the above-mentionedaforementioned researchers. Polski (2001) pointed out “North (1990,1997), Wallis and North (1986), and Williamson (1985, 1999) have all argued that transaction costs are embeddedset in in layers of governance structures,” so that it is not easy to separate and define transaction costs explicitly.Clearing and settlement, howevernevertheless, are consideredwell thought-out importantsignificant procedural components of an exchange particularly those comprising involving financial or cash transactions. On contrarythe other hand, ordering tend to have a non-specific nature for the reason that because it is deemed as a kind of contract. Contacts essentiallyfundamentally designate the conditions of exchange, howevernevertheless, these do not necessarilyessentially includecomprise the execution orders of the exchange that are clearly based on the occurrence of ordering and invoicing. The complexintricate nature of transportation has proved to be challenging to previous researchers becausefor the reason that physical flow seems to be separabledivisible from information flow. Cheung (1998) excluded the transportation costs and production costs from transactions costs because he considered these to be physical while Wallis & North (1986) included the transportation costs as “shipping costs.” If we define an exchange as a transfer of goods or services and related information, it stands to reason that a transaction may be defined as an exchange with a corresponding order of sequential activities required to complete the exchange. Grounded Based on this specificationcondition, our concept of transaction comprises includes transportation by definition. MoreoverFurthermore, it attributes transportation costs to “site specificity” which plays an importantsignificant role in transaction cost economies.In additionAdding together, information flow in transportation such as receiving, tracking, expediting, data processing, accounting, etc., are quite often embedded in the transportation itself and not separable. The costs incurred in moving cargo or changing the location coordinates of plant infrastructure, whether through manual or mechanical means within organization, are most often reflected in management costs. In some instances, the physical transfer or logistical movement of a good signifies the completion of an exchange and in that case, transportation is considered an information flow as well. Physical flow of money is a good example. Some transportation costs are inseparable from the well-accepted nation of “transaction costs”. Furthermore, it is an important cost factor to be considered with regard to the decision of executing transactions and should be given careful consideration.The switching costs related to such activities as training/education, management and business process integration for a newly deployed good/service influence decision on transaction execution as a cost factor and, for that reason, need careful analysis as well. Furubotn and Richter (1977) called attention to political transaction costs and managerial transaction costs along with market transaction costs. Political

transaction costs include “the costs of setting up, maintaining and changing a system’s formal and informal political organization” and “the costs of running a polity”. Managerial transaction costs reduce “the costs of setting up, maintaining or changing an organizational design” and “the costs of running an organization.” Williamson (1985) also included accounting and control systems as well as monitoring systems. The switching cost is considered to be an internal enforcement cost while costs incurred in enforcing compliance measures involving external entities are widely-accepted to be included in transaction costs. Therefore, costs of this nature should be included in the transaction costs.Infrastructure and media are indispensable for the transfer of information and physical objects to occur. For instanceexample, contracts may be sent via the post under the pay per use setup. Alternatively, information can be exchanged via the internet, by using a monthly leased dedicated line, or even by building one’s own communication line or network. Investing in the acquisition of telecom equipment such as digital telephone switches, routers, servers, etc. is often required. In the case of pay per use, it is easily measurable as a transaction cost but the allocation of infrastructure cost, such as durable goods and equipment as well as media cost are embedded and not easily distinguishable. The same can be said of management costs which are highly intangible and are most often dealt with in conventional accounting systems as part of overhead or as a bulk of miscellaneous costs. Most of these are often embedded by nature but comprise involve significantimportant friction factors that merit careful analysis.So we recognize transaction costs as all the costs incurred needed to complete exchange related with the transfer of a physical object and information, including the costs of ordering/invoicing, clearing/settlement, transportation, switching activities such as training/education, integration, and related management, and infrastructure and media. Money transfer is divided into and included in physical and information transfer.TRANSACTIONS COSTS AND THE BREAKDOWN OF SEPARATIONIf transaction costs are nigglingnontrivial, financial intermediaries and marketplaces will providesupply a useful service. In such a world, the borrowing rate will be greater than the lending rate. Financial institutions will pay the lending rate for money deposited with them and then issue funds at a higher rate to borrowers. The difference between the borrowing and lending rates represents their (competitively determined) fee for the economic service provided. Different borrowing and lending rates will have the effect of invalidating the Fisher separation principle. The theory of finance is greatlyvery much easy to understand simplified if we assume that capital markets are perfect. ObviouslyEvidently they are not. The relevant question then is whether the theories that assume frictionless markets fit realityactuality well enough to be useful or whether they need to be refinedpolished in order to providegive greater insights.

SUMMARY

The objective of this chapter was to study consumption and investment decisions made by individuals and firms and to understand the role of interest rates in making these decisions. The decision about what projects to undertake and which to reject is perhaps the single most important decision that a firm can make. Logical development is facilitated if we begin with the simplest of all worlds, a one-person/one-good economy with no uncertainty. The decision maker must choose between consumption now and consumption in the future. Of course the decision not to consume now is the same as investment. The decision is simultaneously one of consumption and investment. In order to decide, he needs two types of information. First, he needs to understand his own subjective trade offs between consumption now and consumption in the future. This information is embodied in the utility and indifference curves. Second he must know the feasible trade offs between present and future

consumption that are technologically possible. From the analysis of an economy we will find that the optimal consumption/investment decision establishes a subjective interest rate for the individual and the economy as a whole.

REVIEW QUESTIONS

1. Jones is endowed with money m0 = 55,000 today and m1 = 88,000 tomorrow. He desires to consume c0 = 80,000 today and c1 = 66,000 tomorrow.(a) If there is no opportunity to borrow or lend can Jones achieve his consumption objective?

Explain?(b) Suppose there is a perfect capital market in which Jones may borrow or lend as much as he

desires at the market interest rate of 10%. Can Jones now achieve his consumption objective? Explain.

(c) Suppose that in addition to a perfect capital market Jones has an opportunity to invest I0 = 100,000 today and receive f(I0) tomorrow. Determine the minimum value for f(I0) for which Jones will be able to exactly achieve his consumption objective.

(d) Explain exactly what Jones must do using the capital market and investment opportunity available to him so that he may exactly achieve his consumption objective.

2. Jones is endowed with money m0 = 40,000 today and m1 = 99,000 tomorrow. He desires to consume c0 = 100,000 today and c1 = 55,000 tomorrow.(a) If there is no opportunity to borrow or lend can Jones achieve his consumption objective?

Explain.(b) Suppose there is a perfect capital market in which Jones may borrow or lend as much as he

desires at the market interest rate of 10%. Can Jones now achieve his consumption objective? Explain.

(c) Suppose that in addition to a perfect capital market Jones has an opportunity to invest I0 = 55,000 today and receive f(I0) tomorrow. Determine the minimum value for f(I0) for which Jones will be able to exactly achieve his consumption objective.

(d) Explain exactly what Jones must do using the capital market and investment opportunity available to him so that he may exactly achieve his consumption objective.

3. Smith’s utility function is U(c0, c1) = ln c0 + 0.4 ln c1. Smith is endowed with money m0 = 90,000 today and m1 = 500,000 tomorrow. There is a perfect capital market for borrowing and lending at the market rate of interest of 25% per period.(a) Determine the optimal consumption plan for Smith. Explain exactly what Smith must do

each period to achieve his optimal consumption plan. (Recall that ddx

∈x=1/ x)

(b) Smith has an opportunity to invest I0 = 80,000 today and receive f(I0) = 135,000 tomorrow. Should he take advantage of this opportunity? If not, explain why not. If so, explain exactly what he should do each period to achieve his new optimal consumption plan.

(c) If Smith no longer has access to the capital market (he cannot borrow or lend), then should he take advantage of the investment opportunity presented in (b)? Explain your reasoning.

4. Suppose that the investment opportunity set has N projects, all of which have the same rate of return, R*. Graph the investment opportunity set.

5. Suppose your production opportunity set in a world of perfect certainty consists of the following possibilities.

Project Investment Outlay

Rate of Return (%)

A TZS1,000,000 8B 1,000,000 20C 2,000,000 4D 3,000,000 30

(a) Graph the production opportunity set in a C0,C1 framework.(b) If the market rate of return is 10%, draw in the capital market line for the optimal

investment decision.

CASE STUDY

Transaction Costs and the Structure of the Market

Introduction

The market is a central building block in the edifice of modem economic theory. Usually pictured as the demand curve crossing the supply curve, the market in standard economics textbooks is free of any institutional structure. Simply reduced as a price-making mechanism, the market serves more as a theoretical construct than as a characterization of the actual process of exchange. Recent development in new institutional economics has drawn much attention to empirical studies of economic institutions. But the emphasis is largely laid on the institutional structure of production, particularly business firms. The institutional structure of exchange is unfortunately overlooked. What Ronald Coase bemoaned more than a decade ago still remains a disturbing reality, "Although economists claim to study the working of the market, in modern economic theory the market itself has an even more shadowy role than the firm" (1988a, 7).

Overlooking the institutional structure of the market leads economists to take the existence of the market for granted. Nothing more than "the law of supply and demand" is required for economists to reach a conclusion on their blackboard that the price mechanism (i.e., the market) is at work. The economists hence seldom bother to investigate how the market as an economic institution emerges and develops in the real world. Even new institutional economists who are committed to studying real life economic institutions are inclined to assume that "in the beginning there were markets" (Williamson 1975, 20; 1985, 87). A casual perusal of the real world, however, makes it clear that this view does not square with the reality. No market is not deliberately created or costly to maintain, be it the Chicago Mercantile Exchange or a Chinese fish market. Taking the existence of the market for granted makes us blind to underlying economic forces that shape the institutional structure of exchange.

Over the past couple of decades, one of the most influential concepts brought into economists' tool-kit is transaction cost (Coase 1937). Admitting the existence of transaction costs helps economists move out of their imaginary world and come down to the reality. In the real world, we can not help but realize that transaction costs are ubiquitous. Without transaction cost as the tool of the trade, it is no wonder that economic analysis in the past was out of touch with real world economic activities. Ever since the introduction of transaction costs, economists have made considerable progress in understanding real world economic institutions, particularly in the field of business firms. It is surprising however that the market--another economic institution of equal if not more significance--has received scant attention.

This paper applies the same approach that has proved to be quite productive in the study of the firm to investigate the working of a real world market. Based upon the development of fish markets in a Chinese fishery community, this paper directs our attention to the institutional structure of exchange. It empirically shows that not only the institutional structure of the market is a response to the cost of carrying out transactions, but the very existence of the market is contingent upon entrepreneurial calculations. In doing so, this paper advances our understanding of the choice between the firm and the market in organizing economic activities for the sake of reducing transaction costs.

The Network of Fish Markets

The fieldwork for this case study was done in the Long Lake area, Hubei province, China. Well known in China as the "land of fish and rice," this area abounds with lakes, ditches, and ponds, which provide favorable natural conditions for freshwater fishery. In spite of rich natural resources, the development of a large scale freshwater fishery did not start until the late 1970s when rural economic reforms were initiated. In the pre-reform era, fishing was exclusively for self-consumption because private commercial activities were largely banned. Under the socialist planned economy, there were virtually no market transactions. In the wake of rural economic reforms, the ban on commercial activities was lifted. "Rural petty commodity markets" were encouraged by the reform policy to cater to the increasing needs of the rural population. The opening of fish markets quickly commercialized fishing and peasant-fishermen began to sell their catches in markets for cash. High profits in fishing stimulated the steady growth of freshwater fishery, which in turn pushed for further market integration.

Fish markets in this region are connected and organized into a hierarchical network. At the bottom of this hierarchy are fishing grounds (e.g., lake shores or fishponds), which are drawn as squares, where fishermen sell their catch. At the top are fish markets in Jingzhou, the prefecture city in this region, where city residents and other consumers buy fish. These fish markets are represented as a big rectangle. In between are intermediate markets, which are located in local towns more or less away from Jingzhou. In the figure, intermediate markets are drawn as ellipses. In this network, the flow of fish starts at the bottom, i.e., fishing grounds, and ends up at the top, i.e., fish markets in Jingzhou. The transportation of fish from the fishing grounds to the final fish markets is carried out by middlemen whose entrepreneurial activities of buying cheap and selling high help to bridge spatially distanced producers and consumers.

The group of middlemen is heterogeneous. Local terms define two types of middlemen, "moving middlemen" and "sitting middlemen." A sitting middleman has a fixed shop in a market. He buys fish in wholesale and retails to individual consumers. Because city residents are the primary group of consumers, most sitting middlemen are in the city. And there are few sitting middlemen at intermediate markets. A moving middleman travels to fishing grounds or intermediate markets to buy fish and sell to other moving middlemen or sitting middlemen in the city.

A moving middleman moves back and forth between different marketplaces, and that is how he earns his name of "moving" middleman. Because fish transportation is a rather arduous task, many moving middlemen work in small groups, which usually consist of two or three middlemen.

To introduce this network of markets, we start with fish markets in Jingzhou. As the biggest city in this region with a population of 330,000, Jingzhou has many permanent "vegetable and food markets" in its residential districts where city residents and retailers freely bargain and make exchange transactions. The marketplace is usually housed in a big hall, separated into many sections. Each section is reserved for one type of goods, such as rice, vegetables, fruit, pork, beef, chicken, or fish. The separation of the marketplace into specialized sections certainly helps consumers reduce their search time when they do shopping. In each section, there are many retailers selling the same type of (certainly not identical) goods and competing against each other for customers. A retailer rents a numbered shop place from the administration office in charge of the marketplace. In most marketplaces, what a retailer rents is actually not a shop, but a counter or just a small area of the floor in the market hall to dis play their goods. After paying a fixed amount of monthly sales tax, the retailer is then licensed to operate. In addition to retailers with fixed shops, there are also occasional retailers. These retailers sporadically sell goods in the market, and do not have a fixed shop place. Occasional retailers are usually peasants who sell their extra agricultural products after consumption. Though entry to the market is not restrained, occasional sellers are required to pay sales tax. However, the amount of tax is not clearly defined, but subject to the arbitrary decision of the market administration office. As a result, tax avoidance on the part of occasional sellers is the rule rather than exception. When they get caught, their goods are usually confiscated or destroyed if the administration officials refuse to take bribes.

In the marketplace, the area of fish shops is usually wet and filthy. Since fresh fish taste much better, most fish are sold alive in the market. Dead fish have a much lower market value. To be kept alive, fish are usually stored in an open container filled with water. Fish of the same species are sorted into one container. Some sellers even build a large container with bricks and keep all their fish together. Depending on the way (i.e., what equipment was used) that fish were captured and transported, as well as their species, fish can be kept alive in a container for one to several days. In a large brick-built container, which is usually equipped with running water, fish can live much longer.

Though the marketplace remains unlocked overnight, the daily schedule of the market starts around five in the morning when first retailers arrive, carrying their goods. After their arrival, retailers begin to display their goods on the counter, waiting for customers. The first customers usually arrive around six. In the beginning, there are not many shops open. Frugal customers usually visit all the open shops one by one, haggling around, and then buy from the one offering the best price. The price differs considerably between shops in the beginning, even for fish of the same species, size and quality. The price information is then communicated to each one of retailers by price-seeking customers. After a while retailers begin to adjust their price levels to attract customers. And the price levels gradually converge. However, this does not mean that the price level stays the same. As more and more customers visit the market, the price gradually increases at almost the same pace across the whole market. In general, the fish price in the morning (between seven and nine o'clock) records the highest. This is not without reason. From the demand side, the morning market has the most local consumers, since most city residents do their shopping before they go to work at eight o'clock. On the supply side, fish are most fresh. After eight, the number of customers decreases sharply and the fish price drops accordingly. After noon, there is a considerable drop in fish price, usually about ten to fifteen percent, partially because some fish can no longer be kept alive and are sold cheap. Prior to the close of the market in the early evening, retailers are eager to sell out their remaining fish, especially those that can not live to the next day, so that the fish price is usually the lowest at this time.

Compared with the market in Jingzhou, the intermediate market in the local town is much smaller, in terms of both its physical size and the number of participants. In general, the larger the town population and hence the consumer group, the bigger the marketplace. The majority of the marketplaces in towns are outdoors. The main street of the town usually serves as the temporary marketplace. Sellers just display their goods along the street, haggling with people passing by. These markets are usually dominated by occasional sellers. There is virtually no administrative restriction, and sales tax is rarely charged. Due to their small size of final consumers, most fish in these markets are bought by middlemen who then transport the fish to Jingzhou or other intermediate markets closer to Jingzhou. On the side of fish sellers, some of them are fishermen in nearby villages who carry their catch to the town market. The others are middlemen who bought fish from fishermen on the fishing grounds or other intermediate markets more far away from Jingzhou and do not want to transport fish any further toward Jingzhou.

Fishing grounds serve as the wholesale spot market where middlemen compete aggressively with each other and haggle hard with fishermen to make a good deal. There are two types of fishing grounds, corresponding to two types of fresh water fishery. In the case of lake fishery, there are many fishermen fishing in the lake every day. But the catch of each fisherman is limited. A middleman usually has to buy fish from several fishermen. In the case of pond fishery, the output of one fishpond is often too large for one middleman to handle. Yet, the information about which pond is going to sell fish becomes highly valuable because only a few number of fishponds sell fish on a given day.

In each fish market, when there are many sellers and buyers, the market works quite competitively. Though parties on two sides may know each other quite well after repeated dealings with each other, personalistic trade or clientation between trading parties is not prevalent. Not a single customer is attached to a particular seller. The customer buys from whoever offers the best deal. Nor does a fisherman sell to a particular middleman. Whoever has the best offer gets the deal. In most cases, the price mechanism prevails. However, this is not to say that non-price mechanisms have no role to play. On the contrary, reputation and personal connections matter a great deal for the success of a middleman. A sitting middleman who is viewed as dishonest can hardly survive. Similarly, a moving middleman without a wide personal network may find it difficult to find enough fishermen from whom he can buy fish.

These geographically separated fish markets are linked together by middlemen, whose buying and selling make the fish market work. In this hierarchical network of fish markets, final markets in Jingzhou are where fish reach final consumers. The price of fish at the final markets, which is determined by the supply and demand conditions, has a strong radiating effect on all fish markets down below, which adjust their price levels accordingly after taking into consideration the cost of transporting fish to Jingzhou. In general, the closer the market to Jingzhou, the higher the fish price. Since there is no special institution to collect and distribute price information across markets, this job is mainly performed by middlemen who travel up and down along the chain of markets. Yet, we can not take it for granted that middlemen will always truthfully reveal the price information to fishermen from whom they buy fish. Instead, the middleman is inclined to take advantage of fishermen who are poorly informed by under- reporting the price in the city and bargaining tough in order to buy as cheap as possible. But when more and more middlemen visit the fishing ground, competition forces them to more candidly disclose their private price information, and this keeps the revealed price closer to the real one in the city.

In this hierarchical network of the fish market, the role played by middlemen is quite obvious. Spatial distance between consumers and producers makes their face-to-face transactions impossible to execute. Even when the price that consumers are willing to pay is much higher than the price at which producers are willing to sell, transactions may not occur and potential trade surplus can not be materialized. This physical distance is reduced, inasmuch as middlemen travel back and forth between producers and consumers. Their buying from producers at

a price of p1 and selling to consumers at a price of p2 (p2 [greater than] p1) brings about the market for producers and consumers, saving the cost for both to locate and meet each other. In addition to physical distance, there may exist a time lag between the demand on the side of consumers and the supply on the side of fishermen. That is, when the consumer has the demand for fish, the fisherman may not have fish available. Likewise, when the fisherman wants to sell, there may be no demand on the side of consumers. The middleman coordinates the demand and supply by storing fish when the supply exceeds the demand and releasing fish when the situation is reversed. Lastly, the middleman is the channel of communication between the fisherman and the consumer, and between the supply and demand of different time periods. Markets separated by geographical and temporal distance are thus connected by middlemen.

The working of the market requires the combination of moving middlemen and sitting middlemen. Moving middlemen buy fish from fishermen at fishing grounds, transport them to city markets, and sell them to sitting middlemen. A moving middleman is not necessarily involved in the whole process. Indeed, the division of labor among moving middlemen exists. Some moving middlemen are specialized in buying fish from fishermen on fishing grounds, utilizing their knowledge of local fishermen. Usually, fish of various species and sizes are bought from fishermen. These fish are then sorted according to their species because fish of different species and sizes are priced differently. These sorted fish are then sold to other moving middlemen who may be specialized in trading fish of certain species and sizes. This differentiation is partly due to the fact that different species usually need different transportation equipment. A sitting middleman is a permanent retailer in the city market. Transactions between sitting middlemen and consumers mark the end point of the fish flow.

This picture intuitively leads us to conceive the market as a continuum of middlemen, like a line of continuum of points. This continuum starts with the middleman who buys from the wholesale spot market on fishing grounds, and ends with the sitting middleman in the city market who retails fish to consumers. The middleman is specialized and each occupies a particular point on the continuum [9] (more middlemen may be located at the same point and competing against each other). A middleman at one point of the continuum buys from the middleman at the preceding position, and sells to the middleman at the succeeding position. Continuous transactions between these middlemen constitute the market. [10] But each point on the continuum, i.e., an individual middleman or a group of middlemen, is a "firm" if by the term we mean "a specialized unit of production" (Demsetz 1995). Every middleman adds certain values to his input and transforms it to his output, as reflected by the fact that for each middleman the selling price is higher than the buying price. The differentials of the market values between input and output (which equals to the product of quantity of the fish transported and the price spread, i.e., p2 - p1) are the added value. When it is greater than the cost associated with conducting two transactions, i.e., buying and selling, and transportation, the firm makes a profit. Otherwise, the firm gets out of the business. When none of the middlemen at one point of the continuum can cover operational costs, the chain of middlemen is disconnected, and the market fails to rise.

The Institutional Structure of Exchange

The single most important contribution to bring our attention to the institutional structure of the economy is Coase's seminal article (1937), in which Coase introduces the concept of what would be later called transaction costs to explain the emergence of the firm. The explosive literature on business firms developed over the past two decades attests the power of Coase's simple idea.

This approach is of no less relevance to the study of the market. The market exists to reduce transaction costs and facilitate economic transactions, when consumers and producers can freely meet each other to execute transactions at no cost, the market as described above does not exist. In the world of zero transaction cost, as

neoclassical economics has nicely demonstrated, the market becomes a price-determining mechanism, and its institutional setting becomes irrelevant. The ubiquitous existence of transaction costs means that the institutional structure of exchange matters. In our case, how does the institutional structure of fish markets that we have described in the preceding sections help to reduce transaction costs?

First of all, the emergence of the middleman-made network of fish markets is a response to the physical distance between fishermen and fish consumers, which poses severe transaction costs for fishermen and consumers wishing to execute transactions directly. Entrepreneurial efforts of the middlemen provide a convenient interface for fishermen and consumers to communicate with each other and make transactions.

Next, keeping in mind the role of the institutional structure of the market in reducing transaction costs, we examine the network structure of fish markets in some detail. Since Jingzhou is the primary consumer market, with numerous anonymous retailers and customers trading with each other, permanent marketplaces were constructed by the city government. In addition to the provision of shelters, the involvement of government is more clearly indicated by the existence of the market administration office. It not only enacts a set of specific rules regulating the behavior of retailers and customers, but also provides third-party enforcement. Both measures have tremendously facilitated market transactions, especially from the perspective of retailers. Third-party enforcement greatly helps retailers secure their property rights.

The market administration office also plays an important role in deterring cheating behavior on the part of retailers, who usually have more information about their goods than consumers. Fraud on the part of retailers usually takes the form of giving short measure or charging high price for low quality goods. Whenever a retailer is caught defrauding, the customer can report the retailer to the market administration office. Of course it is costly for the consumer to detect such fraud on the spot. Moreover, many attributes of a commodity can only be detected after consumption. Spending time on searching may enable the consumer to find the lowest price in the market. But "searching" alone can not help the consumer detect certain qualities of the fish. Only by "experience," i.e., consumption, can the consumer detect whether the fish meets the quality that the retailer claims. This renders third-party enforcement ineffective because the information that the consumer acquires ex post can not be used as evidence for the third-party to judge whether the retailer defrauds.

A notable feature of the city market provides a remedy for this problem. In the city market, a permanent retailer has a fixed shop place, and is thus easily identifiable. This means that reputation can work well as an informal mechanism to check against fraud on the part of retailers. In order for reputation to work, several conditions must be met. First of all, the game is played an infinite number of times (or it is terminated probabilistically). In repeated transactions players' long-term interests outweigh short-term gains. Hence they restrain themselves from opportunistic behavior in order to cultivate good reputation. In a one shot game, reputation does not matter. Secondly, cheating behavior in one round can be easily detected before the next round of the game begins. Thus, the cheated customer is able to punish the dishonest retailer by switching to other retailers. Thirdly, when many individuals are involved in the game, the one with a bad reputation is easily identifiable. As a result, once detected, the defrauding retailer can be readily recognized not only by the cheated consumer, but also by other consumers who are informed of the bad reputation of the dishonest retailer. In the setting of a city fish market, all these conditions are satisfied. A permanent retailer has a long-term stake in the marketplace. With a fixed shop place, he is readily identifiable among other retailers. Although it is costly for the customer to detect frauds on the spot when shopping in the market place, frauds can be easily discovered after the purchase. Accordingly, for a permanent retailer, it is in his interest to maintain a good reputation and behave honestly.

Transactions in intermediate markets and on fishing grounds are fairly free from government regulation. No special physical facilities are provided by the government, nor is any specific rule enacted or enforced. To use North's term (1990, 34-35), transactions in these markets are a mixture of "personal exchange" and "impersonal exchange without third-party enforcement." Nonetheless, these markets work quite well to meet the needs of both sellers and buyers. Several factors contribute to keeping the market in place. First, in each of these marketplaces, the number of sellers and buyers is much smaller. Market participants know each other's identities well. Exchange transactions proceed in a much smaller scale and in a local setting. Even in these markets, however, personalistic trading seldom occurs. Long-term contract between middlemen and fishermen is rare. In addition, fish transactions in intermediate markets and on fishing grounds are mostly conducted in the form of wholesale. large stakes give strong incentives to both sellers and buyers to carefully check the fish before making the deal. Furthermore, most transactions in these markets are between fishermen and middlemen. Both groups are equally experienced in dealing with fish. The problem of asymmetric information, if it exists at all, is not significant. In these markets, defrauding rarely occurs on either side of the transaction.

Another notable feature of fish markets is also an efficient adjustment to the cost of transaction in the market. We have mentioned that middlemen are differentiated into two groups, moving middlemen and sitting middlemen. Why don't moving middlemen sell to consumers directly?

Their reluctance is attributable to the cost of entering into the city market. Transaction costs come from several sources in the city market. First, if a moving middleman sells fish to consumers directly at the city market, he must then pay sales tax to the administration office of the marketplace. However, the amount of the tax is not defined by any rule, but subject to the arbitrary decision of the market administration office. This uncertainty greatly discourages middlemen from moving into retailing, that is, selling fish to consumers. Second, without a fixed shop, occasional sellers are viewed by consumers as less trustworthy because the mechanism of reputation does not work. Without a long-term stake in the market, the occasional seller can not "signal" his honesty, as the permanent seller can. Third, retailing inevitably takes a longer time than wholesaling, and requires special investments (e.g., containers) to keep fish alive. This increases the operation cost for the moving middleman if he engages in retailing. Fourth, the space in the fish section is rented to permanent retailers who usually do not allow occasional sellers to sell fish around their shops. As a result, occasional sellers are usually forced out of the fish section. All these factors push moving middlemen away from selling to consumers directly and keep them selling fish wholesale to sitting middlemen. At the same time, the specialization of moving middlemen allows them to know more thoroughly the marketplaces which they usually visit, and makes them well known in these marketplaces. Both are prerequisites for their success.

To summarize, the institutional structure of the market helps reduce the cost of carrying out exchange transactions. In local markets (i.e., intermediate markets in local towns and wholesale spot markets on fishing grounds) where the identities of people on both sides of transactions are known, the institutional structure of exchange is deeply embedded in and undifferentiated from the local social structure, which provides reliable institutions and helps to lessen any uncertainty involved in transactions. In the city market, the government has a comparative advantage in securing transactions. Indeed, the state is active in regulating market transactions, and this state regulation immensely eases both parties in their commercial transactions.

However, the provision of the market itself is costly. At the very least, for example, some physical facilities have to be provided. Moreover, certain rules governing exchange, such as those regarding property rights and contract, have to be in place and enforced one way or another. The government plays an important role in reducing the cost of providing the market. For instance, the public provision of permanent marketplaces and the existence of the market administration office in Jingzhou greatly reduce transaction costs for both buyers and sellers. Moreover,

the general legal code is vital to protect property rights and enforce the law of contract. Actually, the market obviously has certain features of public good because it can generate strong positive externalities once it is in place. For example, the cost of looking for trading partners is greatly reduced. However, this is not sufficient to justify the public provision of the market. It is true indeed that the market cannot be made and put into p lace by the government alone. The working of the market is crucially dependent on incentives of market participants. That is, the provision of the market requires entrepreneurial efforts. When the costs of making the market shouldered by market participants exceed their potential gains, the market will not be provided. This point is nicely illustrated in the following section where we compare fish marketing at two villages.

Fish Marketing at Two Villages

Dong's Village (DV) and Wens Village (WV) are two fishing villages on the north side of Long Lake. The lake and its many tributary ditches, canals, as well as ponds along the inside lake shore provide peasants at both villages with abundant resources for freshwater fishery. The past two decades of rural reform have encouraged the steady growth of fishery at both villages. Yet, they have developed quite different marketing channels. DV abounds with middlemen who buy fish from fishermen on fishing grounds and transport them to intermediate fish markets or Jingzhou. Contrarily, at WV, there are few middlemen and fishermen transport their fish across Long Lake to fish markets themselves. In other words, fishermen at WV expand their scope of their primary activities along the "value chain" of fishery and get involved in marketing. The forward expansion of fishing "firms" at WV contrasts squarely with the blooming of market at DV. If the provision of the market is subject to economic analysis, this difference suggests that the cost of establishing the market is so high at WV that the market fails to emerge. What makes the cost of establishing the market so different between DV and WV?

With this question in mind, we take a brief look at the two villages. Both DV and WV are on the north side of Long Lake. They are separated from each other by Wang's Bay, an inlet of Long Lake. Wang's Bay is the county borderline between Jiangling and Jingmen. DV is on the west side of the bay and belongs to Jiangling county; WV is on the east side and belongs to Jingmen county. Since final fish markets are located in Jingzhou, fish in both villages have to be transported across Long Lake and carried to Jingzhou one way or another. Long Lake has three ports for passengers traveling across the lake. Port A is on the south side shore. Ports B and C are on the north side. Port B is in DV; Port C in WV. Port A is 20 minutes away from Jingzhou by bus, and there is convenient public transportation between Port A and the city. Close to the highway linking Jingzhou and Wuhan, the capital city of Hubei Province, Port A has the most important intermediate fish market in this area.

Comparing the two routes of fish transportation at DV and WV, we realize that the distance between Ports A and B is much shorter than the one between Ports A and C. This factor gives rise to the considerable difference of transportation costs, which turns out to be crucial in determining the difference of marketing channels between these two villages.

Water transportation across Long Lake is mainly by passenger ship. There are two transportation routes across the lake, the one between Ports A and B, and the other one between Ports A and C. The distance of the first route is about three kilometers and it takes 15 minutes; while the distance of the second one is 15 kilometers and it takes a little more than one hour. Moreover, the passenger ship of the first route travels more frequently. In a normal day, the ship travels back and forth between Ports A and B for as many as 20 times, from six in the early morning until later afternoon. Contrarily, the ship on the second route usually travels four times a day. The much shorter physical distance between Ports A and B is certainly an important factor. Moreover, since DV is in jiangling county, compared with WV, it has more intensive connections with Jingzhou and other towns on the south side of the lake which also belong to the same county. For example, a lot of villagers from DV work in Jingzhou. Peasants in DV

have relatives on the south side of the lake. In addition, intensive interactions between DV and Jingzhou and busy schedules between Ports A and B reinforce each other: more interactions demand more frequent shipping schedules; frequent schedules make interactions more convenient between the two sides of the lake. In a similar way, scant interactions between WV and the south side of the lake require fewer schedules between Ports A and C, which in turn discourage intensive interactions to develop. As a result, longer distance and much fewer schedules between Ports A and C make transportation between Ports A and C much more difficult. This considerably increases the transportation cost that the middleman has to incur if he buys fish from WV.

Moreover, as far as fish transportation is concerned, there is another factor working against the fishermen at WV. In the process of fish transportation, caution must be taken to keep fish alive. This factor makes a huge difference in terms of how fish are transported. The average time of transporting fish from DV to Port A is about 15 minutes. Within this period of time, no special equipment is needed to keep fish alive. Middlemen most often simply use a pair of bamboo baskets to carry fish from DV. On the other hand, the average time of transporting fish from WV to Port A is more than an hour. For such a long period of time, the middleman has to use large containers, such as wooden buckets, to keep fish alive.

High transportation costs make WV less attractive to middlemen, but this does not necessarily prohibit them from going there. If a middleman is able to buy fish at a sufficiently low price at WV so that high transportation costs can be compensated, there is no reason for him not to visit WV. Yet this scenario does not unfold. On the contrary, when the number of middlemen visiting WV is limited, bargaining between them and fishermen becomes hostile and readily breaks up, long before reaching a good deal.

The failure of bargaining between middlemen and fishermen at WV is mainly caused by the problem of private information. As we know, price information in the network of fish markets is primarily distributed by moving middlemen, who travel between the city and local towns as well as fishing grounds. When there are few middlemen competing with each other, the middleman is more likely to over-report the cost of transportation and under-report the fish price in the city; in addition he will probably over-report costs and under-report the price even more than usual, all to seek a larger share of gains from transactions with fishermen. With few middlemen to choose from, the fishermen become highly skeptical of the price information communicated by the middlemen, and they tend to bargain cautiously to avoid being cheated. These two conflicting tendencies are inclined to tear bargaining apart.

A numeric example may help to illustrate the point. Suppose the fisherman values his fish at a price of $4. The middleman reports that the price at Jingzhou is $10. Then, there is a wide range of prices ($4, $10) for them to bargain. For simplicity, suppose both parties expect to equally divide the surplus of trade (which is $10 - $4 = $6), then the outcome is a price of $7 ($4 + 1/2 * $6 = $7). Now, let us take into consideration the problem of information. When there are only a couple of middlemen traveling to WV, the price information revealed by them becomes suspicious to fishermen. Suppose the price in Jingzhou has two states of equal possibility, either a high price of $12 or a low one of $8. This is common knowledge for both the fisherman and the middleman. But the fisherman does not know what the price is on any particular day. If the price information could be credibly communicated between the middleman and fisherman, they would reach a deal at $8 or $6. However, when the number of middlemen visiting WV is quite limited, so the middleman tends to under-report the price, and keep a larger slice of the trade surplus. This strong incentive on the part of middleman to under-report the price makes the fisherman suspicious about the price information communicated by the middleman. Even when the middleman truthfully reveals the price information, the fisherman does not take it at face value simply because he has no way to check whether the middleman is honest. As a result, even when the middleman is honest, it does not increase his chance of making a deal with the fisherman. As for the middleman, he does not have any means to

persuade to the fisherman that he is honest. For example, if the middleman honestly reports that the price is $8, the fisherman is prone to think that the he is under-reporting the price. While the middleman expects a price of $6 as the deal, the fisherman anticipates a price of $8, a price that the middleman can not possibly offer. As a result, bargaining between the middleman and fisher man tends to fail. Overall, bargaining between the middleman and fisherman tends to break up about half of the time when we assume the price has two states of equal likelihood. Since fish prices vary continuously in the real world, bargaining becomes extremely difficult when the fisherman is skeptical of the middleman's price information.

In addition to the problem of bargaining, high transportation cost of carrying fish across the lake from WV to Port A makes WV less attractive to middlemen. Indeed, the middleman takes bigger risk when he buys fish from fishermen at WV: it takes longer to transport fish from WV to Jingzhou and thus creates uncertainty because fish are more likely to die in long distance transportation and the fish price fluctuates. Hence, the middleman must take the factor of risk premium into consideration, in addition to the increased transportation cost, when bargaining with fishermen at WV. Therefore the middleman bargains tough with fishermen at WV to push the price down. When the likelihood of stalemate or bargaining failure increases, the risk premium increases accordingly, and bargaining becomes more aggressive and hence more likely to fail.

To put it in a nutshell, the high cost of transactions between middlemen and fishermen at WV reduces the number of middlemen visiting WV, and increases the risk involved in buying fish there. These two mechanisms reinforce each other to make bargaining more difficult between fishermen and middlemen. On the one hand, as fewer middlemen go to WV, price information is less likely to be credibly transmitted to fishermen; as information asymmetry worsens between fishermen and the middleman, bargaining between them is more likely to fail. At the same time, the high probability of bargaining failure increases the risk premium that the middleman takes into account during price bargaining with the fishermen. This makes middlemen more aggressive in bargaining, and once again the bargaining is more likely to break down. In turn, the bleak chance of successful bargaining discourages middleman from traveling to WV. With fewer and fewer middlemen going to WV, the bargaining becomes even more difficult and more likely to f all into an impasse. This self-perpetuating cycle blocks middlemen from visiting WV. Hence a price mechanism cannot be established and the market fails to emerge. As a result, fishermen at WV transport their fish across Long Lake to the market directly.

From the perspective of the fisherman at WV, the vertical integration of transportation into the value chain is apparently an expansion of the scope of the "firm." At WV, the division of labor between' fishing and fish transportation does not exist. Both activities are performed within one firm. This is sharply different from the case at DV, where fishing and fish transportation are undertaken by separate specialized firms, with the market coordinating transactions between the fishing firm (i.e., fishermen) and the transportation firm (that is, middlemen). From the standpoint of the fisherman at DV, fish transportation is undertaken through the market, not within the "firm."

The market and the firm are two alternatives for conducting fish transportation. At WV, the firm (that is, the fisherman) expands and integrates fish transportation into its scope of activities or value chain. At DV, there is a division of labor between fishing and fish transportation and the fisherman turns to the market for fish transportation. This contrast is consistent with the basic logic embodied in Coase's (1937) classic article, which says that the firm expands and supersedes the market when the cost of using the price mechanism is high. However the situation differs from Coase's theory, which starts with a world of pure market transactions, in that the market does not even exist at WV simply because it is too costly to be set up.

At DV, low transaction costs incurred to middlemen make their entrepreneurial efforts rewarding. When they stay in business, the market is thus at work. The market itself consists of a chain of middlemen, each of whom qualifies as a firm, buying and selling, calculating profit and the cost of making profit. The existence of the market encourages the division of labor between fishing and fish transportation, and hence increases specialization. Viewed this way, the market and the firm are not substitutes for each other; instead, they are complements.

Though each middleman is viewed as a firm, it is a special type of firm whose function is to create a market between producers (e.g., fishermen) and consumers (e.g., city residents). The market is thus dependent on the entrepreneurial efforts of middlemen. Apparently then, its existence can not be taken for granted. Indeed, there were simply no markets in the beginning when commercial activities were criminalized by the law under the doctrine of socialist planned economy. The market comes into place when the middleman is able to profit from buying and selling. Comparing the different marketing channels at DV and WV nicely illustrates the point that the provision of the market is costly and requires entrepreneurial efforts. When the cost of providing the market at WV is too high, the market fails to emerge. The difference between the institutional structure of exchange in the fishery industry at DV and WV brings us back to the central point in new institutional economics, that transaction cost is a primary determinant of the institutional structure of the economy (Coase 1992).

Concluding Remarks

Ever since Adam Smith, the development of the division of labor and the market has been viewed as a natural consequence of human propensity, "the propensity to truck, barter, and exchange." Neither the actual operation of the market system nor the evolution of the institutional structure of the market has seriously interested economists who claim to study the working of the market. Staying aloof from the real world market not only makes economists blind to the cost of using the market system, but also makes them inclined to take the market as given and assume that "in the beginning there were markets" (Fourie 1993). This study shows however that the market did not exist in the beginning. Instead, the provision of the market involves entrepreneurial endeavors and is subject to cost-benefit analysis. When the cost of providing the market exceeds the potential gains, the market will not be provided at all.

Focusing on the operation of fish markets in a Chinese fishery community, this case study brings some realism into economic analysis. In the process of examining how this fish market works, this case study empirically demonstrates that the institutional structure of exchange responds to the cost of carrying out exchange transactions. The differing cost of organizing fish transportation at DV and WV gives rise to the stark difference in the institutional structure of fish transportation at these two villages. From the perspective of fishermen, fish transportation at DV is left to the market; at WV it is performed within the "firm."

In this study, the market consists of a continuum of middlemen, whose entrepreneurial efforts of buying and selling bring the market to work. The raison d'etre for middlemen is to reduce transaction cost for both fishermen and consumers. This is consistent with the Coasean view of the market: "Markets are institutions that exist to facilitate exchange, that is, they exist in order to reduce the cost of carrying out exchange transactions." (Coase 1988a, 7). Unlike most studies which assume "in the beginning there were markets," this study takes one step further to demonstrate that the provision of the market depends on whether the middleman can make a profit and stay in business. By doing so, this paper sheds new light on the relationship between the firm and the market vis-a-vis transaction costs.

Source: The American Journal of Economics and Sociology, Oct, 1999, by Ning Wang

Questions:

1. Do you really think that the institutional structure of exchange responds to the cost of carrying out exchange transactions? Explain.

2. Which factors bring market to work?3. “The provision of the market depends on whether the middleman can make a profit and stay in business”,

do you agree with this statement?

Unit 2

EFFICIENT CAPITAL MARKETS: THEORY

The most important role of the capital market is allocation of possession of the economy’s capital stock. In general terms, the ideal is a market in which prices provide correct signals for resource allocation: that is, a market in which firms can make production-investment decisions, and investors can choose among the securities that symbolize possession of firm’s activities under the assumption that security prices at any time “fully reflect” all available information. A market in which prices always “fully reflect” accessible information is known as “efficient”.The purpose of capital markets is to transfer funds between lenders (savers) and borrowers (producers) efficiently. Individuals or firms may have access the market-determined borrowing rate but not enough funds to take advantage of all these opportunities. Nevertheless, if capital markets exist, they can borrow the needed funds. Lenders, who have excess funds after exhausting all their productive opportunities with expected returns greater than the borrowing rate, will be willing to lend their excess funds for the reason that the borrowing/lending rate is higher than what they might otherwise earn. Hence both borrowers and lenders are better off if efficient capital markets are used to facilitate fund transfers. The borrowing/lending rate is used as a significant piece of information by each producer, who will accept projects until the rate of return on the least profitable project just equals the opportunity cost of external funds (the borrowing/lending rate). Expected Return or “Fair Game” ModelsThe definitional statement that in an efficient market prices “fully reflect” accessible information is so wide-ranging that it has no pragmatically testable entailments. To make the model testable, the process of price formation must be specified in more detail. In essence we must define somewhat more exactly what is intended by the term “fully reflect.”One option would be to hypothesize that equilibrium prices (or expected returns) on securities are generated as in the “two parameter” Sharpe and Linter world. In general, nevertheless, the theoretical models and especially the empirical tests of capital market efficiency have not been this specific. Most of the available work is based only on the assumption that the conditions of market equilibrium can (somehow) be stated in terms of expected returns. In common terms, like the two parameter model such theories would hypothesize that conditional on some relevant information set, the equilibrium expected return on a security is a function of its “risk.” And different theories would differ primarily in how “risk” is defined. All members of the class of such “expected return theories” can, nevertheless, be described notationally as follows:

(1)where E is the expected value operator; p jt is the price of security j at time t; p j,t+1 is its price at t+1 (with reinvestment of any intermediate cash income from the security); r j,t+1 is the one-period percentage return (pj,t+1 – pjt)/pjt; φt is a general symbol for whatever set of information is assumed to be “fully reflected” in the price at t; and the tildes indicate that pj,t+1 and rj,t+1 are random variables at t.

The value of the equilibrium expected return projected on the basis of the information φt

would be determined from the particular expected return theory at hand. The conditional expectation

notation of (1) is meant to imply, nevertheless, that whatever expected return model is assumed to apply, the information in φt is fully utilized in determining equilibrium expected returns. And this is the sense in which φt is “fully reflected” in the formation of the price pjt.But we should note right off that, simple as it is, the assumption that the conditions of market equilibrium can be stated in terms of expected returns elevates the purely mathematical concept of expected value to a status not necessarily implied by the general notion of market efficiency. The expected value is just one of many possible summary measures of a distribution of returns, and market efficiency per se (i.e., the general notion that prices “fully reflect” available information) does not imbue it with any special significance. Thus, the results of tests based on this assumption depend to some extent on its validity as well as on the efficiency of the market. But some such assumption is the unavoidable price must pay to give the theory of efficient markets empirical content.The assumptions that the conditions of market equilibrium can be stated in terms of expected returns and that equilibrium expected return are formed on the basis of (and thus “fully reflect”) the information set φt have a major empirical implication—they rule out the possibility of trading systems based only on information in φt that have expected profits or returns in excess of equilibrium expected profits or returns. Thus let

(2)Then,

(3)which, by definition, says that the sequence {Xjt} us a “fair game” with respect to the information sequence {φt}. Or, equivalently, let

(4)then

()so that the sequence {zjt} is also a “fair game” with respect to the information sequence {φ}.In economic terms, xj,t+1 is the excess market value of security j at time t+1: it is the difference between the observed price and the expected value of the price that was projected at t on the basis of the information φt. And similarly, zj,t+1 is the return at t+1 in excess of the equilibrium expected return projected at t. Let

be any trading system based on φt which tells the investor the amounts αj(φt) of funds available at t that are to be invested in each of the n available securities. The total excess market value at t+1 that will be generated by such a system is

which, from the “fair game” property of (5) has expectation,

The expected return or “fair game” efficient markets model has other significant testable implication, but these are better saved for the later discussion of the empirical work. Now we turn to two special cases of the model, the submartingale and the random walk, that (as we shall see later) play an significant role in the empirical literature.The Submartingale ModelSuppose we assume in (1) that for all t and φt.

, or equivalently, . (6)This is a statement that the price sequence {p jt} for security j follows a submartingale with respect to the information sequence {φt}, which is to say nothing more than that the expected value of next period’s price, as projected on the basis of the information φt, is equal to or greater than the current price. If (6) holds and equality (so that expected returns and price changes are zero), then the price sequence follows a martingale.A submartingale in prices has one significant empirical implication. Consider the set of “one security and cash” mechanical trading rules by which we mean systems that concentrate on individual securities and that define the conditions under which the investor would hold a given security, sell it short, or simply hold cash at any time t. Then the assumption of (6) that expected returns conditional on φ t are non-negative directly implies that such trading rules based only on the information in φ t cannot have greater expected profits than a policy of always buying-and-holding the security during the future period in question. Tests of such rules will be an significant part of the empirical evidence on the efficient markets model.The Random Walk ModelIn the early treatments of the efficient markets model, the statement that the current price of a security “fully reflects” available information was assumed to imply that successive price changes (or more generally, successive one-period returns) are independent. In addition, it was generally assumed that successive changes (or returns) are identically distributed. Together the two hypotheses constitute the random walk model. Formally, the model says

(7)which is the usual statement that the conditional and marginal probability distributions of an independent random variable are identical. In addition, the density function f must be the same for all t.Expression (7) of course says much more than the general expected return model summarized by (1). For example, if we restrict (1) by assuming that the expected return on security j is constant over time, then we have

(8)This says that the mean of the distribution of rj,t+1 is independent of the information available at t, φt, whereas the random walk model of (7) in addition says that the entire distribution is independent of φt.We argue later that it is best to regard the random walk model as an extension of the general expected return or “fair game” efficient markets model in the sense of making a more detailed statement about the economic environment. The “fair game” model just says that the conditions of market equilibrium can be stated in terms of expected returns, and thus it says little about the details of the stochastic process generating returns. A random walk arises within the context of such a model when the

environment is (fortuitously) such that the evolution of investor tastes and the process generating new information combine to produce equilibria in which return distributions repeat themselves through time.Thus it is not surprising that empirical tests of the “random walk” model that are in fact tests of “fair game” properties are more strongly in support of the model than tests of the additional (and, from the viewpoint of expected return market efficiency, superfluous) pure independence assumption. (But it is perhaps equally surprising that, as we shall soon see, the evidence against the independence of returns over time is as weak as it is).Market Conditions Consistent with EfficiencyBefore turning to the empirical work, nevertheless, a few words about the market conditions that might help or hinder efficient adjustment of prices to informations are in order. First, it is easy to determine sufficient conditions for capital market efficiency. For example, consider a market in which (i) there are no transactions costs in trading securities, (ii) all available information is costlessly available to all market participants, and (iii) all agree on the implications of current information for the current price and distributions of future prices of each security. In such a market, the current price of a security obviously “fully reflects” all available information.But a frictionless market in which all information is freely available and investors agree on its implications is, of course, not descriptive of markets met in practice. Fortunately, these conditions are sufficient for market efficiency, but not necessary. For example, as long as transactors take account of all available information, even large transactions costs that inhibit the flow of transactions do not in themselves imply that when transactions do take place, prices will not “fully reflect” available information. Similarly (and speaking, as above, somewhat loosely), the market may be efficient if “sufficient numbers” of investors have ready access to available information. And disagreement among investors about the implications of given information does not in itself imply market inefficiency unless there are investors who can consistently make better evaluations of available information than are implicit in market prices.But though transactions costs, information that is not freely available to all investors, and disagreement among investors about the implications of given information are not necessarily sources of market inefficiency, they are potential sources. And all three exist to some extent in real world markets. Measuring their effects on the process of price formation is, of course, the major goal of empirical work in this area.All the empirical research on the theory of efficient markets has been concerned with whether prices “fully reflect” particular subsets of available of available information. Historically, the empirical work evolved more or less as follows. The initial studies were concerned with what we call weak form tests in which the information subset of interest is just past price (or return) histories. Most of the results here come from the random walk literature. When extensive tests seemed to support the efficiency hypothesis at this level, attention was turned to semi-strong form tests in which the concern is the speed of price adjustment to other obviously publicly available information (e.g., announcements of stock splits, annual reports, new security issues, etc.). Finally, strong form tests in which the concern is whether any investor or groups (e.g., managements of mutual funds) have monopolistic access to any information relevant for the formation of prices have recently appeared. We review the empirical research in more or less this historical sequence.

First, nevertheless, we should note that what we have called the efficient markets model in the discussions of earlier sections is the hypothesis that security prices at any point in time “fully reflect” all available information. Though we shall argue that the model stands up rather well to the data, it is obviously an extreme null hypothesis. And, like any other extreme null hypothesis, we do not expect it to be literally true. The categorization of the tests into weak, semi-strong, and strong form will serve the useful purpose of allowing us to pinpoint the level of information at which the hypothesis breaks down. An we shall contend that there is no significant evidence against the hypothesis in the weak and semi-strong form tests (i.e., prices seem to efficiently adjust to obviously publicly available information), and only limited evidence against the hypothesis in the strong form tests (i.e., monopolistic access to information about prices does not seem to be a prevalent phenomenon in the investment community).Weak Form Tests of the Efficient Markets ModelAs noted earlier, all of the empirical work on efficient markets can be considered within the context of the general expected return or “fair game” model, and much of the evidence bears directly on the special submartingale expected return model of (6). Indeed, in the early literature, discussions of the efficient markets model were phrased in terms of the even more special random walk model, though we shall argue that most of the early authors were in fact concerned with more general versions of the “fair game” model.Some of the confusion in the early random walk writings is understandable. Research on security prices did not begin with the development of a theory of price formation which was then subjected to empirical tests. Rather, the impetus for the development of a theory came from the accumulation of evidence in the middle 1950’s and early 1960’s that the behavior of common stock and other speculative prices could be well approximated by a random walk. Faced with the evidence, economists felt compelled to offer some rationalization. What resulted was a theory of efficient markets stated in terms of random walks, but generally implying some more general “fair game” model.It was not until the work of Samuelson and Mandelbrot in 1965 and 1966 that the role of “fair game” expected return models in the theory of efficient markets and the relationships between these models and the theory of random walks were rigorously studied. As these papers came somewhat after the major empirical work on random walks. In the earlier work, “theoretical” discussions, though generally intuitively appealing, were always lacking in rigor and often either vague or ad hoc. In short, until the Mandelbrot-Samuelson models appeared, there existed a large body of empirical results in search of a rigorous theory.Thus, though his contributions were ignored for sixty years, the first statement and test of the random walk model was that of Bachelier in 1900. But his “fundamental principle” for the behavior of prices was that speculation should be a “fair game”; in particular, the expected profits to the speculator should be zero. With the benefit of the modern theory of stochastic processes, we know now that the process implied by this fundamental principle is a martingale.After Bachelier, research on the behaviour of security prices lagged until the coming of the computer. In 1953 Kendall examined the behavior of weekly changes in nineteen indices of British industrial share prices and in spot prices of cotton (New York) and wheat (Chicago). After extensive analysis of serial correlations, he suggests, in quite graphic terms:

The series looks like a wandering one, almost as if once a week the Demon of Chance drew a random number from a symmetrical population of fixed dispersion and added it to the current price to determine the next week’s price.

Kendall’s conclusion had in fact been suggested earlier by Working though his suggestion lacked the force provided by Kendall’s empirical results. And the implications of the conclusion for stock market research and financial analysis were later underlined by Roberts.But the suggestion by Kendall, Working, and Roberts that series of speculative prices may be well described by random walks was based on observation. None of these authors attempted to provide much economic rationale for the hypothesis, and, indeed, Kendall felt that economists would usually reject it. Osborne suggested market conditions, similar to those assumed by Bachilier, that would lead to a random walk. But in his model, independence of successive price changes derives from the assumption that the decisions of investors in an individual security are independent from transaction to transaction–which is little in the way of an economic model.Whenever economists (prior to Mandelbrot and Samuelson) tried to provide economic justification for the random walk, their arguments generally implied a “fair game.” For example, Alexander states:

If any substantial group of buyers thought prices were too low, their buying would force up the prices. The reserve would be true for sellers. Except for appreciation due to earnings retention, the conditional expectation of tomorrow’s price, given today’s price, is today’s price.In such a world, the only price changes that would occur are those that result from new information. Since there is no reason to expect that information to non-random in appearance, the period-to-period price changes of stock should be random movements, statistically independent of one another.

Although to some extent vague, the last sentence of the first paragraph seems to point to a “fair game” model rather than a random walk. In this light, the second paragraph can be viewed as an attempt to explain environmental conditions that would reduce a “fair game” to a random walk. But the specification imposed on the information generating process is inadequate for this purpose; one would, for instance, also have to say something about investor tastes. By distinction, the stock market trader has a much more practical criterion for judging what constitutes significant dependence in successive price changes. For his purpose the random walk model is valid as long as knowledge of the past behavior of the series of price changes cannot be used to increase expected gains. More particularly, the independence assumption is an sufficient description of reality as long as the actual degree of dependence in the series of price changes is not enough to allow the past history of the series to be used to predict the future in a way which makes expected profits greater than they would be under a naïve buy and hold model.We know now, of course, that this last condition hardly needs a random walk. It will in fact be met by the submartingale model.But one should not be too hard on the hypothetical efforts of the early empirical random walk literature. The arguments were generally appealing; where they fell short was in consciousness of developments in the theory of stochastic processes. Furthermore, we shall now see that most

of the experiential evidence in the random walk literature can easily be interpreted as tests of more common expected return or “fair game” models.Tests of Market Efficiency in the Random Walk LiteratureAs discussed earlier, “fair game” models entail the “impossibility” of different sorts of trading systems. Some of the random walk literature has been concerned with testing the profitability of such systems. More of the literature has, nevertheless, been concerned with tests of serial covariances of a “fair game” are zero, so that these tests are also relevant for the expected return models.If xt is a “fair game,” its unconditional expectation is zero and its serial covariance can be written in general form as:

where f indicates a density function. But is xt is a “fair game,”

From this it follows that for all lags, the serial covariances between lagged values of a “fair game” variable are zero. Thus, observations of a “fair game” variable are linearly independent.But the “fair game” model does not necessarily imply that the serial covariances of one-period returns are zero. In the weak form tests of this model the “fair game” variable is

(9)But the covariance between, for example, rjt and rj,t+1 is

=

and (9) does not imply that = : In the “fair game” efficient markets model, the deviation of the return for t+1 from its conditional expectation is a “fair game” variable, but conditional expectation itself can depend on the return observed for t.In the random walk literature, this problem is not documented, from the time when it is assumed that the predictable return (and indeed the entire distribution of returns) is stationary through time. In practice, this implies estimating serial covariances by taking cross products of deviations of experiential returns from the overall sample mean return. It is rather accidental, then, that is procedure, which represents a rather gross estimate from the point of view of the wide-ranging expected return efficient markets model, does not seem to very much influence the results of the covariance tests, at least for common stocks.FORMAL DEFINITION OF THE VALUE OF INFORMATIONThe notion of efficient capital markets depends on the precise definition of information and the value of information obtained from it. An information structure may be defined as a message about various events that may happen. For instance, the message “there are no clouds in the sky” provides a probability distribution for the likelihood of rain within the next 24 hours. This message may have various values to different people depending on (1) whether or not they can

take any actions based on the message and (2) what net benefits will result from their actions. For example, a message related to rainfall can be of value to farmers, who can act on the information to increase their agricultural products. If there is no rain, the farmers might decide that it would be a good time to harvest hay. On the other hand, messages about rainfall have no value to deep-pit coal miners for the reason that such information probably will not alter the miners’ actions at all.A formal expression of the above concept defines the value of an information structure as

(10)where,q(m) = the prior probability of receiving a message m;p(e|m) = the conditional probability of an event e, given a message m;U(a, e) = the utility resulting from an action a if an event e occurs (benefit function)ν(η0) = the expected utility of the decision maker without the information.According to Eq. (10), a decision maker will evaluate an information structure (which, for the sake of generality, is defined as a set of messages) by choosing an action that will maximize his or her expected utility, given the arrival of the message. For example, if we receive a message (one of many that we could have received) that there is a 20% chance of rain, we may carry an umbrella for the reason that of the high “disutility” of getting drenched and the low cost of carrying it. For each possible message we can determine our optimal action. Mathematically, this is the solution to the problem

Finally, by weighting the expected utility of each optimal action (in response to all possible messages) by the probability, q(m), of receiving the message that gives rise to the action, the decision maker knows the expected utility of the entire set of messages, which we call the expected utility (or utility) or utility value of an information set, v(η).Note that the value of information as defined above in Eq. (10) in a one period setting is very close to the value of a real option in a multi-period setting. The basic idea is the same. Given the arrival of information a decision maker is assumed to have the right, but not the obligation, to take and action. For this reason, the value of information is optimized along three variables, namely, the payoff to the decision maker, the information about the event given the message, and the action assumed to be taken by the decision maker.To illustrate the value of information, consider the following example. The analysis division of a large retail brokerage firm hires and trains analysts to cover events in various regions around the world. The CIO (Chief Information Officer) has to manage the allocation of analysts and deal with the fact that the demand for analyst coverage is uncertain. To keep it simple, suppose we need 300 analysts or 0 analysts with 50-50 probability for coverage in each of three countries. The CIO is considering three possibilities.Policy 1: Hire analysts and keep them on their initial allocations.Policy 2: Switch analysts at a cost of 20 thousand dollars when it makes sense.

Policy 3: Create a catalogue (CAT) that provides better information (e.g., their language capabilities, their technical skills, etc.) for matching analysts to countries when they are switched. The CIO has assigned project weights to various allocation results:Outcome Weight (utility = U(a, e))Analyst skilled in a job 100 unitsAnalysts unsatisfactory in job 30 unitsShortage – no one in job 0 unitsExcess – no demand for job – 30 units

Table 2.1: CAT Provides Better InformationWithout CAT With CAT

Probability of being skilled on the job .7 1.0Probability of being mismatched .3 0.0

We would like to know which is the best policy, what it implies about the number of analysts to hire initially, and what value should be given to the CAT information. The intuition is reasonably clear. If, for example, the CAT information does not help much to make better job allocations, if the cost of mismatch is relatively low, or if the cost of switching is close to zero, then the value differential between policies 1 and 2 will be low. Table 2.1 shows the probability of getting the right analyst into the right job without having the catalogue information (right-hand side).Assuming that analysts are allocated without the benefit of CAT, there is 70% chance of allocating the analyst with the right skills to the right job. Nevertheless, if the CAT information is available, the probability of a correct allocation goes up to 100 percent.If the brokerage company adopts a policy of not retraining and reallocating analysts, it should hire 900 analysts for a value of 20,050. This result follows from the fact that there are four states of nature, with demand of 0, 300, 600, and 900 analysts, respectively.To determine the expected payout from a policy of hiring 600 analysts, but not reallocating them, we can refer to table 2.2. With 600 analysts we would allocate 200 to each of the countries (A, B, and C). If demand turns out to be 300 in total, the country that has that demand will be 100 analysts short and there will be no reallocation. The payout for 140 of the analysts who turn out to be skilled on the job is 100, and the payout for 60 analysts who are unsatisfactory is 30 each. The shortage of 100 has no cost or benefit and there is no excess. Hence, the probability-weighted payment is

.375{[140(100) + 60(30)] + 400(–30)} = 1,425.Table 2.2: Expected Payout from Hiring 600 Analysts and Not Reallocating Them

Demand/SupplyDemand Probability Country

ACountry

BCountry

C0 .125 0/200 0/200 0/200

300 .375 300/200 0/200 0/200600 .375 300/200 300/200 0/200900 .125 300/200 300/200 300/200

PayoutsDemand Skilled Mismatch Shortage Excess Contribution

0 0 0 0(0) 600(–30) –2,250300 (.7)(200)(100) (.3)(200)(30) 100(0) 400(–30) 1,425600 (.7)(400)(100) 2(.3)(400)(30) 200(0) 200(–30) 9,600900 (.7)(600)(100) (.3)(600)(30) 100(0) 0 5,925

14,700Next, if demand turns out to be 600, it would reach that level for the reason that two of the three countries have demand of 300. Each would have been allocated 200 and would have a 100-person shortage. The third country would have an excess of 200. The probability-weighted outcome is

.375{[2[140(100) + 60(30)] + 200(–30) + 0}} = 9,600.To complete the analysis of a strategy that consists of hiring 600 people and not reallocating them, we go the last column in table 2.2, where the probability-weighted contributions add to 14,700.The results for all strategies are given in Table 2.3. The value maximizing strategy without reallocation is to hire 900 people, enough to supply maximum demand for analysts in all three countries. But with the human resources catalogue (CAT) that makes the process of reallocation more accurate, the value-maximizing policy is to hire 600 people with the expectation that many would be reallocated. The difference between the value of reallocation without CAT and reallocation with it is the value of having the CAT (i.e., 28,825 – 22,462).

Table 2.3: Results for All StrategiesAllocation Policy

Number Hired

No Reallocation

Reallocation without CAT

Reallocation with CAT

0 0 0 0300 7,350 16,328 20,625600 14,700 22,462 28,875900 22,050 22,050 28,125

1,200 13,050 13,050 19,125RELATIONSHIP BETWEEN THE VALUE OF INFORMATION AND EFFICIENT CAPITAL MARKETSThe equation 10 above can be used to evaluate any information structure. It also points out some ideas that are implicit in the definition of efficient markers. Fama (1976, 1991) defines efficient capital markets as those where the joint distribution of security prices at a period, given the set of information that the market uses to determine security prices, is identical to the joint distribution of prices that would exist if all relevant information available at that period were used. This implies that there must be no distinction between the information the market uses and the set of all relevant information. Applying information theory, this also implies that net of costs, the utility value of the gain from information to an individual in nil.For instance, the Nigerian Stock Exchange (NSE) has been described as being efficient in the weak form. The relevant information structure, ni, is defined to be the set of historical prices on all assets. Hence, the distribution of security prices today has already incorporated past price histories. For this reason, it is not possible to develop trading rules (courses of action) based on past prices that will allow anyone to ‘outperform’ the market. In other words, the value of the gain from information to an ith individual, must be zero.

ν(ηi) – ν(η0) = 0 (11)Equation (11) says that no one would pay anything for the information set of historical prices. It is pertinent to note that the capital market is efficient relative to a given information set only after consideration of the costs of acquiring messages and taking actions pursuant to a particular information structure.MARKET EFFICIENCY AND RATIONAL EXPECTATIONSOur aim is to have an insight into how the marginal investor’s decision-making process, given the receipt of information, is reflected in the market prices of assets. Nevertheless, it is difficult to observe the quantity and quality of information or the timing of its receipt in the real world. Even the issue of what information is relevant to investors has been an unresolved matter amongst theorists. Forsythe, Palfrey, and Plott (1982) identify four different hypotheses. Each hypothesis assumes that investors know with certainty what their own payoffs will be across time, but they also know that different individuals may pay different prices for the reason that of differing preferences.The first hypothesis, known as naïve hypothesis, asserts that asset prices are completely arbitrary and unrelated either to how much they pay out in the future or to the probabilities of various payouts. The second hypothesis, known as the speculative equilibrium hypothesis, implies that all investors base their investment decisions entirely on their anticipation of other individual’s behaviour without any necessary relationship to the actual payoffs that the assets are expected to provide. The third hypothesis is that asset prices are systematically related to their future payouts. Called the intrinsic value hypothesis, it says that prices will be determined by each individual’s estimate of the payoffs of an asset without consideration of its resale value to other individuals. The fourth hypothesis may be called the rational expectations hypothesis. It predicts that prices are formed on the basis of the expected future payouts of the assets, including their resale value to third parties. Thus, a rational expectations market is an efficient market for the reason that prices will reflect all information. In the rational expectations model, differential payoffs indicate heterogeneous expectations. Heterogeneous expectations could result from information asymmetry amongst individuals. An unresolved issue about market efficiency was whether there is full aggregation or averaging of information in pricing. A fully aggregating market is said to be consistent with the Fama’s (1970) definition of strong form efficiency. In a fully aggregating market, even insiders who possess private information would not be above to profit by it. Nevertheless, empirical evidence on insider dealing suggests that insider can and do make abnormal returns. On this basis, it can be said that capital markets do no instantaneously and fully aggregate information.Market Efficiency with Costly InformationCapital market efficiency relies on the ability of arbitrageurs to recognize that prices are out of line and to make a profit by driving them back to an equilibrium value consistent with available information. Given this type of behavioural paradigm, one often hears the following questions: If capital market efficiency implies that no one can beat the market, then how can analysts be expected to exist since they too, cannot beat the market? If capital markets are efficient, how can we explain the existence of a multibillion naira security analysis industry? The answer, of course, is that neither of these questions is inconsistent with efficient capital markets. First, analysts can and do make profit in a competitive manner amongst themselves. If the profit to analysis becomes abnormally large, then individuals will enter the analysis business until the abnormal profit becomes ‘wiped out’. Cornell and Roll (1981) and Elton, et al (1993) have shown that a sensible asset market equilibrium must leave some room for analysis. Their articles make the more reasonable assumption that information acquisition is costly activity. Cornell and Roll showed that it is reasonable to have efficient markets where people earn different gross rates of return for the reason that they pay differing costs for information.

Nevertheless, net of costs their abnormal rates of return will be equal (to zero0. Elton, et.al (1993) show that gains by professional fund managers appear to no more than cover the expenses of managing the portfolios.Applications of the EMHUsually, a wide range of applications of capital market efficiency includes issues such as accounting information, block trades, new issues of securities, stock splits, and portfolio performance measurement.Accounting InformationBusiness news takes various forms – the announcement of a change in top management, the awarding of a large contract to a competitor, analysts’ reports, or a forecast of earnings by a firm’s management. The most extensively used sources of information, nevertheless, are the firm’s published financial reports. Published statements play a significant role in the dissemination of corporate information. Specified the plethora of complex accounting procedures and principles, we would not expect all market participants to be able to differentiate false or misleading information is based on various assumptions and principles which authorize the use of alternative actions in various situations. Examples comprise the timing of revenue and expense recognition (accrual concept), accounting for lease obligations and mergers, inventory accounting technique, and son on (see Dyckman, et al, 1975:2-4).In accounting terms the market value of assets is the present value of their cash flows discounted at the appropriate risk-adjusted rate. Nonetheless, corporations report accounting earnings, not cash flow, and regularly the two are not related. Empirical evidence shows that if investors really value cash flow and not earnings per share (EPS), we would expect to see stock prices rise when firms.Further, on accounting treatment of mergers and acquisitions, two possibilities exist: pooling or purchase. In a pooling arrangement, the income statements and balance sheets of the merging firms are simply added together. On the contrary, when one company acquires another, the assets of the acquired company are added to the acquiring company’s balance sheet along with an item called goodwill. In an empirical study, hong, Kaplan and Madelker (1978) tested the effect of pooling and purchase techniques on stock prices of acquiring firms. Using monthly data between 1954 and 1964, they compared a sample of 122 firms that used pooling and 37 that used purchase. The acquired firm had to be at least 3% of the net asset value of the acquiring firm. Ball and Brown (1968) used a procedure to assess the speed of adjustment of security prices to earnings announcement in the Wall Street Journal. Ball and Brown reasoned that the market participants would have formed opinions reflected in their forecasts of what the earnings numbers should be, and, collectively, these forecasts would be reflected in a market forecast of the stock’s price. Ball and Brown used monthly data for a sample of 261 firms between 1946 and 1965 to evaluate the usefulness of information in annual financial reports (AFRs). First, they separated the sample into companies that had earnings that were either higher (‘good news’) or lower (‘bad news’) than market’s forecast. The market’s forecast was based on a naïve time series model. Finally, estimated earnings changes were compared with actual earnings changes. If the actual change was greater than estimated, the company was put into a portfolio where returns were expected to be positive, and vice versa. In both cases (good or bad news), the cumulative excess return displayed no significant behaviour in the months following the earnings announcement. The study, thus, indicates that most of the information contained in the AFRs is anticipated by the market before the AFRs are released. Brown and Kennelly (1972) found similar results concerning market reaction to quarterly earnings reports. The Ball and Brown study raised the question of whether or not AFRs contain any new information. Griffin (1977), Foster (1977), Brown (1978), Watts (1978), Aharony and Swary (1980), and Joy, Litzenberger and McEnally (1977) have focused on quarterly earnings reports where information revealed to the market is (perhaps) more timely than AFRs.

Trimonthly earnings reports are sometimes followed by announcements of dividend changes which also have an effect on the stock price. To examine this problem, Aharony and Swary (1980) observe all dividend and earnings announcements within the same period that are at least 11 trading days apart. They conclude that both quarterly earnings announcements and dividend change announcements have statistically significant effects on the stock price. But more significant, they find no evidence of market inefficiency when the two types of announcement effects are separated. Two studies, one by Pettit (1972) and one by Watts (1973), measured the market’s reaction to dividend announcements. Although, the authors arrived at different conclusions concerning the significance of dividend changes to market participants, the results of both studies are consistent with the behaviour implied by the EMH: there was no evidence that a firm’s dividend announcement affected the firm’s security price in the periods following the announcement. Handjinicolaou and Kalay (1984) and Woolridge (1983) have argued that one cannot infer that dividend increases convey positive information about the firm by examing share prices alone, since unexpected dividend increases could cause wealth transfers from bondholders to shareholders by reducing the asset base of the firm. Hence, the observed increase in share price is consistent with both wealth redistribution and positive information. To distinguish between the relative significance of these two effects, Handjinicolaou and Kalay and Woolridge analyze the changes in bond prices around dividend announcements since the two hypotheses (information and wealth transfer) have different predictions for bond price behaviour. In particular, the wealth transfer hypothesis predicts a negative bond price reaction while the information content hypothesis predicts a positive reaction. The findings of these studies give strong support for the hypothesis that informational effects dominate wealth redistribution effects wherever there are unexpected dividend increases.Bhana (1991) analyzed the share market response to considerable changes in dividend policies by Johannesburg Stock Exchange (JSE) listed stocks during the period 1970-1988. The results give strong support for the information content of dividends hypothesis. The empirical evidence suggests that large dividend changes on the JSE convey important information to investors over and above that contained in the earnings announcements. Further, in a subsequent research, Bhana (2002) observes a sample of 100 companies announcing special dividends over the period “1975 – 1994. Daily data on share prices were obtained from the database of the JSE and “McGregor’s Online Information Services. “The study indicates stock price reaction to ‘labelled’ (in the sense that it is only temporal or seasonal and is not expected to be frequent) increases in dividends. It shows that share price reactions are negatively related to dividend declaration frequency. This outcome suggests that market participants look forward to the announcements of special dividends by companies that have numerous declarations of such dividends. This, in turn, shows that request declaration of special dividends convey less information than do rare declarations.Block TradesScholes (1972) and Kraus and Stoll (1972) furnished the first empirical substantiation about the price effects of block trading. Scholes used daily returns data to analyse 345 secondary distributions between July 1961 and December 1965. Secondary distributions, unlike primary ones, are not initiated by the firm but by shareholders who will receive the returns of the sale. The distributions are generally underwritten by an investment banking group that buys the entire block from the seller. The shares are then sold on a subscription basis after normal trading hours. Stock exchange or brokerage commissions are not paid by the subscriber, he only pay subscription price. The issues here spin around the speed with which the market adjusts to the effect of the block trade; the possibility of making abnormal returns from price changes; the liquidity and/or information effects, etc. Fundamentally, the sale of a large block may have two effects. First, if it is assumed to carry with it some new information about the security, the price will change (enduringly) to reflect the new information. Second, if buyers must incur extra costs when they agree to the block, there may be a (temporary) decline in price to reflect what has

been described as a price pressure, or distribution effect or liquidity premium. Scholes’s study focused only on permanent price changes.Kraus-Stoll (1972) study associates to open market block trades. They analyzed price effects for all block trades of 10,000 shares or more carried out on the NYSE between July 1, 1968 and September 30, 1969. They had prices for the close of day prior to the block trade, the price right away before the transaction, the block price, and the closing price the day of the block trade. Abnormal performance indices based on daily data were constant with Scholes’ results. More concerned were intraday price effects. There is clear evidence of a price pressure or distribution effect. The stock price recovers significantly from the block price by the end of the trading day. The recovery averaged 713%. Other studies on block trade effects include Mikkleson and Partch (1985), Dann, Mayers and Raab (1977), and so on.New IssuesVarious articles that have analyzed the pricing of new issues of common stock comprise, but not limited to, Reilly and Hatfield (1969), Stickney (1970), McDonald and Fisher (1972), Logue (1973), Stigler (1964), and Shaw (1971). They all faced an apparently inexplicable problem: How could returns on unseasoned issues be adjusted for risk if time series data on pre-issue prices were nonexistent? An ingenious way around this difficulty was employed by Ibbotson (1975). Portfolios of new issues with identical seasoning (defined as the number of months since issue) were formed. The monthly return on the XYZ company in March 1964, say two months after its issue, was coordinated with the market return that month, resulting in one pair of returns for a portfolio of two months seasoning. Ibbotson could compute a covariance with the market from the collected vector of returns, and thus expected the respective stocks’ systematic risk. Applying the experiential market line, he expected abnormal performance indices in the month of first issue (initial performance from the offering data price to the end of the first month) and in the aftermarket (months following the initial issue). From 2650 new issues between 1960 and 1969, Ibbotson arbitrarily selected one new issue for each of the 120 calendar months.The probable systematic risk in the month of issue was 2.26 and the abnormal return was estimated to be 11.4%. Even after transaction costs, this makes up a statistically important provide abnormal return. Therefore, either the offering price is set too low or investors systematically overvalue new issues at the end of the first month of seasoning. Ibbotson concentrated on the possibility that offering prices determined by the investment banker are systematically set below the fair market value of the security. He asserted that the evidence cannot allow us to reject the null hypothesis that after markets are efficient, although it is interesting to note that returns in 7 out of 10 periods show negative values. Weinstein (1978) studied the price behaviour of newly issued corporate bonds by measuring their excess holding period returns. Surplus returns were defined as the distinction between the return on the ith newly issued bond and a portfolio of seasoned bonds are similar to those of Ibbotson (1975) for newly issued stock, namely, that the offering price is below the market equilibrium price but that the aftermarket is well-organized. Weinstein found a .383% rate of return during the first month and only a .06% rates of return over the next six months.Stock SplitsThe best acknowledged study of stock splits was conducted by Fama, Fisher, Jensen, and Roll (1969). Since stock splits are often linked with increased dividend payouts, it would be expected that split announcements would contain some economic information. Cumulative average residuals were calculated from the simple market model, using monthly data for a period of 60 months around the split ex data for 940 splits between January 1927 and December 1959. Fama et. al. found that there was considerable market reaction prior to the split announcement. In fact, the average cumulative abnormal return for the 30-month period up to the month of announcement was in excess of 30%. This would seem to point out that splits precipitate abnormal returns. Nevertheless, such a conclusion lacks economic logic. The run-up in the cumulative average returns prior to the stock split can be explicated by selection bias. The study recommended that stock splits might be interpreted by investors as a

message about future changes in the firm’s expected cash flows. They speculated that stock splits might be interpreted a message about dividend increases, which in turn entail that the managers of the firm feel secure that it can maintain a enduringly higher of cash flows. To test this hypothesis, the sample was divided into those firms that augmented their dividends beyond the average for the market in the interval following the split and those that paid out lower dividends. The results reveal that stocks in the dividend “increased” class have somewhat positive returns following the split. This is regular with the hypothesis that splits are interpreted as messages about dividend increases. Nevertheless, a dividend boost does not always follow a split. For this reason, the to some extent positive abnormal return for the dividend – increase group reflects small price adjustments that occur when the market is extremely sure of the increase. On the contrary, the cumulative average residuals of split-up stocks with poor dividend performance decline until about a year after the split, by which time it must be very clear that the predictable dividend increase is not imminent. A synthesis that on average, the market makes unbiased estimates about security returns and by extension, prices. The study confirms the semi-strong form of efficiency. The split per se has no effect on shareholder wealth. Rather, it merely serves as a message about the future prospects of the firm. Therefore, splits have benefits as signaling devices. There seems to be no way to use a split to increase one’s expected returns, unless, of course, inside information regarding the split or succeeding dividend behaviour is available. Brennan and Copeland(1987) provide a signaling theory explanation for stock splits and show that it is consistent with the data. This study shows that the lower the target price to which the firm splits, the greater assurance management has, and the larger will be the announcement residual.Mutual Fund PerformanceThe performance of mutual funds has been examined by a number of authors such as Friend, et.al.(1962) who examined 189 funds from December 1952 to September 1958; Friend and Vickers (1965); Sharpe (1966) who scaled the performance of 36 mutual funds from 1954 to 1963; Treynor (1965); Jensen(1968) whose analysis covered the period 1945 to 1964; Jensen (1969) covering 115 mutual funds from 1955 to 1964; Friend, Blume and Crockett (1970) who studied 136 mutual funds from 1960 into 1969; Williamson (1972) studied 180 mutual funds from 1961 to 1970. Even though the samples of firms, time periods, and performance measures differed to some extent between these studies, their results were extraordinarily similar. On the average, net of costs, mutual funds did no better than an individual investor would expect if he purchased a diversified portfolio of similar risk. In fact, when all the fund’s expenses were well thought-out, a majority of funds did worse than a randomly selected portfolio would have done. Therefore, these studies show that the net performance of mutual funds is the same as that for a immature investment strategy.Supporting Models of the EMHThis section presents a concise mathematical representation of the various models that have found extensive use in the EMH research.A. The Expected-Returns ModelThe model, suggested by Fama (1970), is given by:

Zi,t+1 = ri,t+1, – E[ri,t+1/*t]with E[zi,t+1,/*t] = 0

Where Zi,t+1, is the unexpected return for security i in period t+1, the difference between the observed return ri,t+1, and the unexpected return based on the information set *t. The expected return could, for instance, be determined by the CAPM.B. The Capital Asset Pricing Model (CAPM)The CAPM, as developed by Sharpe (1964), Linter (1965) and Mossin (1966), may be mathematically expressed as:E(rit) = rft +[E(rmt) – rft] βi + εit (B1)

Other CAPM – related models are the market model and the empirical market line. The market models argues that returns on any security are linearly related to returns on a “market” portfolio. Mathematically described thus:

rit = ai +βi Rmt + εit (B2)where E(εit) = 0σ(Rmt, εit) = 0σ(εit εit) = 0rit = return on security i in period trmt = general market factor in period tεit = the stochastic portion of the individualistic factor representing the part of security i’s return

which is independent of Rmt.ai, βi = intercept and slope coefficients respectively, which are assumed to be constant over the

time period during which the model is fit to the available data.It is instructive to note that the general market factor in equation (B2) is designed to reflect general market and economic conditions that are related to the returns on a particular security. This is a different notion than the return on the market portfolio in the CAPM given by rmt.Sometimes, we see the empirical market line which is expressed as:

rit = Yot + Yit βit + εit (B3)Although related to the CAPM, it does not require the intercept term to equal the risk – free rate. Instead, both the intercept Yot, and the slope. Yit, are the best linear estimates taken from cross-section data each time period (typically each month). Furthermore, it has the advantage that no parameters are assumed to be constant over time.All three models use the residual term εit, as a measure of risk-adjusted abnormal performance. Nevertheless, only one of the models, the CAPM, relies exactly on the theoretical specification of the Sharpe-Litner-Mossin Model.C. The Abnormal Performance Index (API)Performance measures of mutual funds include the Sharpe Index (reward to variability ratio), Treynor Index and Jensen Abnormal performance.

Shape Index = (rit – rft)/σi (C1)Treynor Index = (rit – rft)/βi (C2)Abnormal Performance = ait = (rit – rft) – [βi (rmt – rft)] (C3)

Whereri = return of the ith mutual fundrf = return on a risk free asset (generally Treasury bills)σi = the standard deviation of return on the ith mutual fundβi = the estimated systematic risk of the ith mutual fund

Accordingly the researchers have used the models above together with the expected-returns and CAPM to examine empirically the effects of accounting numbers. One of other more imaginative developments in this approach was the API by Ball and Brown (1968) to study the association between unexpected changes in accounting earnings and unexpected changes in prices. The unexpected price changes are aggregated (for portfolios formed using the sign of the earnings forecast error) using the relationship:

API = 1 ∏(1 + εit) – 1 (C)Nit

whereT = number of time periods: t = 1,2, …. , TN = number of securities: I = 1,2, …. , Nεit = individualistic component of rit, or, alternatively, the forecast error.

The API traces out the value of a naira invested in equal amounts in each security in the portfolio from time t up to T. At time T the earnings number is assumed to be made public.As Beaver (1972) notes, the API has an appealing intuitive interpretation. It represents one measure of the value of the information contained in the earnings number (actually the sign of the earnings forecast error) T months prior to the release of the earnings number. In this sense, the API concept has some aspects of similarity to the notion of perfect information as the concept is used in decision theory. The analogy is not perfect, nevertheless, for the API is an ex post concept while the value of perfect information is an ex ante notion.It suffices to note that the discussion of the foregoing models given here is intentionally brief. A more extensive coverage is available in Beaver (1972). Dyckman et. al. (1975), Copeland and Mayers (1982), Sharpe and Cooper (1972), and Brealey & Myers (1996). INFORMATION ADEQUACY AND REDUNDANCIES IN ANNUAL FINANCIAL REPORTS (AFRs)A purpose of accounting information is to reduce uncertainty about a state of affairs. To achieve this objective, accounting researchers and writers have, generally, advocated more accounting information revelation. For instance, in the Nigerian context, Inanga (1976) indicated that the information content of published AFRs of public companies are insufficient for the users’ requirements. He has, hence, suggested the disclosure of other information such as protected cash flow and future dividend levels. Some scholars have noted, nevertheless, that the message which contains the principal number of bits of information does not essentially lead to the highest payoff (with respect to extent of vagueness reduction). Also, the incremental benefit of any additional disclosure depends upon the measure of sufficiency of the accessible information set. For this reason, it is noted that it is always preferable to assess the extent to which the existing information set meets the users’ requirements to enable one to determine the extent of the need for extra information disclosure. This view is consistent with the information economics approach to accounting information disclosure decisions. It treats information as any other (normal) economic good, the demand for which constitutes a problem of economic choice.The additional revelation syndrome overlooks problems relating to the limited capacity of people at processing information and those of information overload. Moreover, empirical evidence has indicated that the usefulness of an information item to a decision task may be dependent upon the availability or non-availability of other competing information in the decision environment. For example, empirical research efforts in accounting and efficient markets has been to see if accounting provides information to the market. Studies indicate a market reaction that is anticipatory to the release of accounting data. For this reason, AFRs do not possess a monopoly on these data and competing sources may shovel their formal release. This insight tends to diminish the perceived information content of accounting reports.In a related research, carried our by Ariyo & Soyode (1985), on information sufficiency and redundancies contained in published AFRs. Furthermore, the results of the correlation matric among the variables used for the regression analysis, shows a high degree of intercorrelation among the selected variables (cues). The results, hence, suggest a great deal of redundancies in the simultaneous disclosure in reports of data relating to earnings, dividends, and cash flow. The results further show that the gearing ratio (GR) data is not considerably correlated with any of the other cues, suggesting that information conveyed by GR is different from those of other cues.The Ariyo & Soyode study sought to evolve an optimal trade-off between lack of relevant information and over-abundance of immaterial information. The study suggests that it is desirable to recognize the type of additional information necessary to avoid over-supply of superfluous information.

Summary Numerous scientific research focusing on the stock market has not only developed new theories on capital markets but refined existing ones which are regarded sophisticated and efficient in the explanation of pertinent information. The key focus of this chapter was to review some past financial

studies on market efficiency particularly informational efficiency with a vision to bringing out the behavioral pattern that reflects the Efficient Market Hypothesis dependence on the activities of arbitrageurs and experts who create demand and supply patterns to sustain the market in equilibrium. It is also showed that the value of any information structure should be regarded as net of costs, so that any claim to abnormal returns as a result of monopoly of relevant information may not be important relative to the cost of obtaining the information which is applicable to both portfolio managers and individual investors. In this chapter, we considered the affect of accounting information on stock prices; block trades; new issues; stock splits and mutual fund performance. Most of the evidence are constant with the weak and semistrong forms of market efficiency but irregular with the strong form. In certain states of affairs, individuals with inside information appear to be able to earn abnormal returns. This fact may depict an informational inefficiency rather than general which lend authority to the theory of arbitrage capital markets inefficiency. Furthermore, block traders can earn abnormal returns when they trade at the block price as in purchases of innovative equity issues. Additional research could be performed on the impact of capital structure decisions of firms on stock prices..A rising field of research, referred to as Behavioural Finance, studies how cognitive or emotional biases, which are individual or collective, create anomalies in market prices and returns and other deviations from the EMH. Behavioural models usually put together insights from psychology with neo-classical economic theory. Nevertheless, EMH proponents opine that any observed anomalies will eventually be priced out of the market or explained by appeal to microstructure. They further show the inevitability to differentiate between individual biases and social biases; the former can be averaged out by the market, while the other generate feedback loops that drive the market further away from the equilibrium of the ‘fair price’.

REVIEW QUESTIONS1. Suppose you know with certainty that the ABC Capital Corporation will pay a dividend of TZS10

per share on every January 1 forever. The continuously compounded risk-free rate is 5% (also forever).(a) Graph the price path of Clark Capital common stock over time.(b) Is this (highly artificial) example a random walk? A martingale? A submartingale? (Why?)

2. Given the following situations, determine in each case whether or not the hypothesis of an efficient capital market (semistrong form) is contradicted.(a) Through the introduction of a complex computer program into the analysis of past stock

price changes, a brokerage firm is able to predict price movements well enough to earn a consistent 3% profit, adjusted for risk, above normal market returns.

(b) On the average, investors in the stock market this year are expected to earn a positive return (profit) on their investment. Some investors will earn considerably more than others.

(c) You have discovered that the square root of any given stock price multiplied by the day of the month provides an indication of the direction in price movement of that particular stock with a probability of .7.

(d) A securities and Exchange Commission (SEC) suit was filed against Tanzania Sulphur Company in 1965 for the reason that its corporate employees had made unusually high profits on company stock that they had purchased after exploratory drilling had started in Dar es Salam (in 1959) and before stock prices rose dramatically (in 1964) with the announcement of the discovery of large mineral deposits in Dar es Salam.

3. The First National Bank has been losing money on automobile consumer loans and is considering the implementation of a new loan procedure that requires a credit check on loan applicants. Experience indicates that 82% of the loans were paid off, whereas the remainder defaulted. Nevertheless, if the credit check is run, the probabilities can be revised as follows:

Favourable Credit Check

Unfavourable Credit Check

Loan is paid .9 .5Loan is defaulted

.1 .5

An estimated 80% of the loan applicants receive a favourable credit check. Assume that the blank earns 18% on successful loans, loses 100% on defaulted loans, and suffers an opportunity cost of 0% when the loan is not granted and would have defaulted. If the cost of a credit check is 5% of the value of the loan and the blank is risk neutral, should the blank go ahead with the new policy?

4. Tanzanian Foods Ltd., one of the nation’s largest consumer products firms, is trying to decide whether it should spend TZS5 million to the test market a new ready-to-eat product (called Kidwich), to proceed directly to a nationwide marketing effort, or to cancel the product. The expected payoffs (in millions of TZSs) from cancellation versus nationwide marketing are given below:

ActionMarket conditions

Cancel Go Nationwide

No acceptance 0 –10Marginal 0 10Success 0 80

Prior experience with nationwide marketing efforts has been:

Market Conditions ProbabilityNo acceptance .6Marginal .3Success .1

If the firm decides to test market the product, the following information becomes available:ProbabilityNo Acceptance

Marginal Success

Outcome Predicted by the Test Market

No Acceptance .9 .1 0

Marginal .1 .7 .2Success .1 .3 .6

For example, if the test market results predict a success, there is a 60% chance that the nationwide marketing effort really will be a success but a 30% chance it will be marginal and a 10% chance it will have no acceptance.

(a) If the firm is risk neutral, should it test market the product or not?(b) If the firm is risk averse with a utility function

U(W) = ln(W + 11),should it test market the product or not?

5. The efficient market hypothesis implies that abnormal returns are expected to be zero. Yet in order for markets to be efficient, arbitrageurs must be able to force prices back into equilibrium. If they earn profits in doing so, is this fact inconsistent with market efficiency?

6. Read the following:(a) In a poker game with six players, you can except to lose 83% of the time. How can this

still be martingale?(b) In the options market, call options expire unexercised over 80% of the time. Thus the

option holders frequently lose all their investment. Does this imply that the options market is not a fair game? Not a martingale? Not a submartingale?

7. If securities markets are efficient, what is the NPV of any security, regardless of risk?8. From time to time the federal government considers passing into law an excess profits tax on

U.S. corporations. Given what you know about efficient markets and the CAPM, how would you define excess profits? What would be the effect of an excess profits tax on the investor?

9. State the assumptions inherent in this statement: A condition for market efficiency is that there be no second-order stochastic dominance.

CASE STUDYAccessing East African Capital

Accessing capital markets in Tanzania and East Africa through cross-listing may not be as difficult as at first thought, assert Nimrod Mkono, Clark Arrington, and Bijal Mehta, Mkono & Co Advocates Tanzania embarked on a programme of privatization in the early 1990s as it shifted away from being a quasi-socialist to a more market driven economy. As part of this initiative, efforts were made to develop, broaden and deepen the financial sector. This included providing opportunities for broader local ownership and expanding long term financing alternatives and opportunities through the development of capital markets.

The first step in furthering the development and expansion of capital markets in Tanzania was the creation of an overall regulatory agency, the Capital Markets and Securities Authority (The Securities Authority). The agency was established under the Capital Markets and Securities Act, 1994 (The

Securities Authority) as an autonomous body “for the purpose of promoting and facilitating the development of an orderly, fair and efficient capital market and security authority in Tanzania”. The Securities Authority is the regulatory and supervisory body for the capital markets. It licenses and regulates investment intermediaries and investment professionals, monitors and regulates exchanges and supervises the issuance of and trade in securities. The Securities Authority has since promulgated a number of rules and regulations including those covering guidelines for the issuance of corporate bonds and commercial paper and disclosure guidelines for cross-listing.A necessary adjunct to the creation of the Securities Authority was the establishment of the Dar es Salaam Stock Exchange (DSE). The DSE was incorporated in September 1996 as a private company limited by guarantee with the purpose of facilitating the implementation of financial sector reforms, encouraging wider local share ownership in all companies and facilitating the raising of medium and long-term capital.Trading activities at the DSE commenced on April 15, 1998. Currently the DSE has only the Main Investment Market Segment which is an amalgam of main investments and fixed income securities. An Entrepreneurship Growth Market and an Alternative investment Market are in the process of being created.More than 10 years later, the DSE has 11 domestic companies listed and four foreign cross-listed companies whose total market capitalisation is valued at approximately $3.7bn. In terms of the number of companies listed, the DSE is behind the Nairobi Stock Exchange (NSE) but ahead of the Uganda Securities Exchange (USE) and the Rwandan Capital Markets Advisory Council (CMAC). The NSE was formed in 1954 and is one of the most active capital markets in Africa with more than 50 listed companies. The USE, which was launched in June 1997 and started trading in January 1998, is operated under the jurisdiction of the Capital Markets Authority, which reports to the Central Bank of Uganda and has 12 listed companies of which five are foreign cross-listed companies. The CMAC was established by a prime-ministerial decree in March 2007 and is comprised of a Board of Directors and a Secretariat. The latter oversees the day to day operation of the Rwanda Over-the-Counter Exchange which at present has seven listed companies. Cross border listings in East AfricaCross-border listing, where a firm lists its equity shares for trading in a stock exchange located in a different country has gained significance in East Africa over the past years since the signing of the Treaty for the Establishment of the East African Community (the Treaty). Article 85 (Banking and Capital Market Development) of the Treaty states that the Partner States must undertake to implement within the East African Community (EAC), a capital market development program to be determined by the Council for the purpose of creating a conducive environment for the movement of capital within the EAC. Furthermore the Partner States (which as of July 2009 consisted of Tanzania, Kenya, Uganda and Rwanda) were specifically tasked with promoting co-operation among the stock-exchanges and the capital markets and securities regulators in the EAC. This included establishing within the EAC a mechanism for cross-listing stocks, a rating system of listed companies and an index of trading performance to facilitate the negotiation and sale of shares within and external to the EAC.Further, the Securities Authority issued the Capital Markets and Securities (Foreign Companies Public Offers Eligibility and Cross Listing Requirements) Regulations 2003 and the Capital Markets and Securities (Foreign Companies Public Offers and Cross Listing Requirements) Amendment Regulations, 2005 (“The Regulations”) which are read together and govern the cross listing process. At present, Tanzania has four cross-listed companies (Kenya Airways, East African Breweries, Jubilee Holdings and Kenya Commercial Bank). Their market capitalisation is approximately $2.5bn or more than double the market capitalisation of the domestic listed companies.The process

The regulations apply to all offers of securities to the Tanzanian public by foreign companies which, subsequent to a public offer, intend to list on the DSE as well as to other foreign companies that intend to apply for cross-listing on the DSE. Each foreign company applying for cross-listing must establish a place of business in Tanzania and register as a foreign company under the Companies Act, 2002. A foreign company is then eligible to issue securities to the public in Tanzania subject to the following:

i) Compliance with the provisions of Part XII of The Securities Act, which deals with prospectus requirements, advertisements and compensation;

ii) Listing such securities at a stock exchange in Tanzania subsequent to such public offer; andiii) Compliance with the First Schedule of The Regulations which requires verification that the

laws under which the foreign company is incorporated impose similar requirements as the Companies Act, 2002.

The execution of a Memorandum of Understanding (MOU) between the securities regulator in the jurisdiction where the applicant is listed and the foreign company and the execution of a MOU between the exchange where the applicant’s securities are listed or where listing is being sought and the DSE will satisfy the requirements of the First Schedule of the Regulations. However, the First Schedule can also be satisfied if the securities regulator of the country where the company is listed is a signatory to a relevant multilateral MOU of which Tanzania is also a signatory. An applicant seeking to cross-list in the first tier market of the DSE must be listed in a similar market in other countries where it is listed or cross-listed. Also the company must have minimum issued and fully paid up capital of not less than $700,000 and have net assets of not less than $1.4m. Other requirements include:

• securing at least 1,000 shareholders of which at least 25 percent must be held by the public; and • publication of audited financial statements complying with International Accounting Standards

for at least five years and which show the company was profitable for at least three of the past five years; and

• approval of an Information Memorandum by the Securities Authority.Part Two of the Second Schedule of the Regulations prescribes the form and content requirements of the Information Memorandum. The regulations reflect international norms regarding disclosure requirements. Lastly the securities regulator and the stock exchange of primary listing must each upon notification transmit to the Securities Authority a confidential report on the foreign company regarding the compliance history of the company with the regulations of the stock exchange of primary listing.Tanzania’s commitment In 2002 the Securities Authority enacted the Guidelines on Corporate Governance Practices by Public Listed Companies in Tanzania (the Governance Guidelines) “with the ultimate objective of realising shareholders long-term value while taking into account the interest of other stakeholders”. At the core of the Governance Guidelines is an effort to promote a high standard of self-regulation within public listed companies in Tanzania and achieving an international standard of governance. The provisions highlight practices relating to the board of directors, audit committees, supply and disclosure of information, shareholder participation, etc. Emphasis extends beyond descriptions of accountability within respective roles to principles of operation aimed at fostering efficient communication both within the company and between the company and the Securities Authority. Good corporate governance has become increasingly significant given the multifaceted nature of cross-border listings. Thus, the Partner States are continuing to develop strategies in this area as they work towards attaining a smoother flow of capital in the East African region. Ongoing requirements

In addition to the annual listing fee of 0.05 percent of the market capitalisation of the listed securities and the holding of annual shareholder meetings, listed companies are required to submit semi-annual financial reports to the DES and also to notify the DSE and the public of any previously undisclosed price sensitive information. There are also on-going corporate governance reports regarding the progress attained in complying with the Governance Guidelines. IncentivesTanzania provides numerous incentives for issuers and investors participating in the DSE. They include a reduced corporate income tax rate of 25 percent for three years where 35 percent of the issuer’s shares are held by the public as opposed to a 30 percent rate for unlisted companies. Investors of publicly traded companies realise a zero capital gains tax as opposed to a 10 percent tax on shares of unlisted companies. They also receive a five percent withholding tax on dividend income as opposed to 10 percent on dividends from unlisted companies.The way forwardThe development of cross listing across national stock markets in Tanzania, Kenya, Uganda and Rwanda is a milestone in the EAC’s drive for regional integration. Despite barriers such as wavering political will, differences in settlement procedures and still relative illiquidity, the EAC’s stock markets are braced for continued growth, access and harmonisation. With the emergence of electronic gateways, perhaps a regional stock market is not lingering too far in the future? To properly navigate and take advantage of the dynamic opportunities offered by the East African capital markets it is advisable to hire a reputable and experienced law firm such as Mkono & Co Advocates with regional offices and formal international affiliations.Source: www.mkono.comQuestions

1. What shifted away Tanzania from being a quasi-socialist to a more market driven economy?2. Do you think that the initiative taken by the government of Tanzania to attract foreign

companies and investors would give boost to the national economy? Give your comments with suitable examples.

Unit 3

Efficient Capital Markets: Evidence

Empirical evidence for or against the hypothesis that capital markets are efficient takes many forms. Most of the people and scholars do agree that capital markets are efficient in the weak and semistrong forms but not in the strong form. Generally capital market efficiency has been tested in the large and sophisticated capital markets of developed countries. Hence one must be careful to limit any conclusions to the appropriate arena from which they are drawn. Research into the efficiency of capital markets is an ongoing process, and the work is being extended to include assets other than common stock as well as smaller and less sophisticated marketplaces.

Prices in informationally efficient capital or financial markets should be a sign of all accessible information. A significant result of market efficiency is that investments in publicly traded financial securities, such as stocks and bonds, are zero NPV investments (as opposed to the assumed positive NPVs of most real or productive assets). Investors should anticipate to earn a fair or normal return that is consistent with the risk (defined by CAPM or APT) of the security. Companies should anticipate to get a fair price when they issue securities. Present prices should symbolize a fair and impartial forecast or approximation of the true, intrinsic, or fundamental value of the firm, i.e., the Present Value of all future anticipated cash flows. If this were not the case, we would find a lot of instances where security prices were systematically biased, i.e., either consistently underpriced or overpriced.Some reasons why market efficiency is a critical issue and concept:

1. It impacts the price that the firm will receive for any new stocks and bonds that it may issue. Also, if a firm can sell new stock that is overestimated, it is possibly likely to do such.

2. It affects the cost of capital or necessary rate of return on securities. The cost of capital affects the capital budgeting or new capital expenditure decisions.

3. If you want to link management compensation to stock price or shareholder value, then it is particularly significant that the stock price be representative of the true value of the firm, i.e., stockholders want a stock price that is fair and impartial.

4. An asset’s price should be driven by impartial estimates of future cash flows and the true systematic risk connected with the cash flows. If this were not the case, investors would be able to earn returns that are contradictory with the true level of risk of an asset. Portfolio managers are very interested in any mispricing in the stock market. A mispriced stock would be thought of as cash lying in the street waiting for someone to pick it up.

Normal versus Abnormal ReturnsIf financial markets are not capable, then strategies would survive that can methodically earn above normal or below normal returns, referred to as abnormal returns. Nevertheless, in order to actually calculate any irregular return for any given asset, we first need some Asset Pricing Model such as the Arbitrage Pricing Theory or Capital Asset Pricing Model that gives us an estimate or idea of what the normal or predictable return to that asset should have been.

Abnormal Return = Actual Return observed – Anticipated ReturnThe anticipated or normal return of the asset is based on: (1) the stock’s level of risk and (2) what actually happened with the relevant systematic or macroeconomic source(s) of risk. For instance, in a CAPM world, if the overall market goes down, the stock under investigation would likely also have gone down in price.Reaction of Stock Price to New Information

In an efficient market, prices respond instantaneously to new and material information and fully reflect that information. Delayed responses (under-reaction) and overreaction to new information would suggest that markets are inefficient. Next, we look at three information sets and the corresponding level or form of market efficiency.Information and Forms of Efficient Market Hypothesis (EMH)Three information sets are used to describe the EMH. Note that set 1 is a subset of 2 and that both 1 and 2 are subsets of 3. Each set corresponds to one form of the EMH as discussed below.

1. Historic stock prices and other market related information (e.g., trading volume, etc.)2. Publicly accessible information (this also includes all historical market information)3. All information relevant to a stock (both public and private information)

Three forms of Efficient Market Hypothesis or EMH1. Weak form efficiency: asset prices already replicate all historical market related trading

information such as past prices, returns, trading volume, or trends in volume or prices. The majority of tests show that this information cannot be used to produce or earn abnormal returns after adjusting for risk. This makes sense for the reason that this information is accessible at almost zero cost. If the market is weak-form capable, then technical analysis or chart reading should not produce abnormal returns. Stock prices should strongly follow a random walk. Individuals have a fateful tendency to look for patterns or trends, even in random series. Take a good look at the four graphs on page 10 of these notes. Two of the graphs are actual performance of the U.S. market for a 10-year period, and the other two are computer generated from a random walk model (an upward trend of 10 percent per year with 20 percent annual standard deviation).

2. Semi-strong form efficiency: asset prices already reflect all information that is publicly accessible, i.e., earnings, dividends, analyst forecasts, expectations of the future, etc. Most tests show that material public announcements are accompanied by an immediate change in price. In a semi-strong efficient market, the market's reaction to new and material information should be both instantaneous and unbiased, i.e., without any systematic pattern of over or under reaction. In addition, the market should only react to the extent that new information differs from what had been anticipated. Semi-strong efficiency also means that most financial analysis work or fundamental analysis, based on using public information, should not work. Opportunities may occasionally exist that produce above normal or excess returns. Nevertheless, after the information or strategies become known to the public, they should no longer produce excess returns; e.g., the January effect in small stocks has vanished. Also, a talented investment analyst might still be able to beat the market, provided that he/she is able to consistently interpret information better than the competition can.

3. Strong form efficiency: asset prices by now replicate all private and public information. We know that inside information is very important to anybody that chooses to (illegally in most cases) act upon this information, so the market is undoubtedly not strong form efficient.

Implications of the Semi-Strong Form Market Efficiency: Stock prices are anticipated to increase over time, but future returns are anticipated to be

consistent with the systematic risk. Investments in financial assets are anticipated to be ZERO Net Present Value. This means that

you should expect to earn an average future return that is determined by the systematic risk of the investments.

What if no one performed security analysis? Then the first person that becomes an analyst will find numerous mispriced assets and trading rules that earn excess or abnormal returns. Such profitable opportunities would surely direct to many more individuals entering the analyst field. Competition will speedily start to abolish most of the mispriced assets.

Because of the intense competition, it will become complicated to earn abnormal returns. The marginal benefit of analysis will just equal the marginal cost of analysis for the average analyst or investor.

It therefore follows that individuals should be exceptionally apprehensive of anyone that advertises some investment technique that earns unusual returns. If the technique really works, then any normal person would keep the method undisclosed! This holds for the weak-form market efficiency as well, as many attempt to sell methods for technical analysis.

The Evidence for Market Efficiency Are markets strong form efficient? Certainly not. Are markets weak-form efficient? Evidence strongly suggests yes. Random successes look

appealing, but most tests are not supportive of trading strategies that use market related data. Are markets perfectly semistrong efficient? No. The recent bubble in internet/technology stocks

is a somewhat obvious example of prices not reflecting their fundamental value. Are markets largely semistrong efficient? Yes, based on most of the evidence, especially for

event studies and mutual fund performance.Nevertheless, there are several reported anomalies in stock prices. Do these anomalies represent actual mispricings that a shrewd portfolio manager can convert into abnormal returns (ARs)? No, for the most part. Perhaps they once worked, earlier than they became public knowledge. Also, what might produce apparent or measured abnormal returns on paper might be difficult to really put into practice using actual money.Tests of the EMH fall under three major categories:

1. Tests for random walk in stock prices2. Event Studies3. Performance of professional investors

All these tests evaluate observed stock returns next to returns predicted by the EMH, controlling for systematic risk. We first need to understand what any stock's normal returns should look like, based upon its level of risk. Hence, tests of the EMH are joint tests of market efficiency and the asset pricing model (e.g. the CAPM or APT) used to guess systematic risk.Tests of the Weak Form EMH (do prices follow a Random Walk?)Tests for serial correlation are often used to test the weak form EMH. Positive serial correlation: above (below) normal returns are followed by subsequent above (below) normal returns, or else referred to as momentum. Negative serial correlation: above (below) normal returns are followed by succeeding below (above) normal returns, or else referred to as reversals in returns. Both positive serial correlation and negative serial correlation would entail violations of the weak form EMH if they are found to exist and also if they are economically important after accounting for transactions costs. The greater part of empirical evidence is dependable with the weak form EMH.Tests of Semi-strong Form EfficiencyEvent studies seem at stock price reaction to new information unconfined to the public. Delayed responses or overreaction to new public information entails a violation of the semi-strong form EMH. There is a little varied pragmatic evidence from event studies, but efficiency is generally supported. In addition, the problem of jointly testing the underlying asset pricing model and market efficiency is still unresolved, i.e., first you need to know what the normal returns should look like before any inference can be made relating to market efficiency. Also, always remember that the true asset pricing model is not known.Mutual fund performance studies evaluate the return on actively managed mutual funds to returns on a passive broad-based index mutual fund, e.g., a mutual fund that only attempts to match the S&P 500 or Willshire 5000 indices. Since trading on insider information is illegal, the majority of active fund managers have to depend on public information to prepare their investment strategies. The greater part

of mutual fund studies find that most active mutual fund managers do not do better than passive index funds, supporting the semi-strong form EMH. The Value Line Investment Survey publishes information about stocks that have broad investor interest. Value Line ranks stocks on a scale from 1 to 5. The stocks ranked 1 and 5 are anticipated to have the best and worst future performance, in that order. Research has shown that, on paper, the 1 ranked stocks have without a doubt outperformed the 5 ranked stocks. Nevertheless, the two mutual funds that Value Line manages have in fact underperformed the market, even though they use the Value Line ranking system in formulating their investment decisions. In fact, what looks great on paper cannot always be implemented to produce above normal returns.Some Contrary Views of the EMHSize: Studies have revealed that small firms (measured by market capitalization) earn higher returns than huge firms after adjusting for systematic risk (Beta CAPM risk in these studies). One disagreement on this finding is whether a correct or correctly particular asset pricing model was used to adjust for systematic risk. Returns cannot be said to be above or below normal when you don't have an excellent idea of what the "normal" returns should look approximating.In any case, small firms do not comprise much value in the overall market. Of all the publicly traded firms in the U.S., the largest 10 percent of firms make up around 85 percent of the total value of the U.S. stock market. The smallest 10 percent of firms make up only about 0.3 percent of the overall market value. Even the smallest 30 percent of firms only make up about 2.5 percent of the overall market value.

Temporal Anomalies: Researchers discover several stylish facts (patterns) in stock returns, implying that stock prices do not follow a random walk. Nevertheless, these findings are not strong evidence against the EMH for the reason that these patterns cannot be exploited to earn abnormal profits after accounting for transaction costs. Furthermore, one of these patterns (negative average return on Mondays) has disappeared recently.

Value versus Glamour: This is the latest battleground for the EMH. Several studies show that public information (the ratio of book value of equity to market value of equity or BV/MV) can be used to select stocks that produce abnormal returns. If the finding is not driven by risk, then a challenge to the EMH exists. The verdict on this topic is still not in.

Long-term studies: Many studies, beginning in the early 1990s suggested that the market does not completely or fully react to major corporate events, and long-term (1 to 5 year) abnormal returns exist. This evidence of long-term anomalies following IPOs, stock splits, dividends, stock repurchase announcements, etc. is controversial. The IPO underperformance study by Ritter (1991) is well known and cited in the textbook. Misspecified asset pricing models are the likely cause of most of these findings. More recent studies that use improved asset pricing methods or models have explained or resolved many of these long-term anomalies.

Implications for Corporate Financial Managers1. Can financial managers “fool” investors?

Early studies find that stock prices do not react to changes in accounting methods, unless cash flow is affected. These findings are consistent with the semi-strong form EMH and suggest that restating financial accounting performance or methods is unlikely to increase value unless it can also decrease taxes, bankruptcy costs, or agency costs. Nevertheless, well known recent scandals at Enron, Worldcom, and Tyco, etc., illustrate that managers can sometimes fool the public and investors with deceptive financial accounting practices, but only for so long before the bubble bursts. Perhaps a firm can fool the market only once!

2. Can financial managers “time” security sales?The market usually reacts negatively to an announcement of a sale of new common stock. Nevertheless, early long-term studies found that theses firms have negative abnormal returns in following years (firms that repurchase equity had positive abnormal returns in following years).

These findings suggest that managers “time” equity sales and repurchases correctly. These studies are controversial and have been largely resolved in recent years.

Speculative Bubbles:Sometimes asset prices tend to reach elevated levels (sort of like Alan Greenspan’s 1996 quote of “irrational exuberance”) that are highly inconsistent with fundamental value. Prices of stocks, particularly in a certain industry or sometimes in an entire market, appear to be priced at a level that is far above what would be considered to be a fair price that is based upon risk and realistic anticipated future cash flows. There is what some call the “bigger fool” theory. Someone might have felt like a “fool” for buying Cisco Systems stock in early 2000, but they were hoping that an even more foolish person would soon come along and pay an even higher price. This, of course, can only continue so long before prices eventually crash.In each case of a speculative bubble, asset prices rise to high levels that cannot be sustained. The prices cannot be justified by the asset’s future cash flows, investors begin to realize this, and prices come crashing down and eventually converge around the fundamental price. Some instances of what are thought to be speculative bubbles in the past include the following:

1. Dutch tulip bulb mania of the early 1600s. A classic case of people turning away from their normal productive activities in order to chase the bubble of rising tulip bulb prices.

2. South Sea bubble of the early 1700s in Britain.3. Electric related stocks in the 1880s (yes, this was once new, exciting, and overpriced).4. U.S. stock market of the late 1920s. Crash began in October 1929 and prices fell for the next 3

years. Take a look at Radio Corporation of America (RCA) stock during this period. RCA was the “new economy” stock of its era.

5. “Tronics” boom of the early 1960s. U.S. electronic stocks experienced a bubble.6. “Nifty Fifty” craze of the early 1970s. A bubble in large market capitalization growth stocks, e.g.,

IBM, Xerox, McDonalds, Polaroid, Disney, Sony, etc. These were thought of as Blue Chip stocks that would always be prudent investments for a portfolio, almost regardless of the price paid, since these firms had a history of excellent corporate performance. Price-to-Earnings or P/E ratios of these large company stocks were 60 to 80, while the P/E for the rest of the market was around 15. Nevertheless, no large company can ever grow at a future rate that justifies a current P/E of 60 to 90. This is a classic case of “good company, bad stock”, in which the firm has a bright future, but its stock is simply overpriced (such as the much later case of Cisco Systems in 2000). The Nifty Fifty stocks were hard hit in the recession of 1973 to 1974, typically losing 70 to 80 percent of their value.

7. Japanese stock and real estate bubble of the late 1980s. Nikkei 225 index almost hit 40,000 on the last trading day of 1989. There was no way to justify these prices, which came at a time when Japanese corporate profitability was actually declining. Two years later in 1992 the Nikkei was below 14,000, and even in early 2003, it was below 8000. At the bubble’s peak, the appraised value of all real estate in metropolitan Tokyo exceeded that of the entire United States.

8. Internet and dot.com bubble of 1999 and 2000, accompanied by the idea that we had entered into the “New Economy”, i.e., the “rules” had changed. New industries are often difficult to assess or evaluate and often can be prone to irrational speculation. New industries go on to be part of everyday life, but several have experienced a bubble during their infancy.

During any bubble, many believe that “this time it’s different and the old rules no longer apply”. Such an attitude was certainly true of the Internet bubble. During the Internet bubble one would actually repeatedly hear and read that the “old methods” of stock valuation (essentially discounted Free Cash Flow to Equity) no longer worked in the “New Economy”! The bubbles represent deviations from fundamental or true value and thus represent a violation of market efficiency. Nevertheless, in each

case, the mispricing was eventually corrected in the marketplace. A major downside is that a bubble also represents a temporary misallocation of resources in the economy. The economic slowdown during 2000 and recession of 2001, following the collapse of the Internet/technology bubble, is no mere coincidence. The figures on page 9 illustrates the before and after phases of the Japanese stock market bubble (the Nikkei 225 index) and the Internet bubble (showing the Nasdaq Composite Index, Philadelphia Internet Index, and Cisco Systems stock). The technology laden Nasdaq Composite Index reached 5000 at the peak of the Internet bubble. The figures on page 10 illustrate the U.S. stock market bubble and the RCA (radio) bubble of the late 1920s.EMPIRICAL MODESL USED FOR RESIDUAL ANALYSISBefore discussing the empirical tests of market efficiency, it is worth to review four basic types of empirical models that are frequently employed. The differences between them are crucial. The simplest model, called the market model, simply argues that returns on security j are linearly related to returns on a “market portfolio.” Mathematically, the market model is described by

Rjt = aj + bjRmt + εjt (3.1)The market model is not supported by any theory. It assumes that the slope and intercept terms are constant over the time period during which the model is fit to the accessible data. This is a strong assumption, particularly if the time series is long.The second model uses the capital asset pricing theory. It requires the intercept term to be equal to the risk-free rate, or the rate of return on the minimum-variance zero-beta portfolio, both of which change over time. This CAPM based methodology is given by Eq. (3.1):

Now, nevertheless, that systematic risk is assumed to remain constant over the interval of estimation. The use of CAPM for residual analysis was explained at the end of previous unit.Third, we sometimes see the empirical market line, which was explained earlier in unit 1 was given by equation:

Although related to the CAPM, it does not require the intercept term to equal the risk-free rate. Instead,

both the intercept, , and the slope, , are the best linear estimates taken from cross-section data each time period (typically each month). Furthermore, it has the advantage that no returns, for instance,

(3.2)In this fourth equation, the return of the jth stock in the rth time period is a function of the excess return on the market index over the risk-free rate, the difference in return between a large-capitalization equity portfolio and a small-cap portfolio, and the difference in return between a high and a low book-to-market equity portfolio.All four models use the residual term ε jt, as measure of risk-adjusted abnormal performance. Nevertheless, only one of the models, the second, relies exactly on the theoretical specifications of the Sharpe-Linter capital asset pricing model.In each of the empirical studies discussed, we shall mention the empirical technique by name for the reason that the market model and the multifactor model are not subject to Roll’s critique, whereas the CAPM and the empirical market line are. Thus residual analysis that employs the CAPM or the empirical market line may be subject to criticism.Relevant Information (Earnings Versus Cash Flow)Market efficiency requires that security prices instantaneously and fully reflect all accessible relevant information. But what information is relevant? And how fast do security prices really react to new information? The answers to these questions are of particular interest to corporate officers who report the performance of their firm to the public; to the accounting profession, which audits these reports;

and to the Securities and Exchange Commission, which regulates securities information (For instance, Capital Markets and Securities Authority in Tanzania (CMSA)).The market value of assets is the present value of their cash flow discounted at the appropriate risk-adjusted rate. Investors should care only about the cash flow implications of various corporate decisions. Nevertheless, corporations report accounting definitions, not cash flow, and frequently the two are not related. Does an efficient market look at the effect of managerial decisions on earnings per share (EPS) on cash flow? This is not an unimportant question, for the reason that frequently managers are observed to maximize EPS rather than cash flow for the reason that they believe that the market value of the company depends on reported EPS, when in fact it does not.

Table 3.1: FIFO versus LIFOLIFO FIFO Inventory at

CostRevenue 100 100Cost of goods sold 90 25 Fourth item 90 LIFOOperating income 10 75 Third item 60Taxes at 40 percent 4 30 Second item 40Net income 6 45 First item 25 FIFOEPS (100 shares) .06 .45Cash flow per share .96 .70

Inventory accounting provides a good example of a situation where managerial decisions have opposite effects on EPS and cash flow. During an inflationary economy the cost of producing the most recent inventory continues to rise. On the books, inventory is recorded at cost, so that in the example given in Table 3.1 the fourth item added to the inventory costs more to produce than the first. If management selects to use first-in, first-out (FIFO) accounting, it will record a cost of goods sold of TZS 25 against a revenue of TZS100 when an item is sold from inventory. This result in and EPS of TZS0.45. On the other hand, if LIFO (last-in, first-out) is used, EPS is TZS0.06. The impact of the two accounting treatments on cash flow is in exactly the opposite direction. For the reason that the goods were manufactured in past time periods, the actual costs of production are sunk costs and irrelevant to current decision making. Hence, current cash flows are revenues less taxes. The cost of goods sold is a noncash charge. Hence, with FIFO, cash flow per share is TZS0.70, whereas with LIFO it is TZS0.96. LIFO provides more cash flow for the reason that taxes are lower.If investors really value cash flow and not EPS, we should expect to see stock prices rise when firms announce a switch from FIFO to LIFO accounting during inflationary periods. Sunder (1973, 1975) collected a sample of 110 firms that switched from FIFO to LIFO, between 1946 and 1966 and 22 firms that switched from LIFO to FIFO. His procedure was to look at the pattern of cumulative average residuals from the CAPM. A residual return is the difference between the actual return and the return estimated by the model

The usual technique is to estimate over an interval surrounding the economic event of interest. Taking monthly data, Sunder used all observations of returns except for those occurring plus or minus 12

months around the announcement of the inventory accounting change. He then used the estimated , the actual risk-free rate, and the actual market return during the 24-month period around the announcement date to predict the anticipated return. Differences between estimated and actual returns were then averaged across all companies for each month. The average abnormal return in a given month is

where N = the number of companies.The cumulative average return (CAR) is the sum of average abnormal returns over all months from the start of the data up to an including the current month, T:

whereT = the total number of months being summed (T=1,2, …, M)M = the total number of months in the sample.

If there were no abnormal change in the value of the firm associated with the switch from FIFO to LIFO, we should observe no pattern in the residuals. They would fluctuate around zero and on the average would equal zero. In other words, we would have a fair game. Assuming that risk does not change during the 24-month period, the cumulative announcement of the accounting change. This is consistent with the fact that shareholders actually value cash flow, not EPS. Nevertheless, it does not necessarily mean that a switch to LIFO causes higher value. Almost all studies of this type, which focus on a particular phenomenon, suffer from what has come to be known as postselection bias. In this case, firms may decide to switch to LIFO for the reason that they are already doing well, and their value may have risen for that reason, not for the reason that of the switch in accounting method. Hence, there are two explanations for the big run-up prior to the announcement of a switch from FIFO to LIFO. One is that firms that were doing well were more likely to switch in spite of the lower earnings effect. This interpretation implies higher than anticipated free cash flow (FCF), Cash flow matters. Alternatively, it could be that news of the decision leaked out via the rumor mill, and the market reached to higher anticipated FCF before the announcement. Either way, since FCF is higher, the evidence supports FCF versus earnings as the variable of interest to the market. Either way, Sunder’s results are inconsistent with the fact that shareholders look only at changes in EPS in order to value common stock. He finds no evidence that the switch to LIFO lowered value even though it did lower EPS.Ricks (1982) studied a set of 354 NYSE-and AMEX-listed firms that switched to LIFO in 1974. He computed their earnings “as if” they never switched and found that the firms that switched to LIFO has an average 47 percent increase in their as-if earnings, whereas a matched sample of no change firms had a 2 percent decrease. Ricks also found that the abnormal returns of the switching firms were significantly lower than the matched sample of no change firms. These results are inconsistent with those reported by Sunder.The studies above indicate that investors in efficient markets attempt to evaluate news about the effect of managerial decisions on cash flows–not on EPS. The empirical studies of Sunder [1973, 1975] provide evidence of what is meant by relevant accounting information. By relevant we mean any information about the anticipated distribution of future cash flows. Next, a study by Ball and Brown (1968) provides some evidence about the speed of adjustment as well as the information context of annual reports.Earnings data and cash flows are usually highly correlated. The examples discussed above merely serve to point out some situations where they are not related and hence allow empiricists to distinguish between the two. Ball and Brown used monthly data for a sample of 261 firms between 1946 and 1956 to evaluate the usefulness of information in annual reports. First, they separated the sample into companies that had earnings that were either higher or lower than those predicted by a naïve time series model. Their model for the change in earnings was

(3.3)

Where∆NIjt = the change in earnings per share for the jth firm,∆mt = the change in the average EPS for all firms (other than firm j) in the market.

Next, this regression was used to explain next year’s change in earnings,

(3.4)where

, = coefficients estimated from time-series fits of eq. (3.2) to the data,

= the actual change in market average EPS during the (t+1)th time period.Finally, estimated changes were compared with actual earnings changes. If the actual change was greater than estimated, the company was put into a portfolio where returns were anticipated to be positive, and vice versa.It is quite evident that when earnings are higher than predicted, returns are abnormally high. Furthermore, returns appear to adjust gradually until, by the time of the annual report, almost all the adjustment has occurred. Most of the information contained in the annual report is anticipated by the market before the annual report is released. In fact, anticipation is so accurate that the actual income number does not appear to cause any unusual jumps in the API in the announcement month. Most of the content of the annual report (about 85 percent to 90 percent) is captured by more timely sources of information. Apparently market prices adjust continuously to new information as it becomes publicly accessible throughout the year. The annual report has little new information to add.These results suggest that prices in the marketplace continuously adjust in an unbiased manner to new information. Two implication for chief financial officers are (1) significant new information, which will affect the future cash flows of the firm, should be announced as soon as it becomes accessible so that shareholders can use it without the (presumably greater) expense of discovering it from alternative sources, and (2) it probably does not make any difference whether cash flows are reported in the balance sheet, the income statement, or the footnotes – the market can evaluate the news as long as it is publicly accessible, whatever form it may take.SPEED OF ADJUSTMENT Quarterly Earnings and DividendsThe Ball and Brown study raised the question of whether or not annual reports contain any new information. Studies by Aharony and Swary (1980), Joy, Litzenberger, and McEnally (1977), and Watts (1978) have focused on quarterly earnings reports, where information revealed to the market is (perhaps) more timely than annual reports. They typically use a time-series model to predict quarterly earnings, then form two portfolios of equal risk, one consisting of firms and short in the lower than anticipated earnings firms and short in the lower than portfolio, which is long in the higher than anticipated earnings firms and short in the lower than anticipated earnings firms, is a zero-beta portfolio that (in perfect markets) requires no investment. It is an arbitrage portfolio and should have zero anticipated return. Watts (1978) finds a statistically significant return in the quarter of the announcement–a clear indication that quarterly earnings reports contain new information. Nevertheless, he also finds a statistically significant return in the following quarter and concludes that “the existence of those abnormal returns is evidence that the market is inefficient.”Quarterly earnings reports are sometimes followed by announcements of dividend changes, which also affect the stock price. To study this problem, Aharony and Swary (1980) examine all dividend and earnings announcements within the same quarter that are at least 11 trading days apart. They conclude that both quarterly earnings announcements and dividend change announcements have significant effects on the stock price. But more significantly they find no evidence market inefficiency when the two

types of announcement effects are separated. They used daily data and Watts (1978) used quarterly data, so we cannont be sure that the conclusions of the two studies regarding market inefficiency are inconsistent. All we can say is that unanticipated changes in quarterly dividends and in quarterly earnings both have significant effects on the value of the firm and that more research needs to be done on possible market inefficiencies following the announcement of unanticipated earnings changes.Using intraday records of all transactions for the common stock returns of 96 (large) firms, Patell and Wolfson (1984) were able to estimate the speed of market reaction to disclosures of dividend and earnings information. In a simple trading rule, they bought (sold short) stocks whose dividend and earnings announcements exceeded (fell below) what had been forecast by Value Line Investor Service. The initial price reactions to earnings and dividend change announcements begin with the first pair of price changes following the appearance of the news release on the Broad Tape monitors. Although there was a hint of some activity in the hour or two preceding the Broad Tape news release, by far the largest portion of the price response occurs in the first 5 to 15 minutes after the disclosure. Thus, as per Patell and Wolfson, the market reacts to unanticipated changes in earnings and dividends, and it reacts very quickly.Block TradesDuring a typical day for an actively traded security on a major stock exchange, thousands of shares will be traded, usually in round lots ranging between one hundred and several hundred shares. Nevertheless, occasionally a large block, say 10,000 shares or more, is brought to the floor for trading. The behavior of the marketplace during the time interval around the trading of a large block provides a “laboratory” where the following questions can be investigated:

(1) Does the block trade disrupt the market? (2) If the stock price falls when the block is sold, is the fall a liquidity effect, an information effect, or

both? (3) Can anyone earn abnormal returns from the fall in price? (4) How fast does the market adjust to the effects of a block trade?

In perfect (rather than efficient) capital markets, securities of equal risk are perfect substitutes for each other. For the reason that all individuals are assumed to possess the same information and for the reason that markets are assumed to be frictionless, the number of shares traded in a given security should have no effect on its price. If markets are less than perfect, the sale of a large block may have two effects. First, if it is believed to carry with it some new information about the security, the price will change (permanently) to reflect the new information. Second, if buyers must incur extra costs when they accept the block, there may be a (temporary) decline in price to reflect what has been in various articles described as a price pressure, or distribution effect, or liquidity premium. For instance, if the sale of a large block has both effects, we may expect the price to fall from the price before the trade (–T) to the block price (BP) at the time of the block trade (BT), then to recover quickly from any price pressure effect by the time of the next trade (+T) but to remain at a permanently lower level, which reflects the impact of new information on the value of the security.Scholes (1972) and Kraus and Stoll (1972) provided the first empirical evidence about the price effects of block trading. Scholes used daily returns data to analyze 345 secondary distributions between July 1961 and December 1965. Secondary distributions, unlike primary distributions, are not initiated by the company but by shareholders who will receive the proceeds of the sale. The distributions are usually underwritten by an investment banking group that buys the entire block from the seller. The shares are then sold on a subscription basis after normal trading hours. The subscriber pays only the subscription price and not stock exchange or brokerage commissions. The data accessible to Kraus and Stoll (1972) pertains to open market block trades. They examined price effects for all block trades of 10,000 shares or more carried out on the NYSE between July 1, 1968, and September 30, 1969. They had prices for the close of day before the block trade, the price immediately

prior to the transaction, the block price, and the closing price the day of the block trade. Abnormal performance indices based on daily data were consistent with Scholes’s results. More interesting were intraday price effects. There is clear evidence of a price pressure or distribution effect. The stock price recovers substantially from the block price by the end of the trading day. For instance, a stock that sold for TZS50.00 before the block transaction would have a block price of TZS49.43, but by the end of the day the price would have recorded to TZS49.79.The Scholes and Kraus-Stoll studies find evidence of a permanent price decline that is measured by price drops from the closing price the day before the block trade to the closing price the day of the block transaction. These negative returns seem to persist for at least a month after the block trade. In addition, Kraus and Stoll found evidence of temporary intraday price pressure effects. The implications of these findings are discussed by Dann, Mayers, and Raab (1977), who collected continuous transactions data during the day of a block trade for a sample of 298 blocks between July 1968 and December 1969. The open-to-block price change is at least 4.56 percent for block with purchases of TZS100,000 or more could have earned a net annualized rate of return of 203 percent for the 173 blocks that activated the filter rule. Of course, this represents the maximum realizable rate of return. Nevertheless, it is clear that even after adjusting for risk, transaction costs, and taxes, it is possible to earn rates of return in excess of what any existing theory would call “normal.” This may be interpreted as evidence that capital markets are inefficient in their strong form. Individuals who are notified of the pending block trade and who can participate at the block price before the information becomes publicly accessible do in fact appear to earn excess profits.Nevertheless, Dann, Mayers, and Raab caution us that we may not properly understand all the costs that a buyer faces in a block trade. One possibility is that the specialist (or anyone else) normally holds an optimal utility-maximizing portfolio. In order to accept part of a block trade, which forces the specialist away from the portfolio, he or she will charge a premium rate of return. In this way, what appear to be abnormal returns may actually be fair, competitively determined fees for a service rendered–the service of providing liquidity to a seller.To date, the empirical research into the phenomenon of price changes around a block trade shows that block trades do not disrupt markets and that markets are efficient in the sense that they very quickly (less than 15 minutes) fully reflect all publicly accessible information. There is evidence of both a permanent effect and a (very) temporary liquidity or price pressure effect. The market is efficient in its semistrong form, but the fact that abnormal returns are earned by individuals who participate at the block price may indicate strong-form inefficiency.REJECTION OF STRONG FORM EFFICIENCYA direct test of strong-form efficiency is whether or not insiders with access to information that is not publicly accessible can outperform the market. Jaffe (1974) collected data on insider trading from the Official Summary of Security Transactions and Holdings, published by the Securities and Exchange Commission. He then defined an intensive trading month as one during which there were at least three more insiders selling than buying, or vice versa. If a stock was intensively traded during a given month, it was included in an intensive-trading portfolio. Using the empirical market line, Jaffe then calculated cumulative average residuals. If the stock had intensive selling, its residual (which would presumably be negative) was multiplied by – 1 and added to the portfolio returns, and conversely for intensive buying. For 861 observations during the 1960s, the residuals rose approximately 5 percent in eight months following the intensive-trading event, with 3 percent of the rise occurring in the last six months. These returns are statistically significant and are greater than transactions costs. A sample of insider trading during the 1950s produces similar results. These findings suggest that insiders do earn abnormal returns and that the strong-form hypothesis of market efficiency does not hold.A study by Finnerty (1976) corroborates Haffe’s conclusions. The major difference is that the Finnerty data sample was not restricted to an intensive trading group. By testing the entire population of

insiders, the empirical findings allow an evaluation of the “average” insider returns. The data include over 30,000 individual insider transactions between January 1969 and December 1972. Abnormal returns computed from the market model indicated that insiders are able to “beat the market” on a risk-adjusted basis, both when selling and when buying.A study by Givoly and Palmon (1985) correlates insider trading with subsequent news announcements to see if insiders trade in anticipation of news releases. The surprising result s that there is no relationship between insider trading and news events. Although insider’s transactions are associated with a strong price movement in the direction of the trade during the month following the trade, these price movements occur independently of subsequent publication of news. This leads to conjecture that outside investors accept (blindly) the superior knowledge and follow in the footsteps of insiders.Common Errors of Empirical StudiesThe design and interpretation of empirical studies of market efficiency is a tricky business. This brief selection lists the common pitfalls. Unless research is carefully conducted, results that indicate possible market inefficiencies may simply be faulty research.

1. Biased model of equilibrium returns: Remember that any test of market efficiency is a joint test of the model that defines “normal” returns. If the model is misspecified, then it will not estimate the correct normal returns, and the so-called abnormal returns that result are not evidence of market inefficiency–only a bad model.

2. Specification searches: It is always possible, if one tries long enough, to find a model that seems to beat the market during a test period. Nevertheless, the acid test of whether there is a market inefficiency is to test the model on a different holdout sample period (or periods).

3. Sample selection bias: A portfolio of stocks that split will always show positive abnormal returns prior to the split announcement date. Why? For the reason that only stocks whose prices have risen will decide to split. The very fact that one chooses to study stock splits means that the presplit characteristics of the sample stocks are abnormal. Modeling their behavior relative to the market presplit and expecting their postsplit behavior to be similar is a heroic assumption that is usually wrong.

4. Survivorship bias: If we study the behavior of mutual funds by gathering a list of funds that exist today, then collect data about their historical performance for analysis, we are designing survivorship bias into our research. Why? For the reason that we are looking at only those funds that survived up to today. To avoid survivorship bias we should collect a sample of all funds that existed on a given date in the past, and then collect data about their performance from that date forward. By so doing we will capture all of the funds that ceased to exist between then and now. They probably have lower excess returns, and failure to include them in our sample implies upward-biased results.

5. Biased measures of return: As Fama points out, the calculation of geometric returns over a long period of time can overstate performance. For instance, suppose that an abnormal return of 10 percent is recorded in year 1 on a test portfolio and then compounded, doubling over an additional four years. The cumulative return would be (1.1)(2.0)=2.2. A benchmark portfolio would earn no abnormal return the first year and would also double over the following four years with a cumulative return of (1.0)(2.0) = 2.0. If we were to take the difference between the two, we would conclude that the test portfolio earned an abnormal return of 20 percent over the four years following year 1. This false conclusion can be avoided by measuring cumulative abnormal return as the ratio of geometric returns on the test portfolio to the geometric returns on the benchmark portfolio, for instance, [(1.1)(2.0)/(1.0)(2.0)–1]=10 percent. This approach provides the correct conclusion that 10 percent abnormal return was earned in year 1 and no abnormal return thereafter.

6. Inappropriate portfolio weightings: The economic interpretation of market inefficiency depends on value weightings of outcomes. Often equally weighted abnormal returns are reported with the average being driven by small firms. The researcher then concludes the market is inefficient when in fact only the pricing of small illiquid stocks is inefficient.

7. Failure to distinguish between statistical and economic significance: many studies conclude that the market is inefficient on a statistical basis without actually going the next step. To be economically significant, a trading rule must show arbitrage profits after transaction costs that include paying half of the bid-ask spread, brokerage fees, and net of any price pressure effects. A market is inefficient only if there is both statistical and economic proof of inefficiency.

8. Overestimating the frequency of opportunities for arbitrage: In practice, arbitrage opportunities do not appear every day. For instance, transactions involving very large blocks may offer arbitrage profits, but they do not happen frequently. For the market to be inefficient, the frequency of opportunities must be high enough to allow economically significant arbitrage returns.

STOCK SPLITSWhy do stocks split, and what effect, if any, do splits have on shareholder wealth? The best-known study of stock splits was conducted by Fama, Fisher, Jensen, and Roll (1969). Cumulative average residuals were calculated from the simple market model, using monthly data for an interval of 60 months around the split ex date for 940 splits between January 1927 and December 1959. It plots the cumulative average return for the stock split sample. Positive abnormal returns are observed before the split but not afterward. This would seem to indicate that splits are the cause of the abnormal returns. But such a conclusion has no economic logic to it. The run up in the cumulative average returns prior to the stock split can be explained by selection bias. Stocks split for the reason that their price has increased prior to the split date. Consequently, it should hardly be surprising that when we select a sample of split-up stocks, we observe that they have positive abnormal performance prior to the split date. Selection bias occurs for the reason that we are studying a selected data set of stocks that have been observed to split.Fama et at. (1969) speculated that stock splits might be interpreted by investors as a message about future changes in the firm’s anticipated cash flows. Specifically, they hypothesized that stock splits might be interpreted as a message about dividend increases, which in turn imply that managers of the firm feel confident that it can maintain a permanently higher level of cash flows. To test this hypothesis the sample was divided into those firms that increased their dividends beyond the average for the market in the interval following the split and those that paid out lower dividends. The results reveal that stocks in the dividend “increased” class have slightly positive returns following the split. This is consistent with the hypothesis that splits are interpreted as messages about dividend increases. Of course, a dividend increase does not always follow a split. Therefore the slightly positive abnormal return for the dividend-increase group reflects small price adjustments that occur when the market is absolutely sure of the increase. On the other hand, the cumulative average residuals of split-up stocks with prior dividend performance decline until about a year after the split, by which time it must be very clear that the anticipated dividend increase in not forthcoming. When we combine the results for the dividend increases and decreases, these results are consistent with the hypothesis that on the average the market makes unbiased dividend forecasts for split-up securities and these forecasts are fully reflected in the price of the security by the end of the split month.A study by Grinblatt, Masulis, and Titman (1984) used daily data and looked at shareholder returns on the split announcement date as well as the split ex date. They examined a special subsample of splits where no other announcements were made in the three-day period around the split announcement and where no cash dividends had been declared in the previous three years. For this sample of 125 “pure” stock splits they found a statistically significant announcement return of 3.44 percent. They too interpret stock split announcements as favorable signs about the firm’s future cash flows. Surprisingly, they also

find statistically significant returns (for their sample of 1,360 stock splits) on the ex date. At that time there was no explanation for this result, and it is inconsistent with the earlier Fama et al. study that used monthly returns data.In the same study, Grinblatt, Masulis, and Titman (1984) confirm earlier work on stock dividends by Foster and Vickrey (1978) and Woolridge (1983a, 1983b). The announcement effects for stock dividends are large, 4.90 percent for a sample of 382 stock dividends and 5.89 percent for a smaller sample of 84 stock dividends with no other announcements in a three-day period around the stock dividend announcement. One possible reason for the announcement effect of a stock dividend is that retained earnings must be reduced by the Tanzanian Shilling amount of the stock dividend. Only those companies that are confident they will not run afoul of debt restrictions that require minimum levels of retained earnings will willingly announce a stock dividend. Another reason is that convertible debt is not protected against dilution caused by stock dividends. As with stock splits, there was a significant positive return on the stock dividend ex date (and the day before). No explanation is offered for why the ex date effect is observed.The results of Fama et al. (1969) ate consistent with the semistrong for, of market efficiency. Prices appear to fully reflect information about anticipated cash flows. The split per se has no effect on shareholder wealth. Rather, it merely serves as a message about the future prospects of the firm. Thus splits have benefits as signaling devices. There seems to be no way to use a split to increase one’s anticipated returns, unless, of course, inside information concerning the split or subsequent dividend behavior is accessible.One often hears that stocks split for the reason that there is an “optimal” price range for common stocks. Moving the security price into this range makes the market for trading in the security “wider” or “deeper”; therefore there is more trading liquidity. Copeland (1979) reports that contrary to the above argument, market liquidity is actually lower following a stock split. Trading volume is above argument, market liquidity is actually lower following a stock split. Trading volume is proportionately lower than its presplit level, brokerage revenues (a major portion of transaction costs) are proportionately higher, and bid-ask spreads are higher as a percentage of the bid price.Taken together, these empirical results point to lower postsplit liquidity. Therefore we can say that the market for split-up securities has lower operational efficiency relative to its presplit level. Ohlson and Penman (1985) report that the postsplit return standard deviation for split-up stocks exceeds the presplit return standard deviation by an average of 30 percent. Lower liquidity and higher return variance are both costs of splitting.Brennan and Copeland (1988) provide a costly signaling theory explanation for stock splits and show that it is consistent with the data. The intuition can be explained as follows. Suppose that managers know the future prospects of their firm better than the market does. Furthermore, assume that there are two firms with a price of TZS60 per share that are alike in every way except that the managers of firm A know it has a bright future while the managers of firm B except only average performance. Managers of both firms know that if they decide to announce a split, their shareholders will suffer from the higher transaction cost documented by Copeland (1979). Nevertheless, the successful firm A will bear these costs only temporarily (for the reason that its managers have positive information about its future), while firm B will bear them indefinitely. Therefore firm A will signal its bright future with a stock split, and the signal will not be mimicked by firm B. Costly signaling creates a separating equilibrium. As a result, A’s price will rise at the time of the announcement so as to reflect the present value of its future prospects. Furthermore, the lower the target price to which the firm splits, the greater confidence management has, and the larger will be the announcement residual. Empirical results by Brennan and Copeland (1987) confirm this prediction.Dharan and Ilenberry (1995) and Rankine, and Stice (1996) have reported roughly 7 percent positive abnormal returns in the year after the split ex date. If valid, these results would be inconsistent with

semistrong-form market efficiency. Fama (1998) notes that these studies use buy-and-hold returns in the following way. Suppose that the test group of firms record a 10 percent abnormal return during the announcement year and the benchmark portfolio records a 0 percent abnormal return. Then assume that both portfolios record a cumulative 100 percent buy-and-hold return (BAHR) over the next four years. The BAHR for the test group will be 1.1×2.0=2.20 and the BAHR for the benchmark group will be 1.0 × 2.0, leading to the mistaken conclusion that an abnormal return of 20 percent had been earned over the five-year period including the year of the announcement. One way to correct this problem is to take the ratios of the cumulative returns rather than their difference. In this case we are back to a ratio of 1.1, that is, to a 10 percent abnormal return (in the announcement year)The Value Line Investor SurveyHundreds of investment advisory services sell advice that predicts the performance of various types of assets. Perhaps the largest is the Value Line Investor Survey. Employing over 200 people, it ranks around 1,700 securities each week. Securities are ranked 1 to 5 (with 1 being highest), based on their anticipated price performance relative to the other stocks covered in the survey. Security rankings result from a complex filter rule that utilizes four criteria: (1) the earnings and price rank of each security relative to all others, (2) a price momentum factor, (3) year-to-year relative changes in quarterly earnings, and (4) an earnings “surprise” factor. Roughly 53 percent of the securities are ranked third, 18 percent are ranked second or fourth, and 6 percent are ranked first or fifth.The value line predictions have been the subject of many academic studies for the reason that they represent a clear attempt to use historical date in a complex computerized filter rule to try to predict future performance. Black (1971) performed the first systematic study utilizing Jensen’s abnormal performance measure, which will be given later in eq. (3.7). Black’s results indicate statistically significant abnormal performance for equally weighted portfolios formed from stocks ranked 1, 2, 4, and 5 by Value Line and rebalanced monthly. Before transactions costs, portfolios 1 and 5 had risk-adjusted rates of return of + 10 percent and –10 percent, respectively. Even with round-trip transaction costs of 2 percent, the net rate of return for a long position in portfolio 1 would still have been positive, thereby indicating economically significant performance. One problem with these results in the Jensen methodology for measuring portfolio performance. It has been criticized by Roll (1977, 1978), who argues that any methodology based on the capital asset pricing model will measure either (1) no abnormal performance if the market index portfolio is ex post efficient or (2) a meaningless abnormal performance if the index portfolio is ex post inefficient.Copeland and Mayers (1982) and Chen, Copeland, and Mayers (1987) measured Value Line portfolio performance by using a future benchmark technique that avoids selection bias problems associated with using historic benchmarks as well as the known difficulties of using capital asset pricing model benchmarks. The future benchmark technique uses the market model fit using data after the test period where portfolio performance is being measured. The steps in the procedure are:

1. Using the sample from after the test period, calculate the market model equation for the portfolio being evaluated.

2. Use the parameters of the model as a benchmark for computing the portfolio’s unanticipated return during a test period.

3. Repeat the procedure and the test to see whether the mean unanticipated return is significantly different from zero.

In other words, rather than using a particular (perhaps suspect) model (such as the CAPM) of asset pricing as a benchmark, estimate the anticipated returns directly from the data. The future benchmark technique is not without its problems, nevertheless. It assumes that the portfolio characteristics (e.g., risk and dividend yield) remain essentially the same throughout the test and benchmark periods.

Copeland and Mayers find considerably less abnormal performance than Black, who uses the Jensen methodology. Where Black reported (roughly) 20 percent per year for an investor who was long on portfolio 1 and short on portfolio 5, Copeland and Mayers find and annual rate of return of only 6.8 percent. Moreover, only portfolio 5 had statistically significant returns. Nevertheless, any significant performance is a potential violation of semistrong market efficiency. Thus Value Line remains an enigma.Stickel (1985) uses the future benchmark methodology to assess the abnormal presentation resulting from changes in Value Line rankings. He finds statistically noteworthy returns for reclassifications from rank 2 to rank 1 that are three times as large as the returns from reclassifications from 1 to 2. Upgradings from 5 to 4 were not associated with significant abnormal returns. He concludes that the market reacts to Value Line reclassifications as news events and that the price adjustment is larger for smaller firms.Self Tenders and Share RepurchasesLakonishok and Vermaelon (1990), Ikenberry, Laknonishok, and Vermaelen (1995), and Mitchell and Stafford (1977) all find abnormal returns from a trading rule where one purchases a stock on the day following the announcement of a planned self-tender or open-market repurchases a stock on the day following the announcement of a planned self-tender or open-market repurchase plan. For instance, Mitchell and Stafford report three-year buy-and-hold returns of 9 percent for 475 self-tenders during the 1960-1993 time period, after controlling for size and for book-to-market. They also report a 19 percent abnormal return for a portfolio of 2,542 open-market repurchase plants. These abnormal returns are statistically and economically significant and would be consistent with market inefficiency. Nevertheless, most of the abnormal returns vanish when they use a three-factor benchmark that includes as independent variables (1) the difference between the return on the market portfolio and the risk-free rate, (2) the difference between the returns on a small-cap portfolio minus a large-cap portfolio, and (3) the difference between returns on a high book-to-market and a low book-to-market portfolio. Furthermore, the abnormal returns disappear completely if the portfolio of repurchases is value weighted rather than equally weighted. The lesson learned from these results is that measurement of abnormal returns is extremely sensitive to apparently minor changes in technique.SEMISTRONG FORM ANOMALIES: SHORT TERMMutual FundsMutual funds allege that they can provide two types of service to their clients. First, they minimize the amounts of unsystematic risk an investor must face. This is done through efficient diversification in the face of transaction costs. Second, they may be able to use their professional expertise to earn abnormal returns through successful prediction of security prices. This second claim is contradictory to the semistrong form of capital market efficiency unless, for some reason, mutual fund managers can consistently obtain information that is not publicly accessible.A number of studies have focused their attention on the performance of mutual funds. A partial list includes Friend and Vickers (1965), Sharpe (1996), Treynor (1965), Farrar (1962), Friend, Blume, and Crockett (1970), Jensen (1968), Mains (1977), Henricksson (1984), and Grinblatt and Titman (1989). Various performance measures are used. Among them are

(3.5)

(3.6)

Abnormal performance = (3.7)Where

Rj = the return of the jth mutual fund,Rf = the return on a risk-free asset (usually Treasury bills),σj = the estimated standard deviation of return on the jth mutual fund,

= the estimated systematic risk of the jth mutual fundOf these, the abnormal performance measure of Eq. (3.7) makes use of the CAPM. It was developed by Jensen (1968), who used it to test the abnormal performance of 115 mutual funds, using annual data between 1955 and 1964. If the performance index, α, is positive, then after adjusting for risk and for movements in the market index, the abnormal performance of a portfolio is also positive. The average α for returns measured net of costs (such as research costs, management fees, and brokerage commissions) was –1.1 percent per year over the 10 year period. This suggests that on the average the funds were not able to forecast future security prices well enough to cover their expenses. When returns were measured gross of expenses (except brokerage commissions), the average α was –.4 percent per year. Apparently the gross returns were not sufficient to recoup even brokerage commissions.In sum, Jensen’s study of mutual funds provides evidence that the 115 mutual funds, on the average, were not able to predict security prices well enough to outperform a buy-and-hold strategy. In addition, there was very little evidence that any individual fund was able to do better than what might be anticipated from mere random chance. These conclusions held even when fund returns were measured gross of management expenses and brokerage costs. Results obtained are consistent with the hypothesis of capital market efficiency in its semistrong form, for the reason that we may assume that, at the very least, mutual fund managers have access to publicly accessible information. Nevertheless, they do not necessarily imply that mutual funds will not be held by rational investors. On the average the funds do an excellent job of diversification. This may by itself be a socially desirable service to investors.Mains (1977) reexamined the issue of mutual fund performance. He criticized Jensen’s work on two accounts. First, the rates of return were underestimated for the reason that dividends were assumed to be reinvested at year’s end rather than during the quarter they were received and for the reason that when expenses were added back to obtain gross returns, they were added back at year’s end instead of continuously throughout the year. By using monthly data instead of annual data, Mains is able to better estimate both net and gross returns. Second, Jensen assumed that mutual fund betas were stationary

over long periods of time; note that has not time subscript in Eq. (3.7). Using monthly data, Mains

observes lower estimates of and argues that Jensen’s estimates of risk were too high.The abnormal performance results calculated for a sample of 70 mutual funds indicate that as a groups the mutual funds has neutral risk-adjusted performance on a net return basis. On a gross return basis (i.e., before operating expenses and transaction costs), 80 percent of the funds sampled performed positively. This suggests that mutual funds are able to outperform the market well enough to earn back their operating expenses. It is also consistent with the theory of efficient markets given costly information. As you saw in the previous unit that the theoretical work of Cornell and Roll (1981) and Grossman (1980) predicts a market equilibrium where investors who utilize costly information will have higher gross rates of return than their uninformed competitors. But for the reason that information is costly, the equilibrium net rates of return for informed and uninformed investors will be the same. This is just what Main’s work shows. Mutual funds’ gross rates of return are greater than the rate on a randomly selected portfolio of equivalent risk, but when costs (transaction costs and management fees) are subtracted, the net performance of mutual funds is the same as that for a naïve investment strategy.

There have been several studies of mutual fund performance that have seemingly found evidence of market inefficiency for the reason that there is persistence, or autocorrelation, in the abnormal performance of mutual funds over time. Stickel (1992) and Grinblatt and Titman (1992) approach the question in different ways in the same issue of the Journal of Finance. Stickel compares the accuracy of analyst recommendations for two groups, an award-winning All-American Research Team, selected by the Institutional Investor magazine, and all others. The abnormal returns for the first 11 days following large upward forecast revisions indicates that the All-American portfolio has a 0.21 percent greater average impact on security prices than the non-All-Americans, after controlling for the size of the firm and the magnitude of the revision. There is no difference between the two groups when there are downward revisions. Grinblatt and Titman (1992) study the persistence of mutual fund performance. They find that, after controlling for firm size, dividend yield, past returns, interest rate sensitivity, beta, and skewness, there is statistically significant persistence in returns over time and evidence that mutual fund managers can earn abnormal returns. Carhart (1977) reexamines the issue using a sample that is free of survivorship or selection bias, and demonstrates that the persistence in returns is completely explained by common factors in stock returns and investment expenses. His results do not provide evidence of the existence of skilled or informed mutual fund portfolio managers.Dual Purpose FundsDual purpose funds are companies whose only assets are the securities of other companies. Nevertheless, unlike open-end funds, closed-end dual-purpose funds neither issue new shares nor redeem outstanding ones. Investors who wish to own shares in a closed-end fund must purchase fund shares on the open market. The shares are divided into two types, both of which have claim on the same underlying assets. The capital shares of a dual fund pay no dividends and are redeemable at net asset value at the (predetermined) maturity date of the fund. The income shares receive any dividends or income that the fund may earn, subject to a stated minimum cumulative dividend, and are redeemed at a fixed price at the fund’s maturity date. Dual funds were established on the premise that some investors may wish to own a security providing only income, whereas other investors may desire only potential capital gains.There are two very interesting issues that are raised when one observes the market price of closed-end shares. First, the market value of the fund’s capital shares does not equal the net asset value. Most often, the net asset value per share exceeds the actual price per share of the dual fund. In this case the dual fund is said to sell at a discount. Given that a speculator (especially a tax-exempt institution) could buy all of a fund’s outstanding shares and liquidate the fund for its net asset value, it is a mystery why a discount (or premium) can persist. The second issue has to do with whether or not risk-adjusted abnormal rates of return accrue to investors who buy a dual fund when it is selling at a discount, then hold it for a period of time, possibly to maturity.Ingresoll (1976) shows that the capital shares of a dual fund are analogous to call options written on the dividend-paying securities held by the fund. Holders of income shares are entitled to all income produced by the portfolio plus an amount equal to their initial investment payable when the fund matures. If S is the value of the fund at maturity and X is the promised payment to income shares, then capital shareowners receive the maximum of (S–X) or zero, whichever is larger, at maturity. This payoff can be written as

MAX [S–X,0]and is exactly the same as the payoff to a call option. We know, from option pricing theory, that the present value of a call option is bounded from below by S–Xe–rfT. Nevertheless, dual funds are characterized by the fact that cash disbursements in the form of dividends and management fees are made over the life of the fund. If these disbursements are assumed to be continuous, then the present value of the fund is Se–γT, where γ is the rate of payment. Ingersoll shows that, given the cash disbursements, the lower bound on the value of the capital shares must be Se–γT– Xe–rfT. The dashed line,

S – X, is the net asset value of the capital shares. When the fund is above a critical level, Sc, the capital shares will sell at a discount. Below Sc, they sell at premium.When Ingersoll used the option pricing model to estimate the market value of capital shares, he found that it tracked actual prices very well in spite of the fact that no tax effects were taken into account. The data consisted of prices for seven funds on a weekly basis between May 1967 and December 1973. Furthermore, he simulated a trading rule that bought (sold) capital shares when the option model price was above (below) the market price and financed the investment by an opposite position in the securities of the fund and borrowing or lending at a riskless rate. Thus a hedged portfolio was created that required no investment and that had very low risk. The returns on the hedge portfolio were (almost always) insignificantly different from zero. This suggests that even though the capital shares may sell for a discount or premium, they are efficiently priced in a semistrong form sense.When Ingersoll used the option pricing model to estimate the market value of capital shares, he found that it tracked actual prices very well in spite of the fact that no tax effects were taken into account. The data consisted of prices was above (below) the market price and financed the investment by an opposite position in the securities of the fund and borrowing or lending at a riskless rate. Thus a hedged portfolio was created that required no investment and that had very low risk. The returns on the hedge portfolio were (almost always) insignificantly different from zero. This suggests that even though the capital shares may sell for a discount or premium, they are efficiently priced in a semistrong-form sense.Thompson (1978) measures the performance of closed-end dual funds by using all three versions of the empirical models described at the beginning of this chapter. He used monthly data for 23 closed end funds. The longest-lived fund was in existence from 1940 until 1975 (when the data set ended). A trading rule purchased shares in each fund that was selling at a discount at the previous years’ end, then held the fund and reinvested all distributions until the end of the year. The procedure was repeated for each year that data existed, and abnormal returns were then calculated for a portfolio based on this trading strategy. Thompson found that discounted closed-end fund shares tend to outperform the market, adjusted for risk. By one performance measure the annual abnormal return (i.e., the return above the earned by NYSE stocks of equivalent risk) was in excess of 4 percent. It is interesting to note that a trading strategy that purchased shares of funds selling at a premium would have experienced a –7.9 percent per year abnormal return, although the results were not statistically significant.There are several explanations for Thompson’s results. First, the market may be inefficient, at least for tax-exempt institutions that could seemingly be able to profit from the above-mentioned trading rule. Second, so long as taxable investors persist in holding closed-end shares, the gross rates of return before taxes may have to exceed the market equilibrium rate of return in order to compensate for unrealized tax liabilities. Third, abnormal return measures based on the capital asset pricing model may be inappropriate for measuring the performance of closed-end fund capital shares that are call options.An interesting paper by Brauer (1984) reports on the effects of open-ending 14 closed end funds between 1960 to 1981. Funds that were open-ended had larger discounts from net asset value (23.6 percent versus 16.2 percent) and lower management fees (.78 percent versus 1.00 percent) than funds that were not open-ended. Large discounts provide shareholders with greater incentive to open-end their funds, and lower management fees imply less management resistance. These two variables were actually able to predict which funds would be open-ended. In addition, Brauer reports that most of the (large abnormal) returns that resulted from the announcement of open-ending were realized by the end of the announcement month–a result consistent with semistrong-form market efficiency.The problem of analyzing dual funds is not yet completely resolved. The observed discounts (premia) on capital shares may be attributable to (1) unrealized capital gains tax liabilities, (2) fund holdings of letter stock, (3) management and brokerage costs, or (4) the option nature of capital shares. The relative importance of these factors has not yet been completely resolved. There is no good explanation for why all funds selling at a discount have not been open-ended. In addition, there remains some questions

about whether or not abnormal returns can be earned by utilizing trading rules based on observed discounts (premia). Thompson’s (1978) work suggests that abnormal returns are possible, whereas Ingersoll (1976) finds no evidence of abnormal returns.Timing StrategiesIf it were possible to reweight away from high-risk stocks during market downturns and toward them during upturns, it would be possible to outperform a buy-and-hold strategy. Also, timing would be possible if the market or segments of it were to overreact or under react to the arrival of new information (e.g., earnings news or macroeconomic indicators). Empirical evidence of timing is difficult to find for the reason that the decisions of fund managers simultaneously include both the selection of individual stocks for their own merit as well as reweighting for timing reasons. A few studies, nevertheless, examine trading rules directly. Graham and Harvey (1996) and Copeland and Copeland (1999) examine trading rules designed to exploit new information about volatility. Copeland (1999) examine trading rules designed to exploit new information about volatility. Copeland and Copeland find that when an index for forward-looking implied volatility, called VIX, increases surprisingly from its current level, it is possible to earn statistically significant returns by a significant proportion of mutual fund managers reduce their market exposure during periods of increased market volatility. Fleming, Kirby, and Ostdiek (2001) find that volatility-timing strategies earn economically and statistically significant returns. These timing anomalies, although statistically significant, are usually small in magnitude. Nevertheless they represent evidence of semistrong form inefficiency that has not yet been explained away.Stocks That over or UndershootIt is also widely believed that instead of adjusting smoothly, stocks tend to overshoot the equilibrium price. If so, a strategy that is based on this idea might easily earn significant excess returns. Chopra, Lakonishok, and Ritter (1992), like DeBondt and Thaler (1985) before them, report that stocks that are extreme losers at time zero outperform those that were extreme winners over the following five years. The extent of abnormal performance ranges from 6.5 percent per year using annual data to 9.5 percent per year using monthly data. After adjusting for size but before adjusting for beta, extreme losers outperform extreme winners by 9.7 percent per year. In the context of multiple regressions that control for size, beta, and prior returns, there still remains an economically significant overreaction of 5 percent per year. Nevertheless, much of this excess return in seasonal, being earned in January, and seems to be more pronounces among smaller firms. There was no evidence of overreaction among a set of large market cap firms. Furthermore, Bell, Kothari, and Shanken (1995) report that abnormal returns from the strategy become statistically insignificant when changes in the betas of winners and losers are taken into account.CROSS SECTIONAL PUZZLESAny predictable pattern in asset returns may be exploitable and hence judged as evidence against weak-form market efficiency. Even if the pattern cannot be employed directly in a trading rule for the reason that of prohibitive transactions costs, it may enable people who were going to trade anyway to increase their portfolio returns over what they otherwise may have received without knowledge of the pattern. Two statistically significant patterns in stock market returns are the weekend effect and the turn-of-the year effect.Table 3.2: Summary Statistics for Daily Returns on the S&P 500 Stock Index, 1953-1977Means, Standard Deviations, and t-Statistics of the Percent Return from the Close of the Previous Trading Day to the Close of the Day Indicateda

Monday Tuesday Wednesday Thursday Friday1953-1977 Mean –0.1681 0.0157 0.0967 0.0448 0.0873

Standard-deviation

0.8427 0.7267 0.7483 0.6857 0.6600

t-statistic –6.823c 0.746 4.534c 2.283b 4.599c

Observations 1,170 1,193 1,231 1,221 1,209a. Returns for periods including a holiday are omitted. These returns are defined as Rt =ln(Pt/Pt–1)b. 5 percent significance levelc. 0.5 percent significance level

Source: K. French, “Stock Returns and the Weekend Effect,” reprinted from the Journal of Financial Economics, March 1980, 58.Day of the Week EffectFrench (1980) studied daily returns on the Standard and Poor’s composite portfolio of the 500 largest firms on the New York Stock Exchange over the period 1953 to 1977. Table 3.2 shows the summary statistics for returns by day of the week. The negative returns on Monday were highly significant. They were also significantly negative in each of the five-year subperiods that were studied.An immediate natural reaction to explain this phenomenon is that firms wait until after the close of the market on Fridays to announce bad news. The trouble is that soon people would expect such behaviour and discount Friday prices to account for it. In this way negative returns over the weekend would soon be carried off. Another clarification is that negative proceeds are caused by a general “market-closed” effect. French eliminated this possibility by showing that for days following holidays, only Tuesday returns were negative. All other days of the week that followed holidays had positive returns.At present there is no satisfactory explanation for the weekend effect. It is not directly exploitable by a trading rule for the reason that transaction costs of even .25 percent eliminate all profits. Nevertheless, it may be considered a form of market inefficiency for the reason that people who were going to trade anyway can delay purchases planned for Thursday or Friday until Monday and execute sales scheduled for Monday on the preceding Friday.Year End EffectAnother appealing trend in stock prices is the so-called year-end effect, which has been documented by Dyl (1973), Branch (1977), Kein (1983), Reinganum (1983), Roll (1983), and Gultekin and Gultekin (1983). Stock returns decline in December of each year, particularly for small firms and for firms whose price had already declined during the year. The prices increase during the following January. Roll (1983) reported that for 18 consecutive years from 1963 to 1980, average returns of small firms have been better than average returns of large firms on the first trading day of the calendar year. That day’s difference in returns between equally weighted indices of AMEX-and NYSE-listed stocks averaged 1.16 percent over the 18 years. The t-statistics of the difference was 8.18.Again quoting Roll (1983):To put the turn-of-the-year period into perspective, the annual return differential between equally-weighted and value-weighted indices of NYSE and AMEX stocks was 9.31 percent for calendar year 1963-1980 inclusive. Throughout those same years, the average return for the five days of the turn-of-the –year (last day of December and first five days of January) was 3.45 percent. Thus, about 37 percent of the annual degree of difference is due to just five trading days, 67 percent of the annual discrepancy is due to the first twenty trading days of January plus the last day of December.The most probable cause of the year-end effect is tax selling. At least there is a significant correlation amid the realized rates of return during the year and the size of the turn-of-the-year price recovery. Whether or not this occurrence is credulous with a trading rule remains to be seen. Nevertheless, an individual who is going to transact anyway can benefit by altering his or her timing to buy late in December and to sell in early January. Bhardwaj and Brooks (1992) looked at net returns after transaction costs and bid-ask spreads for turn-of-the-year trading strategies for 300 NYSE stocks during the 1982-1986 time period. They bring to a close conclusion that after transactions costs “the January anomaly of low-price stocks outperforming high-price stocks cannot be used to earn abnormal returns.”

SUMMARY

To the highest degree (but not all) evidence suggests that capital markets are well-organized and competent in their weak and semistrong forms, that security prices do the accepted thing to a fair-game model but not precisely to a random walk for the reason that of small first-order dependencies in prices and nonstationarities in the primary price distribution over time, and that the strong form of market efficiency does not hold. Nevertheless, any conclusions about the strong form of market efficiency need to be experienced by the fact that capital market efficiency must be well thought-out jointly with competition and efficiency in markets for information. If insiders have monopolistic access to information, this fact may be considered an inefficiency in the market for information rather than in capital markets. Filter rules have shown that security prices exhibit no dependencies over time, at least down to the level of transaction costs. Therefore, the capital markets are allocationally competent up to the point of operational efficiency. If transaction costs amounted to a greater percentage of value traded, price dependencies for filter rules greater than 1.5 percent might have been found.REVIEW QUESTIONS

1. Roll’s critique of tests of the CAPM shows that if the index portfolio is ex post efficient, it is mathematically impossible for abnormal returns, as measured by the empirical market line, to be statistically different from zero. Yet the Ibbotson study on new issues uses the cross-section empirical market line and finds significant abnormal returns in the month of issue and none in the following months. Given Roll’s critique, this should have been impossible. How can the empirical results be reconciled with the theory?

2. In a study of corporate disclosure by a special committee of the Securities and Exchange Commission, we find the following statement:The “efficient market hypothesis”–which asserts that the current price of a security reflect all publicly accessible information–even if valid, does not negate the necessity of a mandatory disclosure system. This theory is concerned with how the market reacts to disclosed information and is silent as to the optimum amount of information required or whether that optimum should be achieved on a mandatory or voluntary basis; market forces alone are insufficient to cause all material information to be disclosed.Two questions that arise are:(a) What is the difference between efficient markets for securities and efficient markets for

information?(b) What criteria define “material information”?

3. Describe in your own words, what does the empirical evidence on block trading tell us about market efficiency?

4. Which of the following types of information provides a likely opportunity to earn abnormal returns on the market?(a) The latest copy of a company’s annual report.(b) News coming across the NYSE ticker tape that 100,000 shares of Oracle were just traded in a

single block.(c) Advance notice that the XYZ Company is going to split its common stock three for one but

not increase dividend payout.(d) Advance notice that a large new issue of common stock in the ABC Company will be offered

soon.5. Mr. A has received, over the last three months, a solicitation to purchase a service that claims to

be able to forecast movements in the Dow Jones Industrial index. Normally, he does not believe in such things, but the service provides evidence of amazing accuracy. In each of the last three months, it was always right in predicting whether or not the index would move up more than 10 points, stay within a 10-point range or go down by more than 10 points. Would you advise him to purchase the service? Why or why not?

6. The Sigma Mutual Fund (which is not registered with the SEC) guarantees a 2 percent per month (24 percent per year) return on your money. You have looked into the matter and found that they have indeed been able to pay their shareholders the promised return for each of the 18 months they have been in operation. What implications does this have for capital markets? Should you invest?

7. Empirical evidence indicates the mutual funds that have abnormal returns in a given year are successful in attracting abnormally large numbers of new investors the following year. Is this consistent with capital market efficiency?

8. In each of the following situations, explain the extent to which the empirical results offer reliable evidence for (or against) market efficiency.(a) A research study using data for firms continuously listed on the Compustat computer tapes

from 1953 to 1973 finds no evidence of impending bankruptcy cost reflected in stock prices as a firm’s debt/equity ratio increases.

(b) One thousand stockbrokers are surveyed via questionnaire; and their stated investment preferences are classified as per industry groupings. The results can be used to explain rate of return differences across industries.

(c) A study of the relationships between size of type in the Newspaper headline and size of price change (in either direction) in the subsequent day’s stock index reveals a significant positive correlation. Further, when independent subjects are asked to qualify the headline news as good, neutral, or bad, the direction of the following day’s price change (up or down) is discovered to vary with the quality of news (good or bad).

CASE STUDYTHE EFFICIENT MARKET HYPOTHESIS: EVIDENCE FROM TEN AFRICAN STOCK MARKETS

INTRODUCTIONSince the work by Fama (1965, 1970), the efficient market hypothesis (EMH) has become a central part of finance theory. The vast body of research done around this concept is evidence of the interest that the EMH has drawn in both the investment and academic circles. Nevertheless, the bulk of this evidence is from developed markets in the United States and Europe (Groenewold and Ariff, 1998; Mobarek and Keasey, 2000). Little is known about the efficiency of emerging markets, especially those in Africa. The studies available on African stock markets mostly made use of indices data. This is more so in studies that have included more than one market (e.g., Appiah-Kusi and Menyah, 2003).The majority of studies relating to market efficiency with respect to African stock markets have been conducted on the Johannesburg Stock Exchange (JSE). Smith et al. (2002), Smith and Jefferis (2002) and Magnusson and Wydick (2002), among others, found the JSE to be weak-form efficient. Appiah-Kusi and Menyah (2003), nevertheless, concluded on the contrary that the JSE is not weak-form efficient for the period 1990 to 1995, using weekly data.Appiah-Kusi and Menyah (2003) found the stock markets of Botswana, Ghana and the Ivory Coast not to be weak-form efficient for the respective periods investigated. The findings for Ghana and Botswana are consistent with those by Magnusson and Wydick (2002) who found the two markets not to conform to random walk 3 and random walk 2, respectively. Appiah-Kusi and Menyah (2003) concluded that Kenya, Zimbabwe, Egypt, Morocco and Mauritius are weak-form efficient2. Kiweu (1991) and Dickinson and Muragu (1994) reached the same conclusion for Kenya, for the periods 1986 to 1990 and 1979 to 1988, respectively, while Chiwira (2001) found the Zimbabwe Stock Exchange to be weak-form efficient for the period 1995 to 1999. Smith et al. (2002) concluded, on the contrary, that Egypt, Morocco and Mauritius are not weak-form efficient for the periods January 1990 to August 1998 for Morocco and Mauritius, and January 1993 to August 1998 for Egypt. Bundoo (2000) reached the same conclusion for the Stock

Exchange of Mauritius (SEM) for the period 1992 to 1998. Asal (2000) also found the Egyptian Stock Exchange to be weak-form inefficient for the period 1992 to 1996, although he suggested that it was moving towards efficiency in 1997.Studies that have used data on individual stocks used either monthly or weekly data rather than daily data. The limiting factor, among others, as pointed out by Dickinson and Muragu (1994), has been the nonavailability of computerised databases. The other argument for using data measured over longer time intervals is the problem of thin-trading. Increasing the time interval is argued to reduce the potential biases associated with thin-trading by increasing the probability of having at least one trade in the interval (Dickinson and Muragu, 1994). The trade, most probably, would not have taken place at the end of the interval. The result of this is what Bowie (1994) referred to as the “price age” component. This component of thin-trading, though more critical in long time horizons, is usually ignored.Although some studies on African stock markets have acknowledged the thin-trading problem, very few have gone beyond mere acknowledgement of the existence of the problem. The limited studies that tried to address the thin-trading problem include Asal (2000) on the Egyptian Stock Exchange and Appiah-Kusi and Menyah (2003) on eleven African stock markets. Both studies used the adjustment method by Miller, Muthuswany and Whaley (1994) for index returns. AN OVERVIEW OF AFRICAN STOCK MARKETSThe emergenceAbout two-thirds of African stock markets emerged in the late 1980s and early 1990s (Mlambo and Biekpe, 2001; Moss, 2004). The latest arrival is the Douala Stock Exchange in Cameroon, which was established in 2003, making it the youngest stock market on the African continent. Most of these markets were formed at the instigation of government to act as vehicles to privatise state-owned enterprises (Mlambo and Biekpe, 2001; Moss, 2004). This is contrary to misconceptions in some circles that donor agencies, in particular the World Bank and International Monetary Fund, are the driving forces behind the establishment of stock markets in Africa.African stock exchanges are also the smallest in the world in terms of both number of listed stocks and market capitalisation (see Table 1). They are also small relative to their economies, with the market capitalisation of the Nigerian Stock Exchange only representing 8 percent of gross national product (GNP), while in the case of Zimbabwe, Kenya and Ghana, the market capitalisations ranged from 25 percent to 35 percent of GNP (Kenny and Moss, 1998). The largest African stock market in terms of market capitalisation is the JSE (Table 1). According to Senbet (2000), the market capitalisation of the JSE is ten times the combined capitalisation of the rest of Africa’s stock markets and over 100 times their average. Of the more than 2 000 firms listed on Africa’s stock exchanges, the majority are listed on Egypt’s Cairo and Alexandria Stock Exchanges (CASE), although only a small percentage of these are actively traded.Trading times and trading systemsThe majority of stock markets in Africa trade daily, from Monday to Friday (Sunday to Thursday in Egypt), except for a few such as Ghana4, Tanzania, and Uganda, which, in 2002 were trading three times a week. Ghana was trading on Monday, Wednesday and Friday, while Tanzania and Uganda were trading on Tuesday, Wednesday and Thursday (see Table 2). Trading times also vary, ranging from one hour per trading day in Tanzania to the whole business day from 08h00 to 16h30 in Zimbabwe (see Table 2). Trading methods on African stock exchanges vary from open-outcry to call-over to electronic trading systems (Table 2). The Nigerian stock market has replaced the call-over trading system with the automated trading system (ATS). Clearing, settlement and delivery of transactions on the exchange are now done electronically by the Central Securities Clearing System (CSCS). Following the closure of the open outcry trading floor in June 1996, the JSE introduced an order driven, centralised, automated trading system known as the JSE Equities Trading (JET) system. In May 2002 the JET system was converted to the Stock Exchange Trading Systems (SETS) used on the London Stock Exchange (LSE). SETS

is a world-class, flexible and robust trading platform that promise improved liquidity and ensure more efficient functionality. SETS also allows South African based companies access to offshore privileges without having to move offshore. Other markets that adopted the JSE’s trading system include Ghana and the Namibia Stock Exchanges.Migration from an open outcry to an electronic trading system on the Casablanca Stock Exchange (CSE) took place between 4 March 1997 and 15 June 1998. All securities quoted on the CSE are now traded on the electronic trading system. Orders entered by dealers are automatically sorted by price limit and in chronological order, in the "market order book". On the central market, the less liquid securities are quoted on a call auction or fixing basis (once per session). The more liquid securities are quoted on a continuous basis. The electronic trading system automatically downloads to a market information system. This means that data providers can receive real-time market data (time, price, number of shares traded, etc), just as it appears on the dealers' screens.Trading costs, foreign investment and capital flows restrictionsThe sudden supply of vast amounts of mobile capital in the early 1990s created an environment conducive for the emergence of stock markets in Africa. Nevertheless, up until now, portfolio inflows to Africa have been disappointing. One possible reason for this unfavourable scenario is that the acquisition of shares by foreigners is limited on some African stock markets (see Table 3). The prohibitive institutional barriers, trading costs, and factors such as foreign exchange risk, political risk and informational and institutional barriers, act as disincentives to foreign portfolio investments. Some African stock markets still operate in regulatory environments with restrictive capital flow controls on the remittance of capital, capital gains, dividends, interest payments, returns and other earnings (Mlambo and Biekpe, 2001).

TABLE 1: Age and size of African Stock Market

In Malawi, for example, the Reserve Bank of Malawi (RBM) manages the exchange control. Foreign investment capital, whether in the form of equity or loans, needs to be registered with the RBM. Foreign loans and equity investments, the remittance of dividends and capital, among other transfers, require permission from the RBM. In the case of Mozambique, remittance of funds overseas is restricted. Foreign investors can remit loan repayments, dividends and capital if permission has been obtained for amounts above US$5000. In Zimbabwe, dividend remittances in respect of projects approved by the Zimbabwe Investment Centre are allowed at 100 percent of after tax profits. Capital is blocked and may be remitted through 20-year 4 percent government bonds, denominated in Zimbabwean dollars. Capital is paid in 10 equal annual instalments at the end of years 11 to 20.In some countries such as Botswana, Mauritius, Zambia and Uganda, there are no foreign exchange controls. Profits, dividends and capital can be freely repatriated. In Ghana, the financial regime is relatively flexible and allows the free transfer of foreign currency in and out of Ghana. A foreign investor may, subject to approval, operate a foreign currency account with banks in Ghana. In Kenya, the Foreign Investment Protection Act guarantees that foreign investors can convert and repatriate capital freely. The Exchange Control Act was repealed in 1995, removing the last of the restrictions on profit remittances and borrowing. Namibia, as part of the Common Monetary Area (CMA) with South Africa, Swaziland and Lesotho, has unrestricted capital flows for non-residents. Capital, profits and dividends can be repatriated. Exchange control limitations do apply to resident capital flows. South Africa has adopted the approach of gradually abolishing exchange controls. Restrictions on nonresident capital flows were fully liberalised with the abolition of the financial Rand in 1995. Restrictions on residents are gradually being relaxed. In Swaziland, there are no foreign exchange restrictions between CMA members. Nevertheless, exchange controls between CMA members and the rest of the world may not be less strict than those of South Africa.

TABLE 2: Trading Arrangements on African Stock markets, 2002

TABLE 3: Trading costs and foreign investment restrictions

In Tanzania, foreign exchange controls were removed through the Foreign Exchange Act of 1992. Capital transfers are nevertheless still subject to approval by the Bank of Tanzania. Profits and dividends can be fully repatriated. In Nigeria, foreign investors are guaranteed unconditional transfer of their capital and profits. Importation or exportation of foreign exchange above US$5 000 is declared. A domiciliary account can be opened in foreign currency at banks and cash withdrawals from such accounts are permissible. For purposes of exchange control and monitoring the flow of foreign currencies, authorised dealers are required to inform the central bank whenever transfers larger than US$10 000 are made into a domiciliary account.The market regulatorSome of the African stock markets were established on the back of poor regulatory and legislative frameworks. This, among other things, explains why some of these markets lack the capacity to deal with capital market dynamics. Legislations to prevent insider trading are either inadequate or non-existent, and where they exist, enforcement is often poor. The inadequacy of insider trading laws on

African stock markets has enhanced the perception that these markets are not efficient. Insider trading has been one of the problems historically faced by the JSE. South Africa is one of the countries in Africa with insider trading laws. Nevertheless, no prosecution for insider trading has taken place in South Africa due to the inadequacy of the legislation and the existence of lax and unenforceable laws.The successful integration of African markets into the world financial system requires regulatory frameworks that conform to international standards. An appropriate legal and regulatory framework, sufficiently monitored, is a necessity to protect investors and the integrity of the markets. It also helps to instil confidence, a sense of fairness and financial discipline in the market. In most countries the regulator is a government agency, the central bank, finance ministry or an independent commission (see Table 3). In some countries, the capital market is accorded regulatory powers to become a self-regulatory body, or the power to regulate is a shared responsibility between two or more agencies.Market microstructuresAfrican stock markets are also known to be illiquid and characterised by thin trading (Mlambo and Biekpe, 2005), in comparison to stock markets in other regions. According to Kenny and Moss (1998), 8 of the world’s 12 most illiquid stock exchanges in 1995 were in Africa. Expected annual volatility is also, on average, high on the African markets. In Erb, Harvey and Viskanta. (1996), volatility was an average 35,6 percent on African markets compared to 21,4 percent for the Morgan Stanley Capital International (MSCI) developed equity markets, measured using the countries’ risk ratings. Kenny and Moss (1998) suggested that this extreme volatility is a result of the small size nature, lack of liquidity and, often, unstable political and economic environments. The higher expected volatilities on the African stock markets, implying higher risk, nevertheless, also imply higher expected returns as compensation for the risk. Erb, et al. (1996) found the average expected return for Africa to be quite high at 18.4 percent. The ability of African markets to attract investors, despite the high risks, rests upon the relatively higher returns these markets potentially give on investments.African stock markets lack integration with world equity markets and also with each other. Erb, et al. (1996) found the average correlation of African stock markets with world equity markets to be a low 0.05 percent. The segmentation of African markets implies that investors demand compensation in the form of higher expected returns for their risk exposure. Nevertheless, the lack of integration of African stock markets with global equity markets make them potentially good portfolio diversifiers. African markets, except the JSE, were not affected by the Asian crisis due to the lack of interdependence with other global emerging markets (Collins and Biekpe, 2001).The delay to regional integrationA stock market is perceived by African governments to be an indication of integration into the global economy. It is considered to be a sign of international legitimacy and a measure of a country’s modernisation and commitment to private sector-led development (Moss, 2004). Therefore, the emergence of stock markets on the African continent seems more like an explicit response by African governments to globalisation and a desire to be included. In Africa, a stock market is regarded as a symbol of national prestige and progress, similar to a national airline, or a mark of sovereignty, similar to flags or national anthems (Turner, 1999). According to Moss (2004), both the national airline and the stock exchange serve very significant symbolic and practical purposes, yet they come at a cost that may not rationally justify their existence.The symbolic significance of a stock market to African governments has contributed to the lack of progress in integrating regional stock exchanges. The critical prohibiting factor is location. The African Stock Exchange Association (ASEA) has a long-term plan to consolidate the various national stock exchanges into regional hubs based in Johannesburg, Nairobi, Abidjan, Lagos and Cairo. According to Mlambo and Biekpe (2001), an integrated real time network for SADC stock exchanges was expected to be up and running by the year 2006 but it has not yet materialised. The only regional stock exchange in

Africa, the Bourse Régionale des Valeurs Mobilières (BRVM) was made possible by the fact that Cote d’Ivoire had been the only West African franc zone member country with a Stock Exchange and Abidjan was by far the main commercial centre in the zone.Even though Gabon was pushing to have the exchange located in its capital Libreville, the other members resisted, with each of them wanting to play host (Moss, 2004). The BRVM serves the 8 French speaking member countries of the West African Economic and Monetary Union (UMOEA), namely Benin, Burkina Faso, Cote d’Ivoire, Guinea Bissau, Mali, Niger, Senegal and Togo.The problem of location is not only with regional stock exchanges but national stock exchanges as well. The Nigerian Stock Exchange (NSE) was established as the Lagos Stock Exchange in 1960. When the government moved offices to Abuja in 1991, there were fears that the newly privatised enterprises might still use Lagos as their headquarters instead of moving to the new capital, Abuja. Therefore, propositions were made to open another stock exchange in Abuja and a branch of the NSE was recently opened there. If Africa could embrace Information Technology, a physically existing brick-and-mortar exchange trading floor would become less significant as the trend is moving towards electronic trading. Therefore, the issue of location will inevitably be resolved. Electronic trading will also encourage liquidity and efficiency, two characteristics that have been lacking on the African stock markets.DATA AND METHODOLOGYThe markets studied in this paper are Egypt, Kenya, Zimbabwe, Morocco, Mauritius, Tunisia, Ghana, Namibia, Botswana and the West African Regional Stock Exchange (Bourse Regionale des Valeurs Mobilieres - BRVM) in Cote d’Ivoire. The data used in this study are daily closing stock prices and volume traded for individual stocks. The data for Egypt, Kenya, Zimbabwe, Morocco and Mauritius were obtained from DataStream and comparisons were made with samples from the respective stock exchanges and/or stockbrokers (e.g. the Nairobi Stock Exchange for Kenya, Casablanca Stock Exchange for Morocco, BARDNET and Kingdom Stockbrokers for Zimbabwe) to determine reliability and accuracy.For Botswana, Namibia and the BRVM, the data were obtained from the respective stock exchanges while for Ghana and Tunisia at least two sources were consulted. These are Tunisie Valeurs, Tustex andFinanciere de Placement et de Gestion (FPG) for Tunisia; Databank and SDC Brokerages for Ghana.For all the markets in the study, volume traded data were obtained and used to determine the trading frequencies and durations of non-trading of the different stocks. A stock is included in the sample as long as it has been listed for the entire period under consideration; it has not been part of an acquisition or merger during the period under review; it has not been suspended from trade for a period longer than a week; and it has enough data points to make a meaningful analysis. The period examined for each market is shown in Table 4, which also shows the thin-trading frequencies for each market.As can be seen from Table 4, the markets in this study exhibit serious thin-trading for the periods under investigation. For Namibia, the lowest thin-trading frequency is 62 percent. If such a thin-trading frequency is considered serious, then all the stocks in the Namibian sample can be said to have serious thin-trading. Other relatively thin-traded markets are Botswana with the lowest thin-trading frequency of 34 percent and Mauritius with 14 percent. Interestingly, most of the stocks on the Namibian and Botswana stock exchanges have dual listings on the JSE. Probably, trading in these stocks takes place more on the JSE than on these other markets. Because the majority of stocks on the Namibian Stock Exchange (NSX) are dual-listings on the JSE (68 percent of stocks in our sample), the NSX is usually not open for trade whenever there is a holiday in South Africa.This paper uses continuously compounded returns calculated on a trade-to-trade basis and adjusted for interval variability, following Mlambo et al. (2003)7. According to Mlambo and Biekpe (2005), the conclusion on whether a market is efficient or not is subject to the methodology used, especially in adjusting for the thin-trading effect. Bowie (1994) argued in favour of the trade-to-trade approach in measuring stock returns on a thinly traded market.

TABLE 4: Data and thin-trading properties

Notes: *The variability in the sampling periods is due, among other reasons, to the availability of the data. The BRVM, for example, was established towards the end of 1998 and thus the data series could only start after 1998. Mauritius, though established in 1989, changed to daily trading in 1998 and therefore this period was chosen so as not to distort the data series. The Botswana Stock Exchange only started recording transactions data on a daily basis in March 1998 when it introduced the computerised trading system. A large number of stocks were listed on the Tunisian Stock Exchange after the introduction of electronic trading in 1997 and thus the choice of this starting date. For Ghana it was difficult to access data prior to 1998. Thin-trading frequency is measured as the number of days a stock does not trade out of the total number of days the stock exchange was open for trade for the period studied. A thin trading frequency of 62 percent would imply a stock not trading in 62 out of every 100 trading days.

The adjusted trade-to-trade returns are calculated as follows:

(1)where:

is the trade-to-trade return adjusted for the interval effect

is the stock’s traded price in period t

the price of a stock Kt periods in the pastKt is the length of time (in days) between the trade in period t and the previous successive trade

Nevertheless, even when returns are calculated on a trade-to- trade basis, there is still a high prevalence of zero returns (see Table 5). This implies that in most cases stocks changed hands without having an impact on the stock prices. For Ghana, between 57 percent and 82 percent of the trade-to-trade returns are zero returns, while for the BRVM the trade-to-trade zero returns range from 43 percent to 85 percent of the total returns calculated. Only Egypt has relatively low percentages of trade-to-trade zero returns ranging from 2 percent to 23 percent of the trade-to-trade returns calculated. These zero returns are likely to lead to positive serial correlation in the return series. Therefore, the trade-to-trade approach will only reduce, but not eliminate, the bias on our findings towards the rejection of serial independence.EMPIRICAL RESULTSThe returns were tested for normality and all the stocks rejected the normality assumption using the Kolmogorov-Smirnov test at the 1 percent level of significance. Due to the non-normality of the returns series, a nonparametric measure for independence, that is, the runs test, is used. This test is not affected by any extreme values in the return series (Dickinson and Muragu, 1994) and thus does not require constant variance of the data (Barnes, 1986). The hypothesis of independence is tested using the

significance of the standardised Z-values at the 5 percent level of significance. This statistic compares the observed and expected number of runs. The observed number of runs is the sequences of price changes of the same sign. The total expected number of runs (m) is computed as follows, (see Fama, 1965):

(2)where n is the total number of price changes and i n (I =1,2,3) are the numbers of price changes of each sign - positive, negative and zero.

TABLE 5: Analysis of zero returnsCountry/

Stock Exchange

Stock market trading

days

Days actually traded

Zero returns ( percent)

’True’ zero

returns ( percent)

Botswana 1035 65 to 677 81 to 85 9 to 71BRVM 687 34 to 664 55 to 98 43 to 85Egypt 1342 345 to 1277 6 to 78 2 to 23Ghana 753 143 to 684 65 to 94 57 to 82Kenya 1366 38 to 1338 31 to 98 21 to 45Mauritius 1197 575 to 1033 56 to 76 46 to 56Morocco 1349 33 to 1336 22 to 98 18 to 39Namibia 1341 31 to 503 55 to 97 9 to 61Tunisia 1250 107 to 1293 18 to 94 15 to 72Zimbabwe 1353 81 to 1293 40 to 86 31 to 49

Stock market trading days: These are made up of the number of days a stock exchange was open for trade for the period under investigationDays actually traded: These are the number of days when trade took place for each stockZero returns: These are the number of days recording a zero return for each stock as a proportion of the number of days the respective stock exchange was open for trade, and are thus recorded as percentages.‘True’ zero returns: These are the number of days recording a zero return for each stock when the stock actually traded or changed hands as a proportion of the number of days trade actually took place for the stock, Also given in percentages.

As per Fama (1965), for large n, the distribution of m is approximately normal and the Z-value is calculated as follows:

(3)

where r is the observed number of runs, the “½” is the discontinuity adjustment factor and m σ is the standard error of m and is given by:

… (4)where all the variables are as defined before A negative Z-value implies that the observed number of runs is less than the expected number of runs and thus positively correlated. The opposite is true for a positive Z-value.

The majority of stocks for Ghana and Mauritius rejected the random walk hypothesis using the runs test. Whereas 80 percent of stocks rejected the random walk hypothesis for Ghana, on the Stock Exchange of Mauritius, all the stocks in the sample rejected the random walk hypothesis at the 1 percent level of significance. For the other markets, more than half of the stocks for the BRVM (54 percent) and Egypt (54 percent) and exactly half the stocks for Botswana also rejected the random walk hypothesis using the runs test, whilst for Kenya, Zimbabwe, Tunisia and Morocco, less than half of the stocks in the respective samples rejected the hypothesis. For all the markets, except Kenya, the runs test indicates positive correlation tendencies for most stocks, when disregarding significance (see Table 6). The deviations from the random walk by the stocks on these markets suggest that the possibility of detecting patterns, which can be profitably traded upon, cannot be entirely dismissed. One clear exception is Namibia, where all the stocks in the sample, except one, exhibited random walk behaviour at the 5 percent level of significance. Kenya and Zimbabwe are also considered to be weak-form efficient since the majority of stocks conformed to the random walk hypothesis.Higher order serial correlationBekaert and Harvey (2002) indicated that emerging market stock returns exhibit higher order serial correlation. To investigate this assertion in the current research, a hypothesis that the correlation coefficients of trade-to-trade returns at all lags are zero is tested against the alternative that not all correlation coefficients are zero. This hypothesis is tested using the Box-Ljung Q-statistic, which is chi-square distributed with k degrees of freedom (χK

2 ). The null hypothesis that all the ten correlation coefficients are zero is rejected if the Q-statistic for the 10 lags is significant at the 5 percent level of significance. The results, as summarised in Table 7, show that more than half of the stocks for each market, except Namibia and Zimbabwe, exhibit significant higher order serial correlation at the 5 percent level of significance.

TABLE 6: Number and proportion of stocks with significant Z-values for the Runs test

TABLE 7: Number of significant higher order serial correlation coefficients and the proportions of significant Box-Ljung Q-statistics

Table 7 also shows that the numbers of significant coefficients decrease with increasing lags for both the autocorrelation function (ACF) and the partial autocorrelation function (PACF), but only gradually. Hence, while stock returns in the immediate past provide information that play a significant role in determining future returns, the information becomes less and less useful the further away in the past one looks. Nevertheless, for the African stock markets in this study, the importance of historical price information only dissipates gradually and is hence not totally irrelevant in forecasting future returns.Discussion of resultsThe positive serial correlation observed on most of the markets is not surprising considering that daily data was used for an average period of 5 years. Positive serial correlation is usually considered to be a predictability phenomenon of the short-run, while negative serial correlation is mostly a long run predictability phenomenon. The positive serial correlation on African stock markets might also be a result of institutions imitating spreading their trades over several days to lessen the impact of trades in large volumes on the market (Asal, 2000). Most significantly, the prevalence of zero returns on these African stock markets as discussed in section 3 could have contributed to the nature of the results. The weak-form efficiency of the NSX can probably be explained by the market’s positive correlation with the JSE due to the significant number of stocks that are dual-listed on both markets. Tyandela and Biekpe (2001) found the correlation of the two markets to be 90 percent and the highest for all the market correlations that they studied. Considering some of the recent studies, the JSE was found to be efficient in Magnusson and Wydick (2002), Smith, Jefferis and Ryoo (2002) and Smith and Jefferis (2002). If the JSE is weak-form efficient, then one can argue that this efficiency filters through to the NSX since almost the same stocks are traded on both markets. About two-thirds of the stocks, which make up the sample for Namibia, are dual-listed on the JSE. The efficiency of the NSX can thus be said to be a spill-over from, or a reflection of, the weak-form efficiency of the JSE.The efficiency of Kenya and Zimbabwe is also not surprising since the two markets are among the oldest in Africa. The two markets, probably, gained some sophistication over the years, enabling them to interpret and incorporate information into prices speedily. Fama (1965, p38) highlighted the importance of many sophisticated traders in a market, who can recognise situations where the price of a stock runs well above or below its intrinsic value, and who through their actions would cause such price bubbles to burst before they get underway. The deviation from the random walk portrayed by the sampled stocks listed on the stock exchange of Mauritius cannot be readily explained. The rejection of the random walk hypothesis on this market might indicate the availability of exploitable profit opportunities, which may be due to investors not reacting quickly to new information and thus a slow adjustment of prices. Investors also, probably,

adopt a passive investment strategy thereby failing to identify situations when prices deviate from their intrinsic values. The rejection of the random walk by some Ghanaian stocks could be due to limited trading time on this market. The Ghana Stock Exchange was trading only three times a week, on Mondays, Wednesdays and Fridays for the period under study. This could have resulted in price adjustment delays and thus partly explain why stock prices on this market deviate from the random walk. If new information arrived on a Tuesday, which was a nontrading day on the Ghana Stock Exchange, investors would be in a position to reasonably predict the price movements for Wednesday when the stock market opens for trade.The Z-values of the runs test are, nevertheless, small in magnitude for some stocks, but relatively large for others in comparison to those obtained in, for example, Fama (1965) and Dickinson and Muragu (1994). The largest Z-values, which are significant at the 1 percent level, exceed 5.0 for a number of stocks across markets, reaching to as high as 9,546 for Ghana. In Fama (1965), the largest standardised value for the runs test, also in absolute terms, was 4.23. Dickinson and Muragu (1994) also found very small values in their study of the Nairobi Stock Exchange, with the largest Z-values in absolute terms of 2,987. While Fama (1965) and Dickinson and Muragu (1994) hinted that the serial correlation they found may not be attractive to investors, the same cannot be readily said for the African stock markets studied. One major concern would be the illiquidity of these markets. CONCLUSIONThe paper investigated the weak-form efficiency of ten African stock markets using the runs test methodology for serial dependency. Returns were calculated on a trade-to-trade basis and weighted by the number of days between trades. Serious thin-trading was observed on all markets, and more so for Namibia and Botswana, the two markets with significant dual-listed stocks on the JSE. In all the markets studied (except Namibia), a significant number of stocks rejected the random walk. The weak-form efficiency of the NSX was attributed to its correlation with the JSE. Kenya and Zimbabwe were also concluded as generally weak form efficient, since a significant number of stocks conformed to the random walk. All the stocks in the Mauritius sample rejected the random walk at the 1 percent level of significance using the runs test. This led to the conclusion that stock prices on the Mauritius market tend to deviate from the random walk hypothesis. The same conclusion was made for Ghana. On the BRVM, Egypt and Botswana stock exchanges, chances that one can detect patterns in the stock prices that can used to predict the next price also could not be ruled out. Since rejection of the random walk does not necessary imply weak-form inefficiency, but the presence of serial correlation in stock returns, it is vital to investigate if such serial correlation can be exploited for abnormal returns, net of transaction costs. Hence, there is need for further tests using technical trading methods that try to profitably exploit the serial correlation observed in this study, in order to reach definite conclusions on the efficiency or inefficiency of the African stock markets in this study.The runs test used here only tests for the existence of a linear relationship, which makes it inadequate as a testing method on African stock markets where the return-generating processes are assumed to be nonlinear. This is for the reason that the assumptions on which the EMH is based are believed to be violated due to the heterogeneity of investors and the weak microstructures of the markets. The use of linear models would thus lead to wrong inferences being drawn. Hence, further research is required to test the random walk hypothesis using nonlinear models and to investigate if the observed patterns can be profitably exploited.Source: http://mpra.ub.uni-muenchen.de/25968/Questions:

1. Which factors according to you violated the efficient market hypothesis in the above case study according to you?

2. Do you think that technical trading method will solve the problem which the African stock markets faced earlier?

Unit 4

Capital Structure and the Cost of Capital: Theory

INTRODUCTION

There always is a cost of initiating a business that ought to serve as your benchmark for how you devote in long-term assets. This cost is known as Cost of Capital. Cost of Capital is the price you pay to those who lend or spend money into your business. You can think of Cost of Capital as the rate of return investors need for incurring risk every time they give you money. Cost of Capital applies to long-term funding of assets as contrasting to short-term funding of working capital.Why is Cost of Capital so important? Well, you have to earn by and large rate of return on your assets that is higher than your cost of capital. If not, you end-up destructing value. So how do you calculate Cost of Capital? The popular approach is known as the Capital Asset Pricing Model or CAPM. CAPM estimates your cost of equity by taking a risk free rate and adjusting it by risks that are unique to your company or industry. Long-term government bonds are often used to calculate approximately risk free rates while overall market premiums run around 6 percent.CAPM is not ideal since it has numerous impractical assumptions and variations in estimates. For instance, sources (Bloomberg, S & P, etc.) for reporting market risks of specific companies offer very dissimilar estimates. In addition you might find straightforward estimates are just as accurate as CAPM. For instance, simply adding 3 percent to your cost of debt may provide a reasonably accurate estimate of your cost of capital. You can also see at companies that are very alike to your company. In any event, you necessitate calculating your cost of capital since it is an tremendously significant constituent in financial management decision making.Capital for investment is provided to the firm by investors who hold various types of claims on the firm’s cash flows. Debt holders have bonds that promise to pay them fixed amount of interest and principal in the future in exchange for their cash now. Equity holders provide retained earnings or buy rights offerings (internal equity provided by existing shareholders) or purchase new shares (external equity provided by new shareholders). They do so in return for claims on the residual earnings of the firm in the future. Also, shareholders retain control of the investment decision, whereas bondholders have no direct control except for various types of indenture provisions in the bond that may constrain the decision making of shareholders. In addition to these two basic categories of claimants, there are others such as holders of convertible debentures, leases, preferred stock, nonvoting stock, and warrants.Each investor category is confronted with a different type of risk, and therefore each requires a different expected rate of return in order to provide funds to the firm. The required rate of return is the opportunity cost to the investor of investing scarce resources elsewhere in opportunities with equivalent risk.Whether or not an optimal capital structure exists is one of the most important issues in corporate finance–and one of the most complex. This unit covers the effect of tax-deductible debt on the value of the firm, first in a world with only corporate taxes, then by adding personal taxes as well. Next the effect of business disruption and bankruptcy costs is introduced, and we extend the basic Modigliani-Miller model using the work of Leland. The result is an equilibrium theory of capital structure. The unit also covers nonequilibrium theories that include the pecking order theory, signaling, and the effect of forgoing profitable investments. There is also a discussion of the effect of risky debt, warrants, convertible bonds, and callable bonds.THE VALUE OF THE FIRM GIVEN CORPORATE TAXES ONLYAt the same time as tax consequences are a motivating factor in numerous corporate decisions, managerial actions designed exclusively to reduce corporate tax obligations are thought to be an ever

more significant feature of U.S. corporate activity. Do such activities advance shareholder interests? If prevention activities are costless to investors, the question is trivial as prevention activity outcome in a transfer of value from the state to shareholders. Undeniably, this has been the presumption in the large literature on the effects of taxes on financial decision-making. Corporate tax avoidance activity, nonetheless, may be costly on numerous margins. Aside from the direct costs of engaging in such activities, managers characteristically have to make sure that these actions are obscured from tax authorities. In the process, such machinations may afford managers enlarged latitude to pursue self-serving objectives. Can the latter result be noteworthy enough to change the simple answer that investors fully incarcerate the value of corporate tax avoidance activity?Two little sample studies show that the estimation of tax avoidance activities may not be conventional to the simple story of tax avoidance as a transfer of value to shareholders. First, corporate deportations - transactions where U.S. firms turn upside down their corporate structure so that a supplementary in a tax haven becomes the parent entity - provide important corporate tax savings with limited, if any, operational changes. Nevertheless, markets do not react in a strongly positive fashion – and often react negatively – to U.S. firms announcing such moves (e.g. Desai and Hines, 2002). Second, an event study of an episode of augmented tax enforcement in Russia shows that these enforcement actions are related with positive market reactions (Desai, Dyck and Zingales, 2007). These small sample studies are challenging but leave open questions about the nature of corporate tax avoidance activity in general and in bigger samples. Here we will examine the degree to which corporate tax avoidance activity is valued by investors in a large sample of US firms. While the conventional analysis of corporate tax avoidance suggests that shareholder value should augment with tax avoidance activity, an agency perspective on corporate tax avoidance gives a more nuanced prediction. Exclusively, firm governance should be a significant determinant of the valuation of purported corporate tax savings. While tax avoidance per se should augment the after-tax value of the firm, this consequence is potentially offset, predominantly in poorly-governed firms, by the enlarged opportunities for rent diversion provided by tax shelters. Hence, the net effect on firm value should be greater for firms with stronger governance institutions.The associated advantages of these two views of tax avoidance are evaluated using a dataset with 4,492 observations on 862 firms over the period 1993-2001. This panel is drawn from the Compustat and Execucomp databases, combined with data on institutional possession of firms from the CDA/Spectrum database. Firm value is measured using Tobin’s q, and governance quality is proxied for by the level of institutional possession, reflecting the ability of institutional owners to keep an eye on managerial performance more forcefully. Tax avoidance is considered by inferring the differentiation between income reported to capital markets and tax authorities – the book-tax gap – and controlling for accruals and other measures of earnings management. The analysis shows that, for a given firm, this measure takes on higher values in years when the firm is implicated in litigation relating to aggressive tax sheltering activity than in other years. OLS results show that, controlling for a range of other relevant factors together with firm and year fixed effects, the effect of the tax avoidance measure on q is positive, but not appreciably distinct from zero. As anticipated predicted by the agency perspective on corporate tax prevention, the effect is positive for those firm-years with high levels of institutional ownership. The explanation of these results, nevertheless, is convoluted by the possibility of measurement error in the proxy for tax avoidance and by the potential endogeneity of tax avoidance commotion. In particular, it is probable that firms that are performing worse for exogenous reasons may be more likely to engage in tax avoidance. Luckily, a 1997 regulatory change inadvertently and considerably changed the costs of tax sheltering differentially across firms. This source of exogenous variation permits the execution of an instrumental variables strategy that can be used to address these concerns and to investigate the causal effect of tax avoidance on firm value.

The “check-the-box” set of laws were planned to allow small firms to choose their executive form for tax purposes. Altshuler and Grubert (2005) and different practitioners have observed that these regulations also had the inadvertent outcome of lowering the costs of tax avoidance for firms. Particularly, “hybrid entities” became more and more common. These entities are classified as independently incorporated subsidiaries under the tax rules of one country while at the same time being treated as unincorporated branches under the tax rules of a new country. This flexibility in entity classification creates a sizable tax avoidance opportunity for firms with incentives to capitalize on these regulatory changes. The middle idea underlying the identification strategy is that, for a given enticement to fit into place in tax evasion, a firm will engage in more actual tax evasion after the “check-the-box” regulations were adopted. A crucial determinant of the incentives to engage in tax avoidance is the availability of “tax shields.” Thus, instruments for tax avoidance are constructed by interacting a dummy variable for the period after the “check-the-box” regulations with variables (at the firm-year-level) that proxy for the availability of tax shields, namely NOL carryforwards and two different measures of debt. The estimates using the instruments described beyond lead to results that are in the same direction as the OLS results, but are significantly stronger. The communication between institutional ownership and tax evasion is positive and important, as anticipated by the agency perspective on tax avoidance. This outcome is vigorous to the inclusion of different additional control variables, and to a mixture of extensions to the model. The exclusion restriction underlying the results may be unacceptable if the effect on firm value of the tax shield variables changed over time for reasons not related to the “check-the-box” regulations. Reassuringly, the basic result is robust to together with interactions between these variables and time trends in the model. By and large, the considerably better effects found using the above approach suggests that the OLS results are considerably exaggerated by attenuation bias due to measurement error in the tax avoidance proxy or by the endogeneity of tax avoidance.Theories of Corporate Tax AvoidanceThe ostensible growth in corporate tax avoidance activity has given rise to two substitute perspectives on the motivations and possessions of this activity. Numerous studies investigate corporate tax avoidance as an addition of other tax-favored activity, such as the use of debt. In general, Graham and Tucker (2006) build a sample of firms involved in 44 corporate tax shelter cases over the period 1975-2000. By comparing these firms with a harmonized sample of firms not involved in such litigation, they recognize characteristics (such as size and profitability) that are absolutely related with the use of tax shelters, and argue that tax shelters serve as a replacement for interest deductions in determining capital structure. A substitute notional approach put stress on the interaction of these tax avoidance activities and the agency problems that are intrinsic in publicly held firms. As per this view, obfuscatory tax avoidance activities can create a shield for managerial opportunism and the diversion of rents. This perspective underlies numerous recent studies, together with Desai and Dharmapala (2006a) and Desai, Dyck and Zingales (2007), and forms part of an emerging paradigm that emphasizes the links between firms’ governance arrangements and their responses to taxes. In this view, corporate tax avoidance not only means different costs, but these costs may in fact outweigh the reimbursement to shareholders, given the opportunities for diversion that these vehicles offer. Desai and Dharmapala (2006b) discuss examples of the interaction between tax shelters and various forms of managerial opportunism, exemplifying that straightforward diversion and subtle forms of earnings treatment can be facilitated when managers assume tax avoidance activity.While the conventional view of corporate tax avoidance suggests that shareholder value should augment with tax avoidance activity, the alternative view gives a more nuanced prediction. Particularly, firm governance should be a significant determinant of the valuation of alleged corporate tax savings. While the direct effect of tax avoidance is to augment the after tax value of the firm, these effects are substantially offset, chiefly in poorly-governed firms, by the increased opportunities for managerial rent

diversion. Consequently, the net effect on firm value should be better for firms with stronger governance institutions. Measuring Firm Value, Governance, and Corporate Tax AvoidanceThe data used to test the proposition described above is drawn from three sources. Financial accounting data is drawn from Standard and Poor’s Compustat database, executive compensation data (and certain other control variables) from Standard and Poor’s Execucomp database, and data on institutional ownership of firms from the CDA/Spectrum database. Integrating these variables leads to a dataset with 4,492 explanation at the firm-year level, on 862 firms over the period 1993-2001. The variables are described in detail below; summary statistics are reported in Table 4.1.

Table 4.1: Summary StatisticsMean Standard

DeviationNumber of Observations

Tobin’s q (excluding deferred tax expenses) 2.3537 2.2885 4,492Tobin’s q (including deferred tax expense) 2.3123 2.0011 4,062Market Value (scaled) 1.7825 2.2881 4,470Book-tax Gap (scaled) –0.0074 0.1077 4,492Total Accruals (scaled) –0.0432 0.0787 4,492Ratio of Value of Stock Option Grants to Total Compensation for Top 5 Executives

0.4066 0.2542 4,492

Value of Stock Option Exercises by Top 5 Executives (scaled) 0.0061 0.0257 4,492Sales (millions TZSs) 3.5831 8.3074 4,492Volatility (Black-Scholes measure) 0.4021 0.1864 4,492Net Operating Loss (NOL) Carryforwards (scaled) 0.0351 0.2074 4,492Foreign Income of Loss (absolute value; scaled) 0.0327 0.0401 4,492R&D Expenditures (scaled) 0.0475 0.0666 4,492Long-term Debt (scaled) 0.1748 0.1537 4,492Debt in Current Liabilities (scaled) 0.0387 0.0608 4,492Deferred Tax Expense (scaled) 0.0256 0.0345 4,062Future Sales Growth (fraction) 0.1077 0.2984 4,328Capital Expenditures (scaled) 0.0674 0.0505 4,433Pretax Income (scaled) 0.0544 0.1363 4,492Governance Index, 1998 9.2650 2.8387 3,906Institutional Ownership (fraction) 0.5853 0.1768 4,492

Note: These variables are defined as in the text. “Scaled” variables are deflated by the contemporaneous book value of assets.In stressing the value entailments of corporate tax avoidance, this discussion builds on the extensive literature in corporate finance on the epitopes of firm value. Inside this literature, it has become standard since Demsetz and Lehn (1985) to use Tobin’s q to measure firm value. The definition of q used in Kaplan and Zingales (1997) and Gompers, Ishii and Metrick (2003) is employed in the analysis below, with one modification: deferred tax expense is not incorporated in the definition of q used in the basic results below, as current tax avoidance activity may result in changes to future tax liabilities and thus create a mechanical correlation between the dependent variable and the measure of tax avoidance. While q is the primary dependent variable used in the analysis, alternative measures of firm value lead to consistent results.In summation to drawing on financial statement data, the analysis below necessitates a measure of firm governance. The basic measure of governance used in testing the paper’s chief hypothesis is the fraction of the firm’s shares owned by institutional investors. This fraction (which is reported quarterly) is averaged over each firm-year, and is denoted by Iit є [0, 1] for firm i in year t. The basic motivation

underlying this proxy is that institutional investors have greater incentives and capacity to monitor managerial performance. Thus, the higher is Iit, the greater the degree of examination to which administrative actions are subjected, and the less significant are agency problems between managers and shareholders. In addition, a different measure – the index of antitakeover provisions constructed by Gompers, Ishii and Metrick (2003) – is used in sturdiness checks. While this captures a quite diverse aspect of governance than does Iit (namely, managerial entrenchment rather than the quality of monitoring), its use leads to highly reliable results.Given the efforts undertaken to unclear such activities, tax avoidance is complicated to measure. The analysis adopts an roundabout approach, constructing a measure of corporate tax avoidance that takes as its starting point the gap between financial and taxable income. The difference between income reported to capital markets and to the tax authorities – the so-called book-tax gap – has attracted significant interest in recent years, and has been related to measures of corporate tax avoidance. Given that tax returns are confidential, income reported to tax authorities cannot be observed directly and must be inferred using financial accounting data, as described in Manzon and Plesko (2002) and implemented in Desai and Dharmapala (2006a). This approach uses firms’ reported current Federal tax expense and “grosses up” this tax liability by the US Federal corporate tax rate. For firms with positive current Federal tax expense, the graduated structure of corporate tax rates is used in this calculation. For firms with negative current Federal tax expense, the top statutory rate of 35 percent is used.Given this indirect value of the firm’s taxable income, the book-tax gap can be anticipated by simply subtracting inferred taxable income from the firm’s reported pretax (domestic US) financial income.6 To control for differences in firm scale, and for the reason that the dependent variable is deflated by the book value of assets, the inferred book-tax gap is also scaled by the book value of assets. This yields the measure of the book-tax gap used in the analysis below (denoted BTit for firm i in year t).The book-tax gap does not essentially reflect corporate tax avoidance activity, so any measure of tax avoidance must control for other factors. Specifically, the overreporting of financial income (known in the accounting literature as “earnings management”) may contribute to the measured book-tax gap. Studies of earnings management (e.g. Healy, 1985) have argued that such exploitation is most likely to take place through the exercise of managerial discretion in determining accounting accruals (i.e. adjustments to realized cash flows that are used in calculating the firm’s net income). The basic intuition underlying the measure of tax avoidance used here is that book-tax gaps are attributable either to earnings management or to tax avoidance activity. Accordingly, adjusting for earnings management with an accruals proxy isolates the component of the gap that is due to tax avoidance. In the regressions reported below, BTit is used as a proxy for tax avoidance activity, while earnings management is controlled for by including a measure of total accruals (denoted TAit for firm i in year t) as a control variable.Given the privacy of tax returns, the procedure outlined above yields the best measure of corporate tax prevention that can be obtained using publicly-accessible data. Nevertheless, in view of the limitations associated with inferring taxable income, and as there are alternative explanations for book-tax gaps, it is important to put into practice a validation check of the book-tax gap as a measure of corporate tax sheltering activity before proceeding to determine its valuation effects.Graham and Tucker (2006) build a sample of firms involved in 44 cases of tax shelter litigation over the period 1975-2000, using publicly-available court records and press articles. The validation check undertaken here uses a dataset compiled using a similar methodology. This information can be used to construct a variable that indicates whether tax sheltering activity was alleged in any given firm-year. Specifically, let the indicator variable Lit be equal to one if firm i was alleged to have used a tax shelter in year t, and zero otherwise. This variable is merged with data on the book-tax gap and a set of control variables from the merged Compustat-Execucomp dataset, in order to examine the relationship between involvement in tax shelter litigation and book-tax gaps. The regression specification used is:

(1)where μi and εt are firm and year fixed effects, respectively, and νit is the error term. Xit is a vector of control variables that includes measures of firm size (assets, sales and market value) and the structure of executive compensation.The resulting sample is very small – there are only 14 firms that were involved in tax shelter litigation at some point in the sample period, and for which all the required data is available. Nonetheless, estimating Equation (1) using this sample results in a positive coefficient on Lit, as reported in column 1 of Table 2; i.e. the book-tax gap for a given firm tends to be larger in years when that firm is allegedly using tax shelters, relative to the book-tax gap for the same firm in other years. This result is driven entirely by within-firm variation in Lit, controlling for time-specific changes in sheltering activity. Unsurprisingly, this result is of borderline statistical significance, given the small sample size. Column 2 reports the same specification using all available observations in the merged Compustat-Execucomp dataset; the estimate of β1 is very similar in magnitude to that in Column 1.10

Table 4.2: Book-Tax Gaps and Tax Shelter LitigationDependent Variable Book-Tax Gap

(1)Firms Involved in Tax Shelter

Litigation

(2)All Firms

Indicator (=1) for Alleged Tax Shelter Activity 0.0222* 0.0190(0.0124) (0.0125)

Accruals Measure (Scaled) 0.0440** 0.3646**(0.0200) (0.0574)

Controls for Changes in Sales, Assets, Market Value, and the Ratio of the Value of Stock Option Grants to Total Compensation for Top 5 Executives?

Y Y

Year and Firm Effects? Y YNo. of Firms 14 1,600No. of Obs. 80 6,799R-Squared 0.5258 0.5005

Note: The dependent variable is the book-tax gap, as defined in Section III. The indicator variable of interest = 1, the sample is restricted to those firms in the tax shelter litigation dataset. In Column 1, the sample is restricted to those firms in the tax shelter litigation dataset. In column 2, the sample includes all firms in the merged Compustat and Execucomp databases for which the required data is available. IN each case, the sample period is 1992-2001. The specifications include year effects, firms fixed effects and the set of controls specified. Robust standard errors that are clustered at the firm level are presented in parantheses; *,** and***denotes significance at the 10 percent, 5 percent and 1 percent levels, respectively. Any conclusions from this validation check are necessarily tentative, given the small number of firms that have been involved in tax shelter litigation. Nonetheless, it appears that the measure of tax avoidance employed below captures a critical element of tax sheltering activity, as it takes on higher values for those firm-years for which there is some independent evidence for alleged tax shelter activity.OLS Approach and ResultsWhile the central hypothesis of the paper concerns the interaction of government institutions and tax avoidance activity, the question of whether tax avoidance tends to be associated with increases or decreases in firm value is also of considerable interest. This is addressed using the following specification:

(2)

where the variables BTit and TAit are as defined above, μi and εt are firm and year fixed effects, respectively, and νit is the error term (note all regressions reported here use both firm and year fixed effects).Xit is a vector consisting of the following control variables. Changes in firms size over time are controlled for using sales. The value of stock option grants to executives as a fraction of total compensation is included for the reason that a substantial literature finds stock-based compensation to be a determinant of firm value, presumably through incentive-alignment effects. In addition, the structure of executive compensation plays a central role in Desai and Dharmapala. To control for changes over time in the risk associated with a firm’s stock price a measure of volatility is also included. As net operating loss (NOL) carryforwards are not taken into account in the measure of tax avoidance (and for the reason that NOLs can affect the incentives to engage in tax avoidance), NOL carryforwards scaled by assets (with missing values treated as zeroes) are also included.If a tax avoidance measure is restricted to domestic Tanzanian tax expense and Tanzanian Federal taxes, but liabilities and the incentives for tax avoidance may be influenced by foreign activity under the Tanzanian system of worldwide taxation. Thus, a proxy for foreign activity – the absolute value of foreign income or loss is included in Xit. As tax shields can affect the value of engaging in tax avoidance, changes in firms’ leverage are controlled for by including measures long-term debt and debt in current liabilities. Changes in intangibles that affect q but are imperfectly measured in the book value of assets are proxied for by research and development (R&D) expenditures. A number of additional control variables are used in robustness checks, as described below. Note also that for the reason that firm fixed effects are employed in the specification described below, many of the sources of cross-sectional variations in q across firms that have been discussed in the literature are effectively controlled for here.The specification used to test whether the valuation of corporate tax avoidance is dependent on firm governance extends Equation (2) as follows:

(3)where Iit is the measure of institutional ownership defined above. The hypothesis is section 2 implies that β4>0: i.e., the effect of tax avoidance on q is greater in firm-years in which institutional ownership id higher (and governance is stronger).The results using OLS estimation on Equations (2) and (3) are reported in Table 4.3; note that all results reported in this unit use robust standard errors that are clustered at the firm level. Column 1 presents the results from the estimation of Equation (2). The overall effect on firm value of the proxy avoidance is positive, but insignificant. The test of the hypothesis using Equation (3) is reported in Column 2. Here, the coefficient on the interaction term (Iit*BTit)–β4 in Equation (3)–is positive, consistent with the hypothesis, and is of borderline statistical significance. The intuition can be reinforced by running (2) separately for firm-years with high and low levels of institutional ownership (Columns 3 and 4, respectively), where “high” institutional ownership is defined as being a fraction that exceeds 0.6, which is approximately the mean of the sample. For well-governed firm-years, the effect of tax avoidance on q is positive and of borderline significance. For less well-governed firm-years (with institutional ownership below 0.6), the effect is also positive, but statistically insignificant, and considerably smaller in magnitude. Thus, while the estimated overall effect of tax avoidance on firm value is indistinguishable from zero, the effect appears to be more positive for well-governed firm-years than for poorly-governed firm-years. This finding is consistent with the hypothesis that agency problems mitigate the benefits to shareholders of corporate tax avoidance.

Table 4.3: Tax Avoidance, Firm Value and Government Institutions: OLSDependent Variable Tobin’s q

(1)All Firms

(2)All Firms

(3)Firm-Years

(4)Firm-Years

with High Institutional Ownership

with Low Institutional Ownership

Book-Tax Gap (Scaled) 0.5776 –2.1655 2.7624* 0.2718(0.5590) (1.4957) (1.5394) (0.3678)

Book-Tax Gap Interacted with Institutional Ownership

5.6687*

(3.3307)Institutional Ownership 0.7706**

(0.3287)Total Accruals (Scaled) 1.3267**

(0.3811)1.2689**(0.3613)

0.5313(0.5676)

1.1159*(0.5892)

Ratio of Value of Stock Option Grants to Total Compensation for Top 5 Executives

0.4391**(0.1208)

0.4371**(0.1202)

0.5627**(0.2004)

0.2253(0.1745)

Sales 0.0442*(0.0249)

0.0471*(0.0250)

0.0195(0.0158)

0.0592(0.0404)

Volatility –2.114006**(0.6682)

–1.9465**(0.6466)

–3.8771**(1.2553)

–1.0303*(0.5697)

Controls for Tax Shields (NOLs, Long-term Debt, and Current Debt)

Y Y Y Y

Controls for Foreign Income/Loss and R&D

Y Y Y Y

Year and Firm Effects? Y Y Y YNo. of Firms 862 862 583 614No. of Obs. 4,492 4,492 2,324 2,168R-Squared 0.6483 0.6500 0.7765 0.6213

Note: The dependent variable is Tobin’s q, as defined in Section III. The sample (over the period 1993-2001) is drawn from the merged Comustat and Execucomp databases, and is restricted to those firm-years for which CDA/Spectrum data on institutional ownership is available. All specifications include year effects, firm fixed effects and the controls listed. In column 3, the sample is restricted to firm-years with institutional ownership>0.6. In column 4, the sample is restricted to firm-years with institutional ownership ≤0.3. Robust standard errors that are clustered at the firm level are presented in parentheses; *, ** and *** denote significance at the 10 percent, 5 percent and 1 percent levels, respectively.Instrumental-Variables Approach and ResultsIV ApproachOLS estimation of Equations (2) and (3) gives rise to two types of potential problems. The first is measurement error in the proxy for tax avoidance, particularly if the extent of measurement error differs by governance institutions. For example, if the proxies used for earnings management are incomplete, then the remaining component of the book-tax gap may be mischaracterized as tax avoidance when it actually represents earnings management. Accordingly, it is possible that the results are driven by differential market reactions to earnings management by well-governed and poorly-governed firms. The second is the potential endogeneity of tax avoidance activity. For example, firms that are performing worse for other reasons may be more likely to engage in tax avoidance.In order to address these concerns, an exogenous source of variation in firms’ opportunities for tax avoidance is required. Fortunately, a 1997 regulatory change with unrelated objectives lowered the costs of tax avoidance for a subset of firms. In late 1996, the Treasury issued what are known as the “check-the-box” (CTB) regulations. These regulations enable firms to choose their organizational form for tax purposes – for example, whether to be taxed as a Corporation or as a pass-through entity such as a partnership or sole proprietorship – by filing a one-page form on which they could simply check the

appropriate box. In replacing a complex set of rules by which the IRS determined firms’ tax status, the CTB regulations were intended to reduce the administrative burdens faced by small firms. Researchers studying international taxation argue that the CTB regulations also had the unintended consequence of facilitating tax avoidance by large US-based multinational firms through the use of what are known as “hybrid entities” (see in particular Altshuler and Grubert (2005)). Hybrid entities are classified in two distinct ways – as separately incorporated subsidiaries under the tax rules of one country and as unincorporated branches under the tax rules of another country.The instruments for tax avoidance involve interacting a dummy variable for the post-CTB time period (i.e. the years since 1997) with firm-year-level variables that capture the incentive to engage in tax avoidance. The central idea underlying the identification strategy is that, for a given incentive to engage in tax avoidance, a firm will engage in more actual tax avoidance after the CTB regulations were adopted than it would have before, other things equal. A crucial determinant of the incentives to engage in tax avoidance is the availability of “tax shields” (i.e. tax deductions from other sources, such as interest deductions or NOL carryforwards resulting from losses in previous years); for instance, Graham and Tucker (2006) emphasize the substitutability of tax shelters and other kinds of tax shields. Instruments for tax avoidance can thus be constructed by interacting a dummy variable for the period after the CTB regulations with each of the following variables: NOL carryforwards and two different measures of debt (long-term debt and debt in current liabilities).

Table 4.4: Book-Tax Gaps, Tax Shields, and the “Check-the -Box” Regulations: First-Stage IV ResultsDependent Variable Book-Tax Gap

(1) (2)PostCTB*(NOL Carryforwards) –0.0349**

(0.0139)–0.0162(0.0356)

PostCTB*(Long-Term Debt) –0.0127(0.0160)

–0.0474(0.0384)

PostCTB*(Current Debt) –0.0879*(0.0507)

–0.3103**(0.1197)

PostCTB*(NOL Carryforwards)*(Institutional Ownership)

–0.0579(0.1197)

PostCTB*(Long-Term Debt)*(Institutional Ownership) 0.0548(0.0570)

PostCTB*(Current Debt)*(Institutional Ownership) 0.4149**(0.1644)

F-statistic for Joint Significance of the Instruments (P-value)

3.33*(0.0189)

3.00***(0.00063)

Controls for Total Accruals, Executive Compensation, Sales, Volatility, Foreign Income/Loss, and R&D

Y Y

Controls for Tax Shields (NOLs, Long-term Debt, and Current Debt)

Y Y

Control for Institutional Ownership N YYear and Firm Effects? Y YNo. of Firms 862 862No. of Obs. 4,492 4,492R-Squared 0.5993 0.6009

Note: The dependent variable is the book-tax gap, as defined in the text. “PostCTB” is an indicator variable for years after the “Check-the-box” regulations were introduced (1997-2001). The sample (over the period 1993-2001) is drawn from the merged Compustat and Execucomp databases, and is restricted to those firm-years for which CDA/Spectrum data on institutional ownership is available. All specifications include year effects, firm fixed effects and the controls listed. Robust standard errors

that are clustered at the firm level are presented in parentheses; *, ** and *** denote significance at the 10 percent, 5 percent and 1 percent levels, respectively.The IV approach involves instrumenting for the endogenous variable BTit in Equation (2) using as the set of instruments the variables listed above, each interacted with a dummy variable for the post-CTB period. Let Pt be the dummy for the post-CTB period, NOLit be the NOL carryforwards for firm i in year t, DLit be long-term debt for firm i in year t, and DCit be debt in current liabilities for firm i in year t. Then, the instruments for BTit are (Pt*NOLit), (Pt*DLit), and (Pt*DCit). The first-stage regression (reported in Column 1 of Table 4) shows that these instruments are indeed strongly related to BTit. Specifically, the coefficients are negative, as expected; i.e. lower values of tax shields (which imply a greater incentive to engage in tax avoidance) are associated with larger values of BTit after the CTB regulations than before, controlling for other factors. The instruments are jointly significant at the 5 percent level. In Equation (3), there are effectively two endogenous variables – BTit and (Iit*BTit) – and the set of instruments thus includes interactions with Iit. In particular, the instruments for BTit and (Iit*BTit) are the following: (Pt*NOLit), (Pt*DLit), (Pt*DCit), (Iit*Pt*NOLit), (Iit*Pt*DLit), and (Iit*Pt*DCit). The first-stage results (presented in Column 2 of Table 4) show the expected negative relationship between each of the first three of these instruments and BTit; the full set of instruments is also strongly jointly significant.The basic rationale for this IV approach is that a given incentive to engage in tax avoidance should lead to more actual tax avoidance after the CTB regulations than before. The crucial exclusion restriction underlying the use of these instruments is the following. The underlying tax shield variables (NOLs and the debt measures) used in constructing the instruments may directly affect firm value; this direct effect is controlled for by including the tax shield variables in the specification. Nevertheless, the tax shield variables should not affect firm value differently after the CTB regulations, other than through their influence on incentives for tax avoidance. This restriction is conditional on the controls included in the model: for example, even if firm valuations were in general higher in the late 1990’s, the year dummies included in the specification would account for this. Of course, the validity of the exclusion restriction depends on there being no other changes over time in the effect of the tax shield variables on firm value. To test for possible violations of this exclusion restriction, interactions between the tax shield variables and time trends are included in the model as a robustness check.IV ResultsThe second-stage results from the IV analysis are presented in Table 5. Column 1 reports the results from estimating Equation (2), using the instruments for BTit described above. The overall estimated effect of tax avoidance on firm value is substantially larger than in the OLS results in Table 3. Column 2 reports the results from estimating Equation (3), using the instruments for BTit and (Iit*BTit) described above. The coefficient β4 of the interaction between governance and tax avoidance is positive and highly significant, consistent with the paper’s main hypothesis. The IV results thus support the notion that the benefits to shareholders from corporate tax avoidance depend on firms’ governance institutions.

Table 4.5: Tax Avoidance, Firm Value and Government Institutions: Second-Stage IV ResultsDependent Variable Tobin’s q Tobin’s q Market

Value (Scaled)

Tobin’s q

(1) (2) (3) (4)Book-Tax Gap (Scaled) 14.523

(12.288)–5.8710(5.1474)

–6.9313(4.9001)

–4.2465(4.0640)

Book-Tax Gap Interacted with Institutional Ownership

32.8204**(13.0267)

31.4461**(12.8540)

21.4885**(10.8603)

Institutional Ownership 1.0331**(0.4376)

1.0593**(0.4250)

0.9614**(0.3976)

Total Accruals (Scaled) –2.8305(3.6467)

–1.3588(2.3935)

–0.4447(2.3651)

–0.2585(1.8085)

Ratio of Value of Stock Option Grants to Total Compensation for Top 5 Executives

0.3495(0.3092)

0.4839**(0.2142)

0.5532**(0.1919)

0.4732**(0.1609)

Sales 0.0434(0.0276)

0.0473*(0.0264)

0.0581(0.0245)

0.0418(0.0255)

Volatility –1.0221(0.9795)

–1.2096(0.7703)

–1.2202(0.7869)

–1.5364**(0.7227)

Controls for Tax Shields (NOLs, Long-term Debt, and Current Debt)

Y Y Y Y

Controls for Foreign Income/Loss and R&D Y Y Y YInteractions between Tax Shield Variables and Time Trends

N N N N

Year and Firm Effects? Y Y Y YNo. of Firms 862 862 862 862No. of Obs. 4,492 4,492 4,470 4,492

Note: The dependent variable in Columns 1, 2 and 4 is Tobin's q , as defined in Section III. The dependent variable in Column 3 is market value (scaled by the book value of assets), as defined in Section V. The sample (over the period 1993-2001) is drawn from the merged Compustat and Execucomp databases, and is restricted to those firm-years for which CDA/Spectrum data on institutional ownership is available. All specifications include year effects, firm fixed effects and the controls listed. The book-tax gap variable and the interaction between the book-tax gap and institutional ownership are instrumented, as described in the text. Robust standard errors that are clustered at the firm level are presented in parentheses; *, ** and *** denote significance at the 10 percent, 5 percent and 1 percent levels, respectively.The results in Table 4.5 are in the same direction as the OLS results in Table 3, but are considerably stronger. The coefficients from Tables 4.3 and 4.5 can be interpreted as reflecting an expected duration of a particular tax shelter or the ability to engage in tax planning for a given period. Suppose a firm unexpectedly increases its book-tax gap by $1. The current-year tax benefit (including federal and state tax benefits) would be approximately $0.40. Market reactions are given by the coefficient on the book tax gap and should incorporate an expectation of how long tax sheltering activity will continue in the future. Using reasonable discount rates, the estimated coefficient (2.76) for the well-governed sample in the OLS specification (column 3 of Table 3) would correspond to an expected life of tax savings of seven to nine years for well-governed firms. Nevertheless, interpreting this OLS coefficient in this manner is intricate by the identification concerns discussed above and the marginal significance of the coefficient in Table 4.3.The IV estimate in column 3 of Table 4.5 can be used to overcome these complexity. Taking a firm with a mean value of institutional ownership, these coefficients entail that the market interprets an increase in the book-tax gap as a quasi-permanent change in tax obligations. As such, changes in the book-tax gap are not interpreted as momentary items associated with particular shelters but as signals of general tax planning ability. The bigger effects using the IV approach suggest that measurement error in the tax avoidance proxy may lead to attenuation bias in the OLS estimates. Alternatively, the OLS estimate of the effect of tax avoidance on firm value may be biased towards zero by the form of endogeneity noted above, where firms that are performing poorly for other reasons are more apt to engage in tax avoidance.Robustness of the IV ResultsThese IV results come out to be robust to concerns regarding the measurement of the booktax gap, governance, and firm value. For example, the basic result in Column 2 of Table 4.5 is robust to the (unreported) inclusion of additional variables – particularly, deferred tax expense, depreciation expense, investment tax credits, interest expense, pension expense, and a proxy for employees’ stock option exercises - that control for the potential mismeasurement of the booktax gap. It is also robust to adding

lagged tax prevention activity to the model; this does not change the effect of that period tax avoidance (interacted with Iit), and the effect of the lagged variable is small and insignificant. Thus, there is no evidence to suggest a substantial delayed market reaction to firms’ tax avoidance activity.The consequences are also robust to using alternative actions of governance. Specially, the findings are unaffected when Iit is replaced by the index of antitakeover provisions constructed by Gompers et al. (2003). This index represents a count of antitakeover provisions that apply to a firm (either through its corporate charter or state law). It takes on values up to 18, with lower values representing better governance. As the cardinal properties of this index are unclear, the robustness check involves constructing an indicator variable for better-governed firms by dividing the sample at the mean (with values of 9 or lower corresponding to “well-governed” firms). The interaction between this indicator variable for well-governed firms and BTit is very similar in magnitude and significance to that in Column 2 of Table 5. This suggests that the results are robust to alternative notions of governance, as the Gompers et al. (2003) index measures managerial entrenchment rather than the quality of monitoring. Moreover, this also indicates that the results are unaffected by the potential endogeneity of Iit, where institutional investors may choose to buy firms that are expected to increase in value; this is less applicable to the Gompers et al. index, as its values were predominantly determined in the 1980’s, and generally do not change during the sample period.While the baseline condition in Table 4.5 comprises an extensive set of controls, it is probable that unobserved changes in firms’ investment opportunities or anticipated future performance may affect q. To address these concerns, it is probable to comprise capital expenditures and future revenue growth as additional controls; together with these controls leads to consistent results. In addition, although Tobin’s q is a standard measure of firm value in the literature, it is however significant to consider alternative proxies. As q takes account of the book as well as market value of equity and the value of debt, a simpler measure is the market value of common stock (Execucomp variable MKTVAL, the closing share price for the fiscal year multiplied by the number of common shares outstanding). This is scaled by the book value of assets in order to be conventional to the scaling of the independent variables. As exposed in Column 3 of Table 5, using this variable instead of q leads to basically identical results.In conclusion, the identification strategy used above depends on the validity of the exclusion restrictions. In particular, it necessitates that there are no changes over time in the effect on firm value of the tax shield variables that are used in constructing the instruments (other than the change due to the impact of the CTB regulations on tax avoidance activity). This postulation may be violated if there are trends unrelated to the CTB regulations in the effect of the tax shield variables on q. It is possible to test for this possibility by adding to the model interactions between a time trend and each of the tax shield variables. Particularly, these additional control variables are (NOLit*(t – 1997)), (DLit*(t – 1997)), and (DCit*(t – 1997)), where t is the year (1997 – which is the midpoint of the sample period – is used as the base year). The second-stage IV results with the addition of these controls are presented in Column 4 of Table 4.5. While the coefficient of the interaction term of interest is to some extent smaller, it remains significant at the 5 percent level. Thus, it does not appear that the effect of the instrumental variables is driven by time trends in the impact of the tax shield variables on firm value. Fairly, the results seem to depend only on the sporadic change in the effect of tax shields on firm value that is connected with the CTB regulations.

The simple belief that corporate tax avoidance symbolizes a transfer of value from the state to shareholders does not appear to be validated in the data. Rather, the patterns in the data are more regular with the agency perspective on corporate tax avoidance, which emphasizes the mediating role of governance. The basic result that higher quality firm governance leads to a larger effect of tax avoidance on firm value is toughened by using an exogenous source of variation due to changes in tax regulations to build instrumental variables for tax evading activity. The results are robust to a wide variety of tests

Security Market Line

βi = 1

RF

S(RM)

S(Ri)

for alternative explanations. Furthermore, the magnitude of the effect entails that changes in the book-tax gap are interpreted by the market as signals of overall tax preparation ability for well-governed firms, rather than simply reflecting the use of particular tax sheltering strategies.INTRODUCING RISK—A SYNTHESIS OF MODIGLIANI-MILLER MODEL AND CAPMGenerally, CAPM provides a natural theory to determinate the market price for risk and the suitable measure of risk. This model was developed more or less at the same time by Sharpe (1963-1964) and Treynor (1961), and Mossin (1966), Lintner (1965-1969) and Black (1972) developed it further. This model show that the equilibrium rates of return on all risky assets are function of their covariance with the market portfolio. But CAPM has a lot of implications for some corporate policy decision. So, the cost of equity capital for a firm is given directly by CAPM. After all, the company's beta is measured by calculating the covariance between the return on its common stock and the market index. Consequently, the beta measures the systematic risk of the common stock, and if we know the systematic risk, we can use the CAPM to determine the require rate of return on equity. Equation (4) is known as the capital asset pricing model, CAPM:

(4)where:S(Ri) = the expected rate of return on asset i;RF = the risk-free rate (constant);S(RM) = the expected rate of return on the market portfolio.It is shown graphically in fig.1, where it is also called the security market line. All securities fall exactly on the security market line. The required rate of return on any asset, S(R i) in eq (4), is equal to the risk-free rate of return plus a risk premium. The risk premium is the price of risk multiplied by the quantity of risk. In the terminology of the CAPM, the price of risk is the slope of the line, the difference between the expected rate of return on the market portfolio and the risk-free rate of return. The quantity of risk is often called beta, βi.

(5)

Figure 4.1: The Capital Asset Pricing ModelIt is the covariance between returns on the risky asset, i, and market portfolio, M, divided by the variance of the market portfolio. The risk-free asset has a beta of zero for the reason that its covariance with the market portfolio is zero. The market portfolio has a beta of one for the reason that the covariance of the market portfolio with itself is identical to variance of the market portfolio:

(6)

The Slope = (7)is definition for the price of risk.If it is possible to estimate the systematic risk of a company's equity as well as the market rate of return, then S(Ri) is the required rate of return on equity, i.e., the cost of equity for the firm. If we designate the cost of equity as Rc, then:

(8)When we combine CAPM with the cost of capital definitions derivated by Modigliani and Miller Theory (1958-1963), it provides a unified approach to the cost of capital.Figure 4.2 illustrates the difference between the original Modigliani-Miller cost of capital and the CAPM.

Figure 4.2: The security market lineMM assumed that all projects within the firm had the same business or operating risk. Consequently, the WACC in the MM theory, is represented by the horizontal line in fig. 4.2. The WACC for the firm does not change as function of systematic risk. This assumption, of course must be modified for the reason that firms and projects differ in risk.Table 6 shows expressions for the cost of debt before taxes, r, unlevered equity, Rc(N), levered equity, Rc(I) , and weighted average cost of capital, WACC, in both the MM and CAPM frameworks.

Table 4.6: Comparison of MM and CAPM Cost of Capital EquationsType of cost of capital CAPM Definition MM DefinitionInterest rate r= RF+βd(RM–RF) r=RF; βd = 0The cost of equity for the unlevered firm

Rc(N)=RF + β(N) (RM–RF) Rc(N)= Rc(N)

The cost of equity for the levered firm

Rc(I) = RF + β(I) (RM–RF) Rc(I) = Rc(N) + (1–T) (Rc(N)–r).DC

WACCWACC(I)=Rc(I)+r(1–T).

CC+D

WACC(I) = Rc(N).[1–T.D

C+D]

Now we will demonstrate, that the traditional and MM definition of WACC (the last line in table 1) are identical.By definition:

(9)Substitute into the right hand side the CAPM definition of the cost of levered equity, Rc(I), from (10):

(10)We have:

(11)Modigliani-Miller assumed, for convenience, that corporate debt is risk free; i.e., that its price is sensitive to changes in interest rates and either that it has no default risk or that default risk is completely diversifiable (βd = 0).The Modigliani-Miller definition of the cost of equity for the unlevered firm was tautological, i.e., R c(N) = Rc(N), for the reason that the concept of systematic risk had not yet been developed in 1958. We know that it depends on the systematic risk of the firm’s after-tax operating cash flows, β (N). Unfortunately for empirical work, the unlevered beta, β(N), is not directly observable. We can, nevertheless, easily estimate the levered equity beta, β(I). To derive the relationship between the levered and unlevered betas, begin by equating the Modigliani-Miller and CAPM definitions of the cost of levered equity (line 3 in table 6).

(12)Next, use the simplifying assumptions that r=RF, to write:

(13)Then substitute into the right-hand side the CAPM definition of the cost of unlevered equity, RC(N),

(14)By cancelling terms and rearranging the equation, we have:

(15)

(16)

(17)

The implication of eq. (17) is that if we can observe the levered beta by using observed rates of return on equity capital in the stock market, we can estimate the unlevered risk of the firm’s operating cash flows.To obtain the graph in which is shown changes in the cost of capital in function of leverage increase, begin by eq. (11)

(18)

For ∞we will obtain

Fig. 4.3. Changes in the cost of capital as leverage increasesNote that the cost of equity increases with increasing leverage. This simply reflects the fact that shareholders face more risk with higher financial leverage and that they require a higher return to compensate them for it. A more difficult problem is to decide what to do if the project's risk is different from that of the firm. In this case, the cost of capital must be risk-adjusted for capital budgeting purposes. Each project must be evaluated at a cost of capital that reflects the systematic risk of its operating cash flows as well as the financial leverage appropriate for the project. Estimates of correct opportunity cost of capital are derivated from a through understanding of the Modigliani-Miller cost of capital and the CAPM.THE COST OF CAPITAL WITH RISKY DEBTIn most cases the debt is not free of risk. But, in simple discussions when we talk about the weighted average cost of capital (WACC), we take the debt for risk-free. At the same, in the calculation of the WACC, we may use a value for the cost of debt d that is higher than the risk-free rate rf. In assuming that the debt is risk free, there are several advantages.

First, when we assume that the debt is risk-free, we do not have to model the risk of default on the debt.

Second, with risk-free debt, the expected return on the debt is equal to the promised return and we do not have to distinguish between the expected return on the debt and the promised return.

Third, since the debt is risk free, the cost of debt is constant and not a function of the amount of debt. Thus, we do not have to model how the cost of debt d varies as a function of the amount of debt D.

Taxes raise additional complications. With risky debt, we have to use the expected rate of return on the debt rather than the promised rate of return on the debt in the formula for the WACC. Furthermore, we model the expected cost of debt risky as an increasing function of the amount of debt.Here we will examine the probability of risk in both single period binomial model and multi-period model.Single Period Binomial ModelWe start by reviewing the standard formula for the WACC. In words, the WACC is a weighted average of the market cost of debt and the return to equity, where the weights are the market values of the debt and equity, as percentages of the total value.

WACC = percentD*d + percentE*e (19)

Where d is the market cost of debt,e is the return to equity, percentD is the market value of debt as a percent of the total value and percentE is the market value of equity as a percent of the total value.

Single Period Binomial ModelWe briefly review the single period binomial model. Consider a simple one period binomial model with two states of nature. Let FCF(i,j) be the value of the FCF in the jth state of the ith year. In the up state of nature, the value of the FCF is TZS 190 and the down state of nature, the value of the FCF is 50.

FCG(1, 1) = (20)FCF (1,2) = (21)

We assume that both states of nature are equally likely. Let P be the set of probabilities for two states of nature, where pU is the probability of the up state of nature and (1–pU) is the probability of the down state of nature.Let Ep{ECF(1, 1:2)} be the expectation of the FCF for the two states of nature in year 1, with respect to P. The expected FCF is TZS120.

Ep{FCF(1, 1:2)} = pU*FCF(1, 1) + (1 – pU)*FCF(1, 2)= 50 percent *190 + 50 percent *50 = 120.00

(22)We assume that the return to unlevered equity ρ is 20 percent . Thus, with respect to year 0, the (present) value of the FCF, discounted with ρ, is TZS100.

(23)Equivalent probabilities (risk-neutral probabilities)We can calculate the value of the FCF with respect to year 0 in another equivalent way. Rather than using the set o probabilities P, we define another set of equivalent probabilities Q = (q U, (1 –qU)). The set of equivalent probabilities Q (also known as risk-neutral probabilities), unlike the set of probabilities P,

are very useful for the reason that they are appropriate (and convenient) for taking the expectation of any cash flow structure.1

To find the value of ECF, we take the expectation of the ECF with respect to Q and discount with the risk-free rate rf. Assume that the risk-free rate is 12 percent . We verify that the appropriate value of q U is 44.29 percent .Let EQ{FCF(1, 1:2)} be the expectation of the FCF for the two states of nature in year 1, with the respect to Q. The expected FCF is TZS112.

(24)Thus, with respect to year 0, the (present) value of the FCF, discounted with rf, is TZS100.

(25)Debt FinancingNext, we introduce debt financing. Let X(1) be the total payment (principal interest) for the debt in year 1 and let D(0) be the value of the debt in year 0. The maximum amount that can be repaid in year 1, with no risk, is equal to the value of FCF(1,2). That is,

D(0)*(1 + rf) = FCF(1,2) (26)Solving for D(0), we obtain:

(27)The critical value of the debt is TZS 44.64. If the value of the debt at the end of year 0 is higher than $44.64, then there is a positive probability of default and the cost of debt will be higher than the risk-free rate of 12 percent .For example, suppose the project “promises” to pay X(1) to the debt holder at the end of year 1, where the value of X(1) is TZS60. At the end of year 0, what is the market value of debt? Let CFD(I, j) be the cash flow to debt (CFD) in the jth state of nature in the ith year. If the up state of nature occurs, there will be no problem in making the payment of TZS to the debt holder. Nevertheless, if the down state of nature occurs, then the debt holder simply receive TZS50, which is the value of the FCF, and the equity holder will receive nothing.

CFD(1, 1) = 60 (28)CFD(1, 2) = 50 (29)

To find the market value of the debt, we take the expectation of the CFD with the risk-neutral probabilities and discount with the risk-free rate.

1 There are two equivalent ways to calculate the value of an expected cash flow. First, we take the expectation of the cash flow with respect to a set of probabilities P and discount with the risk-adjusted discount rate. Second, we take the expectation of the cash flow with respect to Q, which is an equivalent set of probabilities, and discount with the risk-free rate. In terms of valuation, it is more convenient to take expectation with respect to the set of risk-neutral probabilities Q rather than the set of probabilities P.Since the risk-free rate is lower than the risk-adjusted rate, in the alternative method, the expectation of the cash flow has to be lowered accordingly by decreasing the probability for the up state of nature. Thus the value of q U, which is the equivalent probability for the up state of nature, is lower than the value of pU, which is the original probability for the up state of nature. Another way to think of the change in probabilities is as follows. Taking the expectation with respect to the equivalent set of probabilities Q, we “subtract” the risk premium from the cash flow to obtain the certainty equivalent. Consequently, we can discount the certainty equivalent with the risk-free discount rate.

Let EQ{CFD(1, 1:2)} be the expectation of the CFD for the two states of nature in year 1, with respect to Q. The expected CFD is TZS54.43.

EQ{CFD(1, 1:2)} = qU*CFD(1, 1) + (1–qU)*CFD(1, 2)= 44.29 percent*60 + (1 – 44.29 percent)*50 = 54.43(30)

Thus, with respect to year 0, the (present) value of the FCF, discounted with rf, is TZS48.60.

(31)The value of the debt is 48.60 percent of the total value of TZS100. Below we show and explain the calculation for the promised and expected rate of return to debt.Promised versus expected return on the debtSince the debt is risky, we distinguish the promised rate of return on the debt dProm from the expected rate of return on the debt dExp. In the calculation of the WACC we use the expected rate of return on the debt rather than the promised rate of return.Promised return to debtThe promised payment is X(1) and equal to TZS60. The promised rate of return is equal to 23.46 percent.

(32)In the down state of nature, the FCF of TZS50 is less than X(1) and the “promise” will not be fulfilled.Expected return to debtThe relationship for the expected rate of return to debt is as follows.

(33)The expectation of the CFD in year 1 with respect to Q and discounted with the risk-free rate is equal to the expectation of the CFD in year 1 with respect to P and discounted with the expected rate of return.Let EP{CFD(1, 1:2)} be the expected value of the CFD for the two states of nature in year 1, with respect to P. The expected CFD is TZS55.

EP{CFD(1, 1:2)} = pU*CFD(1,1) + (1 –pU)*CFD(1, 2)= 50 percent*60+50 percent*50 = 55.00

(34)Solving for the expected rate of return to debt in equation 13 and substituting the appropriate values, we obtain that the expected rate of return is 13.18 percent.

(35)It can be shown that the expression for the expected rate of return to risky debt is as follows.

(36)Rate of return to equityWe can show that in the single period case, if the debt is risky, the rate of return to equity e is constant and the formula for the return to equity is as follows.

= (37)Weighted average cost of capitalWe verify that the values for the cost of debt and return to equity are correct by showing that the WACC is equal to the return to unlevered equity ρ.

EACC = percent D*d+ percentE*e= 48.60 percent*13.18 percent + (1–48.60 percent)*26.44= 20.00 percent (38)

Expected return and promised return on risky debtThus, if the debt is risky, we can use equations 36 and 37 to estimate the cost of the debt as a function of the value of debt. For example, if the value of the debt in year 0 D(0) is TZS60, then the expected return on the debt is 15.71 percent

(39)To borrow TZS60 at the end of year 0, the project must promise to pay TZS88.84 at the end of year 1, with a promised return of 48 percent. We verify that these numbers are correct.

EQ{CFD(1,1:2)}=qU*(CFD)(1,1)+(1–qU)*CFD(1,2)= 44.29 percent*88.84 + (1–44.29 percent)*50 = 67.20

(40)Thus, with respect to year 0, the (present) value of the CFD, discounted with rf, is TZS60.

(41)The expected rate of return to debt versus the value of debt in year 0In the following graph, we plot the relationship between the expected rate of return to debt and equity versus the value of the debt in year 0. As expected, for values of debt less than or equal to TZS44.64, the expected rate of return to debt is equal to the risk-free rate of 12 percent. As the value of debt increases above TZS44.64, the expected rate of return to debt increases, at a decreasing rate.For values of debt less than TZS44.64, when the cost of debt is constant, the rate of return to equity increases, at an increasing rate. As the value of debt increases above TZS44.64, the rate of return to equity is constant.Graph 1: Relationship between the expected rates of return and the value of debt D(0)

MULTI-PERIOD BINOMIAL MODELIn this section, we present a four period binomial model and use one-way tables to show the relationship between the cost of debt and the amount of debt. With a binomial model, we can see clearly the impact of the passage of time on the expected return to debt and equity. For convenience we use a recombining tree to represent the (present) value of the FCF rather than the FCF process. Let VUn(i,j) be the (present) value of the FCF in the jth state of nature in the ith year. We assume that VUn(i,j) either increases or decreases by 20 percent.Graph 2: Process for the unlevered value VUn(i,j)

There are no free cash flows in years 1 to 3. The only cash flows occur in year 4.FCF(i,j) = 0 for all i ≤ 3 (42)

FCF(i,j) = VUn(i,j) for i = 4 (43)For example, under the third state of nature in year 4, the equity holder receives TZS92.16.You can easily verify that the correct values of the parameters for the above unlevered value process are as follows.

P = {pU, (1–pU)} = {70 percent,30 percent}(44)

Q = {qU, (1–qU)} = {62.5 percent, 37.5 percent}(45)

Unlevered return ρ = 8 percent (46)Risk-free return rf = 5 percent

(47)Debt Financing (1)Let X(4) be the payment to the debt holder in year 4. There are no cash flows to debt in years 1 to 3. If the value of X(4) is less than or equal to TZS40.96, which is the FCF under the fifth state of nature in year 4, then the debt is risk-free and the return to debt is equal to the risk-free rate. If the value of X(4) is higher than TZS40.96, then the debt is risky for the reason that there is a positive probability that the debt will not be fully repaid and the return to debt will be higher than the risk-free rate.Suppose the value of X(4) is TZS60, which is between the values of the FCF under the fourth and fifth states of nature. With respect to year 0, what is the value of the debt D(0)? To find the value of the debt, first we calculate the CFE and then the CFD.In year 4, under each state of nature, the cash flow to equity is as follows:

CFE(4,j) = max{[FCF4(4,j)–X(4)],0} (48)If the FCF is greater than the promised payment for debt, then the CFE is the difference between FCF and the payment. Otherwise, the equity holder receives zero. The cash flow to debt is simply the difference between the FCF and the CFE.Below, we show the process for the value of equity EL(i,j) and the value of debt D(i,j). Under the first four states of nature in year 4, the FCF is sufficient to pay the debt holder. In the fifth state of nature, the FCF is insufficient to pay the debt holder.Under the first four states of nature in year 4, the debt holder receives the promised amount X(4) and in the fifth state of nature, the debt holder receives the FCF of TZS40.96.Graph 3: Process for the value of (levered) equity EL(i,j) with X(40 = 60)

To find the value at any node in the value trees for the equity and debt, we take the expectation with respect to the set of risk-neutral probabilities and discount with the risk-free rate rf.

Graph 4: Process for the value of debt D(i,j) with X(4) = 60

Return to equity and return to debtNext we show the returns to equity and debt at each node of the value trees. The average (expected) return to equity increases from 10.76 percent in year 0 to 11.72 percent in year 3.Graph 5: Return to levered equity e(i,j) with X(4) = 60

The average (expected) return to debt decreases from 5.13 percent in year 0 to 5.08 percent in year 3.Graph 6: Return to debt D(i,j) with X(4) = 60

We verify that the WACC in year 0, calculated with the expected return to equity and debt, is equal to the return to unlevered equity ρ.

WACC(0) = percentD*d + percentE*e= 49.05 percent*5.13 percent + (1–49.05 percent)*10.76 percent= 8.00 percent

(49)Example:The expectation of the value of the debt at the subsequent two nodes in year 3 with respect to Q, discounted with the risk-free rate rf is equal to the expectation of the value of the debt at the subsequent two nodes with respect to P, discounted with d(2,3).

= 70 percent*57.14 + 30 percent*50.34 = 55.10

= 62.5 percent*57.14 + 37.5 percent*50.34 = 54.59Substituting the appropriate values, we obtain that the expected rate of return on the debt is 5.98 percent.

= Debt financing (2)Next we consider a higher level of debt financing. Suppose the value of X(4) is TZS90, which is between the values of the FCF under the third and fourth states of nature. With respect to year 0, what is the value of the debt D(0)? Again, we can construct the tree processes for the return to equity and the return to debt. Under the fourth state of nature in year 3, the return to equity is undefined for the reason that CFE is less for both the fourth and fifth states of nature in year 4.Graph 7: Return to levered equity e(i,j) with X(4) = 90

With the increase in X(4) from TZS60 to TSZ90, in year 0, the expected return to equity has increased from 10.76 percent to 13.18 percent and the expected return to debt has increased from 5.13 percent to 5.80 percent.Graph 8: Return to debt D(i,j) with X(4) = 90

With X(4) equal to TZS90, the reader can verify that the value of the debt in year 0 is TZS70.15.Debt financing (3)Next, we examine a very high level of debt and set X(4) equal to TZS150, which is between the values of the FCF under the first and second states of nature.Graph 9: Return to levered equity e(i,j) with X(4) = 150

With such a high level of debt, the CFE will be positive in the first state of nature in year 4 and zero in all the other states. Now the expected return to equity is constant at 17.60 percent.Graph 10: Return to debt D(i,j) with X(4)=150

For the debt holder, the debt will only be repaid in full under the first state of nature in year 4. Under all the other states of nature in year 4, the debt holder will receive less than the promised amount. The expected return to debt is 7.26 percent.The expected rate of return to debt and equity versus the value of debt in year 0In the following graph, we plot the relationship between the expected rate of return to debt and equity versus the value of the debt in year 0.Note that when the value of X(4) is between the values of the FCF under the first and second states of nature, as discussed previously, the expected return to equity is constant at 17.60 percent.Using the binomial model and a one-way table, we can estimate the expected return to equity and risky debt as a function of the value of debt in year zero2.Graph 11: Relationship between the expected rates of return and the value of debt D(0)

2 It is interesting that numerically we can use the formula for the expected cost of debt for the single period given in equation 16 to generate the expected cost of debt in the multi-period example. Nevertheless, we have not attempted to show rigorously that the formula in equation 36 applies in general.

Here we uses simple binomial models to illustrate the WACC with risky debt and no taxes. In the single period case, when the debt is risk-free, the cost of debt is constant and the return to equity is a function of the debt-equity ratio. When the debt is risky, the return to equity is constant and the expected return to debt is a function of the value of debt. In the multi-period case, we use one-tables to show the relationship between the rates of return to equity and debt as functions of the value of debt in year 0.THE MATURITY STRUCTURE OF DEBTFinancial economics has made significant progress in explaining the incentives that lead large public corporations to choose particular financing policies. Increasingly, the profession is moving beyond an examination of the basic leverage choice to more detailed aspects of the financing decision. In this discussion, we extend this literature by employing data from a large sample of firms to examine hypotheses about the determinants of the maturity structure of the firm’s debt. We group the hypotheses that have been offered to explain corporate debt maturity into three categories: contracting-cost hypotheses, signaling hypotheses, and tax hypotheses. Our evidence supports the contracting-cost hypotheses. Consistent with Myers (1977), firms with more growth options in their investment opportunity sets have less long-term debt in their capital structure. Large firms and regulated firms have more long-term debt. The evidence on signaling hypotheses is mixed. We find limited evidence that firms use debt maturity to signal information to the market. Nevertheless, our evidence is consistent with a pooling equilibrium in which firms with larger potential information asymmetries (measured by the amount of growth options in their investment opportunity sets) issue more short-term debt. Our evidence also suggests a nonmonotonic relation between credit standing and debt maturity as predicted by Diamond (1993). We find no evidence that taxes affect debt maturity.The Theory of Debt MaturityThree nonmutually exclusive sets of hypotheses have been proposed to explain corporate debt maturity: contracting-cost hypotheses, signaling hypotheses, and tax hypotheses. In this section, we

summarize these basic hypotheses to guide the structure of our empirical tests and the interpretation of our results.Contracting-Cost HypothesesInvestment Opportunity SetMyers (1977) argues that a corporation’s future investment opportunities are like options. The value of these options depends on the likelihood that the firm will exercise them optimally. With risky fixed claims in the firm’s capital structure, the benefits from undertaking profitable investment projects are capture enough of the benefits so that profitable project does not offer stockholders a normal return. In these cases, stockholders have incentives to reject positive net present projects. Myers calls this the underinvestment problem.With more growth options in the firm’s investment opportunity set, the conflict between stockholders and bondholders over the exercise of these options is greater. Myers argues that a firm can control this incentive problem in several ways: by including less debt in its capital structure, by including restrictive covenants in its indenture agreements, or by shortening the effective maturity of its debt. Specifically, Myers notes that if the debt matures before any opportunity to exercise the real investment options, this disincentive to invest is eliminated. Hence, firms with more growth options in their investment opportunity sets should employ shorter-maturity debt.If the firm could costlessly recapitalize itself, this disincentive to invest would be eliminates (Fama 1978). For example, consider a firm with a growth opportunity that requires an investment in one year. This firm has the same incentive to invest whether it has one-year debt. This firm has the same incentive to invest whether it has one-year debt outstanding or ten-year debt that it repays after one year. But the recapitalization strategy requires that the firm repurchase its outstanding long-term debt at prices that do not reflect the value of the new project. If information about the project causes the prices of its long-term debt to rise, then the benefits from the project are again shared between the stockholders and bondholders—whether the bonds are repurchased or not — and the underinvestment incentive remains. Issuing short-term debt avoids this problem. If fixes the price at which the firm repurchases its debt and allows stockholders to capture more (if not all) of the benefits from its new investments?Myers’ analysis thus provides a rationale for value-maximizing firms to match the effective maturities of their assets and liabilities. At the end of an asset’s life, the firm faces a reinvestment decision. Issuing debt that matures at this time helps to reestablish the appropriate investment incentives when new investment is required. More importantly, nevertheless, this analysis indicates that the maturity of a firm’s tangible assets is not the sole determinant of its debt maturity. The firm’s intangible assets (its growth options) play a critical role as well.Implicitly, Myers must also assume that the cost of rolling short-term debt is greater than the cost of issuing long-term debt. If there were no differential costs of short-term debt, then all firms would prefer short-term debt under the contracting-cost hypothesis. The higher costs of short-term debt potentially include: (1) higher out-of-pocket flotation costs, (2) greater opportunity costs of management time in dealing with more frequent debt issues, and (3) reinvestment risk and potential costs of illiquidity.Stulz and Johnson (1985) suggest that the firm can control the underinvestment problem by retaining the ability to issue fixed claims with high priority (such as secured debt). Financing new investment projects with high-priority claims limits wealth transfers from stockholders to existing bondholders and thus reduces the incentives for stockholders to forego these projects. Ho and Singer (1982) argue that even if short-term and long-term debt have the same priority in bankruptcy, short-term debt has a higher effective priority outside bankruptcy for the reason that it is paid first. Thus issuing short-term debt to finance new investment projects offers potential benefits that are similar to those from issuing secured debt for controlling the underinvestment problem.Finally, Smith and Warner (1979) hypothesize that risky firms benefit from placing more restrictive covenants in their debt. Covenants reduce moral- hazard problems (including the underinvestment

problem) that occur after the debt is issued. When a restrictive debt covenant is violated, it is often optimal to renegotiate the debt agreement rather than forcing the borrower into bankruptcy. The Trust Indenture Act of 1939, nevertheless, limits the discretion that may be allocated to the trustee in a public debt issue. It thus is costly to renegotiate covenants in public debt agreements outside the bankruptcy process. Private lenders, therefore, have a comparative advantage in writing debt contracts with restrictive covenants. Fama (1985) argues that banks have a comparative advantage over other private lenders in monitoring such loans. (Consistent with this hypothesis, James (1987) finds a positive and statistically significant price reaction associated with bank loan announcements, but insignificant price reactions for public and nonbank private debt announcements.) To maximize the effectiveness of these monitoring activities, most bank loans are short term. By reducing the term of the debt, the bank maintains a stronger bargaining position, and that can affect the firm’s investment policy. For example, consider a firm with outstanding public debt that would forego the exercise of a growth option. With short-term debt, the bank can make exercise of the option a condition for refunding. This suggests that high-growth firms that choose bank financing over public or nonbank private debt will have more short-term debt. RegulationSmith (1986) argues that managers of regulated firms have less discretion over future investment decisions than managers of unregulated firms. This reduction in managerial discretion reduces the adverse incentive effects of long-term debt. The contracting-cost hypothesis thus implies that regulated firms will have longer-maturity debt than unregulated firms.Firm SizeFirm size is potentially correlated with debt maturity for several reasons. Issuance costs for public issues have a large fixed component resulting in significant scale economies. Smaller firms are less able to take advantage of these scale economies; they typically opt for private debt with its lower fixed costs and consequently lower overall costs. Small firms that choose bank debt over public debt for the reason that of the lower flotation costs will have more short-term debt. Firms with foreign operations are more likely to issue foreign debt to manage their currency exposure. Many foreign debt markets are less liquid than U.S. debt markets, especially for longer maturities. Thus, multinational firms are likely to have more short-term debt. If larger firms are more likely to have foreign operations, this induces a negative relation between size and debt maturity.Signaling HypothesesFirm QualityFlannery (1986) and Kale and Noe (1990) examine signaling implications of the firm’s debt-maturity choice. The pricing of long-term debt is more sensitive to changes in firm value than the pricing of short-term debt. Thus, although mispricing of the firm results in both long-term and short-term debt being mispriced, the mispricing of the long-term debt is greater. If the bond market cannot distinguish between high-quality and low-quality firms, high-quality (undervalued) firms will want to issue the less underpriced short-term debt. Low-quality (overvalued) firms will want to issue the more overpriced long-term debt. Rational investors understand these incentives when valuing risky corporate debt. In equilibrium, nevertheless, with risky debt of both maturities issued, high-quality firms will issue more short-term debt and low-quality firms will issue more long-term debt.Pooling EquilibriaIn addition to a separating signaling equilibrium (in which the firm’s choice of debt maturity reveals its type) a pooling equilibrium is also possible. In a pooling equilibrium, both high and low quality firms issue debt of the same maturity. Flannery argues that firms with large potential information asymmetries (such as high-growth firms) are likely to issue short-term debt for the reason that of the larger information costs associated with long-term debt. Firms with smaller potential information

asymmetries will be less concerned about the signaling effects of their debt maturity choice, and are more likely to issue long-term debt.Credit RiskAs in other signaling models, Diamond (1991, 1993) assumes that firms with favorable private information about future profitability will prefer to issue short-term debt. Diamond argues, nevertheless, that short-term borrowing also exposes firms to the risk of excessive liquidations. Lenders are reluctant to refinance the debt if bad news arrives. In Diamond’s analysis, firms with the highest credit ratings issue short-term debt for the reason that this refinancing risk is small. Firms with lower credit ratings prefer long-term debt to reduce this refinancing risk. Firms with very poor credit ratings, nevertheless, are unable to borrow long-term for the reason that of the extreme adverse-selection costs. Thus, there are two types of short-term borrowers: those with very good credit ratings and those with very poor credit ratings; firms in between are more likely to issue long-term debt.Rajan (1992) argues that owners of firms with public debt outstanding will continue some projects that would be abandoned without the debt. He assumes that an optimal liquidation policy would be based on private information. Public debt contracts, therefore, cannot be contingent on such an optimal policy. If private lenders have access to better information about the value of the firm’s assets, nevertheless, sh0rt—term private debt contracts can induce more efficient liquidation policies.Tax HypothesesTerm StructureBrick and Ravid (1985) analyze tax implications of the debt maturity choice. For the reason that the firm can default on its promised debt payments, the expected value of the firm’s tax liabilities depends on the maturity structure of its debt whenever the term structure of interest rates is not flat. They assume that the probability of default increases with time, and the value of the firm’s interest tax shield is reduced upon default. (This occurs, for example, if in reorganization, the firm faces binding constraints on the use of tax-loss carry-backs.) If the yield curve is upward sloping, the expectations hypothesis implies that in early years the interest expense from issuing long-term debt is greater than the expected interest expense from rolling short-term debt. But the interest expense is less in later years. In this case, Brick and Ravid argue that issuing long-term debt reduces the firm’s expected tax liability and consequently increases the firm’s current market value. Conversely, if the term structure is downward sloping, issuing short-term debt increases firm value. Thus, the tax hypothesis implies that firms employ more long—term debt when the term structure has a positive slope. Lewis (1990) argues that taxes have no effect on optimal debt maturity. He notes that Brick and Ravid assume that the firm selects its leverage before the debt-maturity structure is chosen. If optimal leverage and debt-maturity structures are chosen simultaneously, then the debt-maturity structure is irrelevant.DataFor our empirical investigation of the determinants of corporate debt maturity, we construct a large sample of firms. Our data set merges the COMPUSTAT expanded annual industrial and full-coverage files, the COMPUSTAT research annual industrial file, and the Center for Research in Securities Prices (CRSP) New York Stock Exchange (NYSE) / American Stock Exchange (AMEX) and NASDAQ files. We restrict our sample to firms with Standard Industrial Classification (SIC) codes from 2000 to 5999 to focus on the industrial corporate sector. Our data span the years 1974 through 1992.Debt MaturityMeasuring Debt MaturityCOMPUSTAT reports the amount of long-term debt payable in years one through five from the firm’s fiscal year end. To measure the maturity structure of a firm’s debt, we examine the percentage of the firm’s total debt (long-term debt plus debt in current liabilities) that has a maturity of more than three years. Several firms on the COMPUSTAT file report less than 0 percent or more than 100 percent of their

total debt maturing in more than three years. Since these observations apparently reflect data-coding errors we discard them. The choice of three years, admittedly, is arbitrary. Our results are qualitatively similar, nevertheless, using the percentage of debt maturing in more than one, two, four, or five years.“

Table 4.7: Summary Statistics for the Percentage of Total Debt that Matures in More than One, Two, Three, Four, and Five Years from the Fiscal Year-End

Percentage of debt that matures in more than

Mean Standard Deviation

25th

Percentile Median 75th

Percentile Value Weighted Mean*

One year 71.8 28.2 59.8 82.9 92.7 73.0Two years 60.9 29.1 42.6 69.5 84.0 65.7Three years 51.7 28.9 29.5 57.1 75.1 58.7Four years 43.7 28.1 19.7 46.0 66.4 52.2Five years 36.6 27.2 11.5 35.8 58.1 45.9*Weighted by total debt outstanding

We measure debt maturity as the ratio of long-term debt to total debt in order to separate the debt maturity decision from the leverage decision. Titman and Wessels (1988) examine the ratio of long-term debt to total assets and the ratio of short-term debt to total assets. Their evidence shows that firms with higher leverage issue both more long—term debt and more short-term debt. Their specification, nevertheless, does not provide a clear picture of how the mix of long-term and short-term debt varies with firm characteristics. By examining long-term debt as a fraction of total debt, we more carefully focus on the debt maturity decision.

Descriptive Statistics on Debt Maturity

Table 4.7 provides debt-maturity summary statistics for the pooled time series and cross-section of firms on CRSP and COMPUSTAT between 1974 and 1992 (39,949 firm-year observations). The table reports the percentage of total debt that matures in more than one through five years from the firm’s fiscal year-end. On average, these firms have 71.8 percent of their debt due in more than one year, 51.7 percent of their debt due in more than three years, and 36.6 percent of their debt due in more than five years. There is also substantial cross-sectional variation in debt maturity. For instance, the interquartile range for the percentage of debt due in more than three years rises from 29.5 to 75.1 percent.

To reflect the potential affect of firm size on the debt-maturity decision, we also report the value-weighted average debt maturity. By weighting annotations by the firm’s total debt outstanding, this average provides a more representative approximation of economy-wide debt maturity. On the short end of the maturity spectrum, the value-weighted and equally Weighted averages are similar. For longer debt maturities, nevertheless, the difference is larger. The average percentage of debt that matures in more than five years increases from 36.6 percent (equally weighted) to 45.9 percent (value-weighted).

Potential Measurement Problems

Using COMPUSTAT balance-sheet data gives a broad outlook of corporate debt. In addition to bonds and mortgages, long-term debt also includes capitalize lease obligations, forestry and paper companies’ timber contracts, publishing companies’ royalty contracts payable, and similar long-term fixed claims. Short-term debt (debt in current liabilities) comprises short-term notes, bank acceptances and overdrafts, the current segment of long-term debt, and sinking funds or installments on loans.

In principle, a theory of debt maturity should be appropriate to this broad spectrum of corporate liabilities. In practice, together with these non-debt fixed claims is unlikely to have a material effect on the analysis. In our illustration, the most widely used non-debt fixed claims are capitalized leases. Capitalized leases appear on the balance sheet in 44 percent of our firm-year observations. For these firms, the lease obligations represent 20 percent of total debt, for all firm-years in the sample, lease obligations represent 9 percent of total debt.

To the extent that the firm’s debt comprises imbedded options in the call or sinking-fund provisions, our measure overstates the effective maturity. This measurement-error problem introduces a potential bias if the use of these provisions is correlated with our independent variables. For instance, firms with more growth options might comprise more imbedded options in their debt contracts. This entails that the firms that are anticipated to employ the shortest-term debt will have the largest overstatement. Consequently, such a measurement problem would bias our tests against finding results regular with the contracting-cost hypothesis.

An analogous measurement quandary occurs with affirmative covenants in debt agreements. Affirmative covenants necessitate firms to keep up specified levels in accounting-based financial ratios. (For instance, an affirmative covenant might specify a minimum level of working capital or net worth.) The violation of an affirmative covenant is an event of default typically giving the lender the right to speed up the maturity of the debt. Consequently to the extent that the debt includes affirmative covenants, our measure also overstates effective maturities. Affirmative compacts are practically completely employed in private debt agreements. This implies that small, risky firms, which we expect to employ confidential debt agreements, have the biggest overstatement. This again biases our tests against finding results consistent with the contracting-cost hypothesis.

Exogenous Variables

Investment Opportunity Set

Smith and Watts (1992) and Gaver and Gaver (1993) study the relationship between measures of the firm’s investment opportunity set and the firm’s financing, compensation, and dividend policies. These studies make use of the ratio of the market value of the firm’s assets to their book value as a proxy for growth options. The firm’s balance sheet does not include intangible assets like growth options. More growth options thus increase the firm’s market value in relation to its book value. Regular with incentive-contracting hypotheses, Smith and Watts find that the market-to-book ratio is considerably connected with the firm’s observed policy choices. As a result, the principal variable used in this study to proxy for the firm’s investment opportunity set is the market-to-book ratio. We approximate the market value of the firm’s assets as the book value of assets minus the book value of equity plus the market

value of equity. The market—to-book ratio is the predictable market value of assets divided by the book value of assets. This estimated market-to-book ratio has several extreme observations. For instance, 98 percent of the ratios are between 0.5 and 6.9. The range for this variable, nevertheless, is 0.1 to 291.3. To get rid of the influence of these extreme observations on the regression results, we discard observations if the market-to-book ratio is greater than ten. (Discarding these observations reduces the statistical meaning of this variable in the regressions, but increases the size of the coefficient.)

Regulation

To estimate the effects of regulation, we construct a dummy variable that is set equal to one for firms in regulated industries and zero otherwise. Regulated industries include railroads (SIC code 4011) through 1980, trucking (4210 and 4213) through 1980, airlines (4512) through 1978, telecommunications (4812 and 4813) through 1982, and gas and electric utilities (4900 to 4939).Firm SizeWe measure firm size as the natural logarithm of our estimate of the market value of the firm in constant 1972 dollars.Firm QualityThe signaling literature has given little attention to defining quality in a way that is empirically observable. Nevertheless, signaling models all assume that managers have better (or more timely) information about firm value than investors. To estimate quality empirically, We use the firm’s abnormal future earnings. We assume that high-quality (undervalued) firms have positive future abnormal earnings and low-quality (over-valued) firms have negative future abnormal earnings. Assuming that earnings follow a random walk, We define abnormal earnings in year t + 1 as earnings per share in year t + 1 (excluding extraordinary items and discontinued operations and adjusted for any changes in shares outstanding) minus earnings per share in year t, divided by the year t share price. This variable also has several extreme observations. For example, 98 percent of the abnormal earnings are between — 1.2 and 1.6, while the range for this variable is 92,77 3 to 1,451. We discard observations if the absolute value of the abnormal earnings variable is greater than five. Discarding these observations has a material effect on the estimated coefficient for the abnormal-earnings variable. This coefficient is not significantly different from zero in any regression that includes these extreme observations. Discarding these observations, nevertheless, has little effect on the other variables in the regression.Term StructureTo measure the term structure of interest rates, we subtract the month-end yield on six-month government bonds from the month-end yield on ten-year government bonds. This yield spread is then matched with the month of the companies’ fiscal year-end. Data on government bond yields are from Citibase.The Determinants of Debt Maturity StructureThe debt-maturity is one of numerous financing choices that the firm must make at the same time. When determining how to finance itself, the firm must choose between debt and equity. If it chooses debt, it must also choose the maturity of that debt, the priority, whether the debt is public or private, and other contract provisions, together with call and convertibility provisions and restrictive covenants. In an ideal world, we would like to provide a system of instantaneous equations to control for these joint decisions. Unluckily, the theory currently is not rich enough to provide the essential identifying limitations for this system.Although these financing decisions are decided jointly, theory suggests that they are driven by the same underlying firm characteristics. For case in point, a firm with more growth options in its investment opportunity set is likely to have less debt its capital structure, and the debt it issues is likely to have

shorter average maturity. A growth firm is also more likely to issue private debt with higher priority and more restrictive covenants. Our regressions should be interpreted as reduced-form regressions from this system. Since we are unable to provide the identifying restrictions for the simultaneous system, we restrict our choice of independent variables to those that we believe are most likely to be exogenous—i.e., measures of the firm’s investment opportunity set, its informational environment, and its regulatory and tax status.Table 4.8 reports pooled time series, cross-sectional regressions of the percentage of a firm’s debt payable in more than three years on its market-to-book ratio, a regulation dummy, the log of firm value, the firm’s future abnormal earnings, and the risk-free term structure. Regression (1) in table 4.8 is a ordinary-least-squares (OLS) regression on the entire pooled time series and cross section (37,979 firm-year observations.)3

Since the normal OLS assumption of independent errors is unlikely to be satisfied in this regression, t-statistics are potentially overstated. To account for the potential error-dependence problem, regression (2) is a cross-account for the potential using the time-series mean of each variable by firm (5,545 observations). Running the regression in a single cross-section eliminates the problem of serially correlated errors. The cross-sectional regression preserves the dispersion across firms, but does not exploit any time-series variation in the observations.In regression (3), we use a fixed-effects regression as an alternate method for dealing with the problem of serially correlated errors. In this regression, we subtract the firm-specific time-series mean for each variable from each observation. (This technique is equivalent to adding a dummy variable for each firm in the sample.) Thus, although this regression preserves the time-series dispersion in the sample, it ignores most of the information from differences across firms.The Percentage of debt that matures in more than three years is regressed on the firm’s market-to-book ratio, a dummy variable for firms in regulated industries, the natural log of firm value (market value of equity plus book value of liabilities in constant amount of money), the firm’s future abnormal earnings, and the risk-free term structure (the difference between ten-year government bond yields and six-month government bond yields). The table reports estimates from pooled ordinary least squares, cross sectional ordinary least squares and fixed-effects regressions. Sample: All firms on both Center for Research in Securities Prices and COMPUSTAT between 1974 and 1991 with Standard Industrial Classification (SIC) codes from 2000 to 5999.(White adjusted t-statistics in parentheses)

Table 4.8: Regressions Estimating the Determinants of Corporate Debt Maturity(1) (2) (3)

Independent Variable Predicted sign

Pooled Regression

Cross-Sectional Regression

Fixed Effects

Intercept 36.74(89.50)

31.49(27.98)

N.A.

Market-to-book ratio – –4.37(–25.42)

–4.78(–15.12)

–2.33(–9.51)

Regulation dummy + 6.66(8.96)

8.91(3.97)

3.54(3.29)

Log of firm value + 5.28(75.98)

6.31(35.98)

5.03(18.81)

3 The number of observations in Table 4.8 is smaller than the number of observations in Table 4.7 for the reason that the abnormal earnings variable cannot be calculated for the final data year, and for the reason that observations with extreme values are excluded from the regressions. Less than 1 percent of the sample (386 firm-years) is discarded due to extreme observations.

Abnormal earnings – –1.94(–5.04)

–2.94(–2.38)

—

Term Structure + –0.32(–3.07)

–0.63(–1.21)

–0.21(–2.84)

R2 0.16 0.26 0.61a

F 1414.42 398.14 116.17Number of observations 37969 5545 116.17

aThe adjusted R2 for the fixed effects regression includes the fixed effects. The adjusted R 2 excluding the fixed effects is 0.02.b. The fixed effects regression excludes firms with only one observation.

Regression ResultsConsistent with the contracting-cost hypothesis, the coefficient on the market-to-book ratio is negative and highly significant in all three regressions. The t-statistic for this variable ranges from –9.51 for the fixed-effects regression to ~25/12 for the pooled OLS regression. This coefficient indicates that, on average, firms with more growth options (as proxied by the market-to-book ratio) have significantly less long-term debt. It is also consistent with the hypothesis that firms with larger potential information asymmetries issue long-term debt.To derive a better understanding of the economic significance of this result, we calculate the impact of moving from the tenth to the ninetieth percentile for the market-to-book ratio in our sample. The estimated coefficient from the pooled regression implies that this move reduces the fraction of long-term debt by 9.6 percentage points.The dummy variable for firms in regulated industries is significantly positive; t-statistics range from 3.29 to 8.96. Other things equal, regulation increases the proportion of long-term debt by 6.6 percentage points.The coefficient for the log of firm value is positive and significant in all three regressions; t-statistics range from 18.81 to 75.98. The economic significance of this variable is also dramatic. The pooled-regression coefficient implies that moving from the tenth to the ninetieth percentile for firm size increases the fraction of long-term debt by 27.7 percentage points.The regression provide some support for the signaling hypothesis; t-statistics range from –2.38 to –5.04, implying that firms with higher future earnings have more short-term debt. But the economic significance of this factor is questionable. For abnormal earnings, the pooled regression coefficient implies that moving from the tenth to the ninetieth percentile reduces the fraction of long-term debt by 0.6 percentage points. While the time-series stability of a firm’s investment opportunities is plausible, firm quality is less likely to be stable over time. For example, such stability would be difficult to reconcile with the existing evidence on market efficiency. The prediction of the signaling hypothesis is thus primarily time-series in nature. Consequently, it is not surprising that the coefficient on the abnormal-earnings variable has lower significance in the cross-sectional regression.Table 4.8 results provide no support for the Brick and Ravid tax hypothesis. Inconsistent with their hypothesis, the coefficient for the term-structure variable is significantly negative in the pooled and fixed-effects regressions. For the term-structure variable, the point estimate from the pooled regression implies that moving from the tenth to the ninetieth percentile of the term-structure variable would increase long-term debt by only 1.1 percentage points. This limited impact is broadly consistent with Lewis’ (1990) analysis; he argues that if optimal debt-asset ratios and debt-maturity structures are chosen simultaneously, then taxes do not affect optimal debt maturity. The prediction of the tax hypothesis is also primarily time-series in nature.

The F-statistics indicate that each of the regressions in Table 4.8 is significant at reasonable levels. Adjusted R2s of 0.16 and 0.26 in the pooled and cross-sectional regressions with the much smaller R2 of 0.02 in the fixed-effects regression (when the influence of the fixed effects are excluded) suggest that it is primarily the variation in maturity structure across firms that provides the explanatory power in these regressions.Correlations among Financial Policy VariablesSmith and Watts (1992) argue that if the firm’s investment opportunity set is an important determinant of its financing, dividend, and compensation policies, then the investment opportunity set induces correlations among these corporate policies. If the firm’s investment opportunity set is also an important determinant of its debt maturity structure, then debt maturity should be correlated with leverage, payout, and compensation policies.In table 4.9, we report the Pearson correlation coefficients between the percentage of debt payable in more than three years, leverage (book value of debt dividend by the market value of the firm), and dividend yield (dividend per share divided by price.) As predicted by the incentive-contracting hypothesis, the correlation between debt maturity and leverage is significantly positive, indicating that firms with greater leverage have more long-term debt. The correlation between debt-maturity and dividend yield is also significantly positive, indicating that firms with higher dividend yields tend to have more long-term debt.Figure 4.9: Pearson Correlation Coefficients between Debt Maturity, Leverage, and Dividend YieldSample: All firms on both Center for Research in Securities Prices and COMPUSTAT between 1974 and 1992 with Standard Industrial Classification (SIC) codes from 2000 to 5999. 39,949 observations. (p-values in parentheses.)

Percentage of Debt Due in More than 3 years

Leverage Dividend Yield

Percentage of debt due in more than 3 years

1.000

Leverage 0.120(0.0001)

1.000

Dividend yield 0.049(0.0001)

0.040(0.0001)

Specification ChecksInvestment Opportunity SetThe market-to-book ratio is an admittedly noisy proxy for growth options. To analyze the robustness of our results to alternate proxies, we reestimate the regressions reported in table 4.9. In table 4.10, we examine the firms’ annual research and development expenses as a percentage of firm value, depreciation as a percentage of firm value, the earnings-price ratio, and the asset-return standard deviation (the firm’s stock-return standard deviation in percent times its equity-to-value ratio). We expect firms with lower research and development expenses, more depreciation expenses, higher earnings-price ratios and lower return standard deviations, to have more tangible assets and fewer growth options in their investment opportunity sets. In these regressions, the coefficients on the investment-opportunity set proxies are all statistically significant (t-statistics range from 12.06 to 28.75 in absolute value) and have the signs predicted by the contracting cost hypothesis. (The sample sizes for these regressions are smaller than the sample size in table 4.8 due to missing COMPUSTAT data.)Table 4.10: Specification Checks for Regressions Estimating the Determinants of Corporate Debt Maturity

The percentage of debt that matures in more than three years is regressed on a proxy for the firm’s investment opportunity set, a dummy variable for firms in regulated industries, the natural log of firm value (market value of equity plus book value of liabilities in constant amount of money), the firm’s future abnormal earnings, and the risk-free term structure (the difference between ten-year government bond yields and six-month government bond yields). The investment-opportunity-set variables include the firm’s research and development (R&D) expense divided by firm value, depreciation expense divided by firm value, the earnings-price ratio, and the firm’s asset return standard deviation (stock return standard deviation times the equity-to-firm-value ratio).Sample: All firms on both Center for Research in Securities Prices and COMPUSTAT between 1974 and 1991 with Standard Industrial Classification (SIC) codes from 2000 to 5999. (White-adjusted t-statistics in parentheses)

Independent Variable Predicted Sign

(1) (2) (3) (4)

Intercept 35.83(71.77)

28.30(66.25)

31.72(89.98)

44.87(79.92)

R&D/firm value – –0.77(–15.05)

Depreciation/firm value + 0.74(12.06)

Earnings-price ratio + 13.21(20.59)

Asset return std. deviation

– 4.50(2.87)

6.17(8.10)

7.37(9.95)

6.32(8.72)

Log of firm value + 4.37(49.20)

5.31(75.46)

4.96(69.33)

4.05(50.39)

Abnormal earnings – –1.78(–3.33)

–2.39(–6.16)

2.30(4.57)

–1.60(–4.06)

Term structure + –0.77(–5.58)

–0.59(–5.78)

–0.37(–3.55)

–0.25(–2.42)

R2 0.12 0.14 0.15 0.16F 627.80 126.15 1296.79 1347.16Number of observations

22072 37955 37643 35274

Including these alternate measures of the investment opportunity set generally leaves unaffected the other estimated coefficients. A noteworthy exception is in regression (3). If the earnings-price ratio is used as the measure of growth options, the abnormal-earnings variable changes sign and becomes significantly positive (the t-statistics is 4.57). This suggests potential regression misspecification and multicollinearity between the variables. In an efficient market, firms with high expected growth rates in earnings will have high current prices; thus, on average, firms with large abnormal future earnings tend to have low current earnings-price ratios.Scale Economies, Public Debt, and Commercial PaperScale EconomiesFirms with little outstanding debt are likely to receive short-term loans from a bank rather than issue long-term bonds and incur the higher fixed costs of a public issue. Thus, there is a potential mechanical association between leverage and debt maturity involving firms with low levels of debt. Smith and Watts (1992) find a negative relation between leverage and the firm’s market-to-book ratio. Hence, a

mechanical association between leverage and debt maturity could generate a spurious relation between debt maturity and the firm’s market-to-book ratio.To test for this potential spurious correlation, we exclude firms that have small amounts of total debt. Regression (1) in Table 4.11 reports the coefficients from our basic regression excluding firms that have less than TZS50 million of total debt. (The TZS50 million cutoff is approximately equal to the tenth percentile of our sample firms that have public debt ratings.) Excluding these firms reduces the number of observations from 37,969 to 11,798 and the number of different firms from 5,545 to 1,580.The t-statistics for the estimated coefficients generally decline with the sample size. Nevertheless the coefficient on the market-to-book ratio remains negative and significant, and the coefficient on the regulation dummy remains positive and significant as in all previous regressions. Thus, a mechanical relation debt maturity and the total amount of debt or leverage does not appear to drive our basic results.The only material change in this regression is that the coefficient on firm size is small and only marginally significant. If we exclude firms with less that TZS100 million of total debt (approximately equal to the twenty-fifth percentile of firms with public debt ratings in our sample), the coefficients for the market-to-book ratio and the regulation dummy are again unaffected. Nevertheless, the coefficient on firm size now becomes negative and statistically significant. This effect on the firm-size coefficient indicates a nonmonotonic relation between firm size and debt maturity. Debt maturity increases with firm size for firms smaller than TZS1 billion of market value. After that point, there appears to be a negative relation size and maturity.Public Debt and Commercial PaperAs noted above, scale economies associated with public debt issues make this market inaccessible to many companies with small amounts of total debt. Table 4.11: Specification Checks for Regression Estimating the Determinants of Corporate Debt MaturityThe percentage of debt that matures in more than three years is regressed on the firm’s market-to-book ratio, a dummy variable for firms in regulated industries, the natural log of firm value (market value of equity plus book value of liabilities in constant amount of money), the firm’s future abnormal earnings, the risk-free term structure (the difference between ten-year government bond yields and six-month government bond yields), dummy variables for firms with Standard and Poors bond and commercial-paper ratings, a numeric variable for the firm’s bond rating (set equal to one for AAA through 27 for CCC or below, and equal to 28 for nonrated firms), and three-digit SIC industry dummy variables. Sample: All firms on both Center for Research in Securities Prices and COMPUSTAT between 1974 and 1991 with Standard Industrial Classification (SIC) codes from 2000 to 5999. (White-adjusted t-statistics are in parentheses)

Independent Variable

Firms with more than

TZS50 Million of Total Debt

(1)

Includes Public Bond Ratings and Commercial

Paper Dummies (2) (3)

Includes 3-digit SIC Industry

Dummies(4)

Intercept 64.98(60.67)

37.38(87.40)

–2.76(–1.23)

NAa

Market-to-book ratio

–2.53(–4.63)

–4.12(–24.12)

–4.05(–23.84)

–3.58(–19.84)

Regulation dummy 7.48(10.04)

6.86(8.93)

6.70(8.69)

5.09(4.55)

Log of firm value 0.39(2.51)

4.72(55.69)

5.00(57.67)

4.99(62.44)

Abnormal earnings –1.26(–1.57)

–1.95(–5.06)

–1.94(–4.94)

–1.72(–4.54)

Term structure 0.04 –0.30(–2.90)

–0.39(–3.83)

–0.30(–2.98)

Bond rating dummy 11.21(28.22)

33.69(25.97)

Commercial paper dummy

–13.72(–20.67)

–9.52(–14.20)

Bond rating for rated firms

1.40(18.44)

F-statistic for industry dummiesb

[25.82]

R2 0.01 0.17 0.18 0.21F 21.03 1138.51 1043.92 932.49Number of observations

11798 37969 37969 37969

aThis regression is rum without an intercept for the reason that there is a dummy variable for each 3-digit SIC industry.bThe F-statistic for the industry dummies (in parentheses) tests whether the industry dummies are jointly different from zero.

Since public debt is issues tend to have long maturities, this implies a potential mechanical relation between access to the public debt market and a firm’s debt maturity.Similarly, the high fixed cost of commercial paper programs excludes all but the largest firms from this source of funds. Commercial paper (short-term promissory notes issued in the open market) offers an alternative to bank borrowing for large corporations with strong credit ratings. The most common maturity of commercial paper is 30 to 50 days. (The maturity of commercial paper is generally less than 270 days for the reason that the Securities Act of 1933 exempts such issues from registration.) Thus, firms with commercial paper programs are likely to have more short-term debt.We examine the sensitivity of our results to the firms’ use of public debt and commercial paper. In regression (2) of table 4.11 we include dummy variables to indicate whether the firm has an Standard and Poor’s (S&P) bond rating or an S&P commercial paper rating. As expected, the coefficient for the bond-rating dummy is significantly positive and the coefficient for the commercial-paper dummy is significantly negative. Nevertheless, the inclusion of these variables does not affect the size or the statistical significance of the other coefficient. Thus, access to public debt and commercial paper markets does not appear to be driving our basic results.Signaling and Credit Risk We have used the bond and commercial paper ratings to provide information about potential mechanical relations between the firm’s choice of debt instruments and debt maturity; these results are also potentially relevant for examining signaling models that relate credit risk to debt maturity. Although the data are not ideally suited to test these hypotheses, we examine the relation between bond ratings and debt maturity. Regression (3) in table 4.11 includes a bond-rating variable; it equals to one if the firm’s S&P bond rating is AAA, through 27 if the bond rating is CCC or below. This variable is set equal to 28 for firms without S&P bond ratings. The coefficient on the bond-rating variable is positive and

significant (t-statistic is 18.44), indicating that lower-rated firms issue more long-term debt than higher-rated firms. The positive coefficient for the bond-rating dummy variable, nevertheless, indicates that non-rated firms (which tend to be small firms with the lowest credit standing) have more short-term debt. Thus, the evidence suggests a nonmonotonic relation between credit standing and debt maturity as predicted by Diamond (1993).Although there is evidence of a nonmonotonic relation between debt maturity and credit standing, this evidence should be interpreted with some caution. For firms with public debt ratings, the relation between debt maturity and bond rating is strictly monotonic; the nonmonotonicity is driven solely by the nonrated firms. As one moves from the sample of firms with S&P bond ratings to those that do not have ratings, nevertheless, many things change. In particular, firms typically get bond ratings only if they issue public debt. Thus, firms without S&P bond ratings have mostly private debt which, on average, has shorter maturity regardless of the firm’s credit rating.Industry Effects, Deregulation, and Time TrendsIndustry EffectsTo determine the relative importance of industry-specific versus firm-specific effects, we reestimate the regressions in Table 4.8 including dummy variables for 2-digit and 3-digit SIC industries. We report the results using 3-digit industry dummies in regression (4) of Table 4.11. The F-statistic for the industry dummies indicates that they are jointly different from zero (F=25.82). Nevertheless, the adjusted R2 in this regression increases from 0.16 in regression (1) of Table 4.8 to only 0.21 with the 3-digit industry dummies. The adjusted R2 for the regression with 2-digit industry dummies (not reported here) is 0.19. In addition, adding these dummy variables has no material effect on any of the other regression coefficients. Evidence of standard industry practices still would be consistent with the hypothesis in section I of this topic. For example, knowledge of a firm’s industry may be highly correlated with its investment opportunities. Nevertheless, the evidence suggests that firm-specific characteristics (such as size and market-to-book ratios) are more important than industry-specific effects in determining a firm’s debt maturity.DeregulationThe four major industries in our sample (airlines, railroads, trucking, and telecommunications) experienced significant deregulation during our sample period. If the effect of regulation is to decrease managerial discretion over future investment decisions, and thus increase debt maturity, then these firms should shorten their debt maturity following deregulation. As predicted, all four industries have significantly more short-term debt after deregulation. Following deregulation, the average percentage of debt due in more than three years fell from 67 to 64 percent in the airline industry, from 76 to 67 percent in the railroad industry, from 57 to 46 percent in the trucking industry, can be explained almost entirely by the more general trend in debt maturities. Including this more general time trend in debt maturity eliminates the significance of the effect of deregulation.Time TrendsThe fact that the effects of deregulation are covered by the general time trend in debt maturity raises the likelihood that this trend has a significant effect on other regression coefficients. We test the importance of this time trend and for other temporal effects (such as business cycle effects). We reestimate the regressions in Table 4.8 using both a linear time-trend variable and annual dummy variables. As expected both the time-trend and dummy-variable approaches indicate that economy-wide debt maturity declined over this period. None of the other regression coefficients are substantially affected in magnitude or statistical significance, nevertheless, by the inclusion of a time trend or year dummies. (Since there is no material change in any of the regression coefficients of interest, we do not report this regression.)Our examination of the determinants of corporate debt maturity supports the hypothesis that firms with more growth options in their investment opportunity sets issue more short-term debt. This result is

consistent with Myers’ (1977) prediction that reducing debt maturity helps control the underinvestment problem. We also find that regulated firms issue more long-term debt. This is consistent with Smith’s (1986) argument that regulation reduces the firm’s discretion over corporate investment policy, thus controlling the underinvestment problem. The empirical results are robust to alternate measures of the investment opportunity set. Our methods also suggest that growth options in the firm’s investment opportunity set are important in explaining both the time-series and cross sectional variation in the firm’s maturity structure.We provide evidence of a strong association between firm size and debt maturity: large firms issue a significantly higher proportion of long-term debt. This is consistent with the observation that small firms rely more heavily on bank debt that typically has shorter maturity than public debt. We also document a reliably positive association between the existence of an S&P bond rating and debt maturity, and a reliably negative association between the existence of a commercial paper rating and debt maturity.Our evidence provides less support for the hypothesis that firms use the maturity of their debt to signal information to the market. Although the estimated coefficient on the abnormal earnings variable is significantly negative, the economic impact of this variable is trivial. The results are consistent, nevertheless, with the hypothesis that firms with potentially large information asymmetries (such as high-growth firms) issue more short-term debt. Consistent with Diamond’s (1993) hypothesis, our evidence also suggests that firms with the highest and lowest credit risk issue short-term debt while firms with intermediate credit risk issue long-term debt. The tax hypothesis is not important in explaining the debt maturity choice.Our analysis has some potentially significant limitations. First, the power of the hypothesis tests are not all equal. For instance, we examine variation in debt maturity across all of the firms’ outstanding debt. More powerful tests of the signaling hypotheses may come from examining the variation in debt maturity at issuance.Second our data employ a broad definition of debt. Although we believe the behavior of the aggregate is more significant than the behaviour of any single component, more disaggregated data would facilitate examination of the richness of the debt-maturity policy. For example, our evidence indicates that firms with more growth options in their investment opportunity sets issue more short-term debt. Yet we do not know how much of the total variation is (such as long-term public debt and short-term bank debt) and how much is due to variation in the maturity of a given set of instruments (such as 20 year and 12 year public debentures). Similarly, we do not have data on foreign borrowing. Thus, we cannot identify the variation in debt maturity attributable to the differential maturity of Tanzania versus foreign debt. A more detailed examination of the mix of debt instruments issued by different types of firms would provide important texture to our understanding of how firms achieve their observed debt maturity structure.To end with, we have simply taken the COMPUSTAT definition of the amount of debt that matures in given year to analyze the maturity choice. Many debt issues have provisions that specify imbedded options to modify the repayment schedule. For instance, both call provisions and optional sinking fund payments reduce the effective maturity of a debt issue. Debt covenants, particularly affirmative covenants in private debt agreements, also allow the lender to accelerate the payment schedule if the borrower breaches the covenant. The appropriate adjustment in our methods to reflect this maturity flexibility is not obvious.POSSIBLE REASONS FOR AN “OPTIMAL” MIX OF DEBT AND EQUITYWhat is the optimal mix of debt and equity for a firm? In the previous discussion we looked at the qualitative trade-off between debt and equity, but we did not develop the tools we need to analyze whether debt should be 0 percent, 20 percent, 40 percent, or 60 percent of capital. Debt is always cheaper than equity, but using debt increases risk in terms of default risk to lenders and higher earnings volatility for equity investors. Thus, using more debt can increase value for some firms and decrease

value for others, and for the same firm, debt can be beneficial up to a point and destroy value beyond that point. We have to consider ways of going beyond the generalities in the last chapter to specific ways of identifying the right mix of debt and equity.In the following discussion we explore four ways to find an optimal mix. The first approach begins with a distribution of future operating income; we can then decide how much debt to carry by defining the maximum possibility of default we are willing to bear. The second approach is to choose the debt ratio that minimizes the cost of capital. We review the role of cost of capital in valuation and discuss its relationship to the optimal debt ratio. The third approach, like the second, also attempts to maximize firm value, but it does so by adding the value of the unlevered firm to the present value of tax benefits and then netting out the expected bankruptcy costs. The final approach is to base the financing mix on the way comparable firms finance their operations.Operating Income ApproachThe operating income approach is the simplest and one of the most intuitive ways of determining how much a firm can afford to borrow. We determine a firm’s maximum acceptable probability of default as our starting point, and based on the distribution of operating income and cash flows, we then estimate how much debt the firm can carry.Steps in Applying Operating Income ApproachWe begin with an analysis of a firm’s operating income and cash flows, and we consider how much debt it can afford to carry based on its cash flows. The steps in the operating income approach are as follows:

1. We assess the firm’s capacity to generate operating income based on both current conditions and past history. The result is a distribution for expected operating income, with probabilities attached to different levels of income.

2. For any given level of debt, we estimate the interest and principal payments that have to be made over time.

3. Given the probability distribution of operating income and the debt payments, we estimate the probability that the firm will be unable to make those payments.

4. We set a limit or constraint on the probability of its being unable to meet debt payments. Clearly, the more conservative the management of the firm, the tighter this probability constraint will be.

5. We compare the estimated probability of default at a given level of debt to the probability constraint. If the probability of default is higher than the constraint, the firm chooses a lower level of debt; if it is lower than the constraint, the firm chooses a higher level of debt.

Limitations of the Operating Income ApproachAlthough this approach may be intuitive and simple, it has key drawbacks. First, estimating a distribution for operating income is not as easy as it sounds, especially for firms in businesses that are changing and volatile. The operating income of firms can vary widely from year to year, depending on the success or failure of individual products. Second, even when we can estimate a distribution, the distribution may not fit the parameters of a normal distribution, and the annual changes in operating income may not reflect the risk of consecutive bad years. This can be remedied by calculating the statistics based on multiple years of data. This approach is also an extremely conservative way of setting debt policy for the reason that it assumes that debt payments have to be made out of a firm’s operating income and that the firm has no access to financial markets or pre-existing cash balance. Finally, the probability constraint set by management is subjective and may reflect management concerns more than stockholder interests. For instance, management may decide that it wants no chance of default and refuse to borrow money as consequence.Refinements on the Operating Income Approach

The operating income approach described in this section is simplistic for the reason that it is based on historical data and the assumption that operating income changes are normally distributed. We can make it more sophisticated and robust my making relatively small changes.

We can look at simulations of different possible outcomes for operating income, rather than looking at historical data; the distributions of the outcomes can be based both on past data and on expectations for the future.

Instead of evaluating just the risk of defaulting on debt, we can consider the indirect bankruptcy costs that can accrue to a firm if operating income drops below a specified level.

We can compute the present value of the tax benefits from the interest payments on the debt, across simulation, and thus compare the expected cost of bankruptcy to the expected tax benefits from borrowing.

With these changes, we can look at different financing mixes for a firm and estimate the optimal debt ratio as that mix that maximizes the firm’s value.

Cost of Capital ApproachIn the beginning of this unit we estimated the minimum acceptable hurdle rates for equity investors (the cost of equity), and for all investors in the firm (the cost of capital). We defined the cost of capital to be the weighted average of the costs of the different components of financing–including debt, equity and hybrid securities–used by a firm to fund its investments. By altering the weights of the different components, firms might be able to change their cost of capital. In the cost of capital approach, we estimate the costs of debt and equity at different debt ratios, use these costs to compute the costs of capital, and look for the mix of debt and equity that yields the lowest cost of capital for the firm. At this cost of capital, we will argue that firm value is maximized.Cost of Capital and Maximizing Firm ValueIn the previous unit we laid the foundations for estimating the cost of capital for a firm. We argued that the cost of equity should reflect the risk as perceived by the marginal investors in the firm. If those marginal investors are diversified, the only risk that should be priced in should be the risk that cannot be diversified away, captured in a beta (in the CAPM) or betas (in multi factor models). If the marginal investors are not diversified, the cost of equity may reflect some or all of the firm-specific risk in the firm.The cost of debt is a function of the default risk of the firm and reflects the current cost of long term borrowing to the firm. Since interest is tax deductible, we adjust the cost of debt for the tax savings, using the marginal tax rate, to estimate an after-tax cost. In summary, the cost of capital is a weighted average of the costs of equity and debt, with the weights based upon market values:

To understand the relationship between the cost of capital and optimal capital structure, we first have to establish the relationship between firm value and the cost of capital. In the earlier sections of this unit we noted that the value of a project to a firm could be computed by discounting the expected cash flows on it at a rate that reflected the riskiness of the cash flows, and that the analysis could be done either from the viewpoint of equity investors alone or from the viewpoint of the entire firm. In the latter approach, we discounted the cash flows to the firm on the project, that is, the project cash flows prior to debt payments but after taxes, at the project’s cost of capital.Extending this principle, the value of the entire firm can be estimated by discounting the aggregate expected cash flows to the firm over time at the firm’s cost of capital. The firm’s aggregate cash flows can estimated as cash flows after operating expenses, taxes, and any capital investments needed to create future growth in both fixed assets and working capital, but before debt payments.

Cash Flow to Firm = EBIT (1–t) – (Capital Expenditures – Depreciation) – Change in Non-cash Working CapitalThe value of the firm can then be written as

The value of a firm is therefore a function of its cash flows and its costs of capital. In the special case where the cash flows to the firm remain constant as the debt/equity mix is changed, the value of the firm will increase as the cost of capital decreases. If the objective in choosing the financing mix for the firm is the maximum of firm value, this can be accomplished, in this case, by minimizing the cost of capital. In the more general case where the cash flows to the firm themselves change as the debt ratio changes, the optimal financing mix is the one that maximizes firm value.The Cost of Capital Approach – BasicsTo use the cost of capital approach in its simplest form, where the cash flows are fixed and only the cost of capital changes, we need estimates of the cost of capital at every debt ratio. In making these estimates, the one thing we cannot do is keep the costs of debt and equity fixed, while changing the debt ratio. In addition to being unrealistic in its assessment of risk as the debt ratio changes, this analysis will yield the unsurprising conclusion that the cost of capital is minimized at a 100 percent debt ratio.As the debt ratio increases, each of the components in the cost of capital will change. Let us start with the equity component. Equity investors are entitled to the residual earnings and cash flows in a firm, after interest and principal payments have been made. As that firm borrows more money to fund a given level of assets, debt payments will increase, and equity earnings will become more volatile. This higher earnings volatility, in turn, will translate into a higher cost of equity. In the language of the CAPM and multi-factor models, the beta or betas we use for equity should increase as the debt ratio goes up. The debt holders will also see their risk increase as the firm borrows more. Holding operating income constant, a firm that contracts to pay more to debt holders has a greater change of defaulting, which will result in a higher cost of debt. As an added complication, the tax benefits of interest expenses can be put at risk, if these expenses become greater than the earnings.The key to using the cost of capital approach is coming up with realistic estimates of the cost of equity and debt at different ratios. The optimal financing mix for a firm is trivial to compute if one is provided with a schedule that relates to costs of equity and debt to the debt ratio of the firm. Computing the optimal debt ratio then becomes purely mechanical. To illustrate, assume that you are given the costs of equity and debt at different debt levels for a hypothetical firm and that the after-tax cash flow to this firm is currently TZS200 million. Assume also that these cash flows are expected to grow at 3 percent a year forever, and are unaffected by the debt ratio of the firm. The cost of capital schedule is provided in Table 4.12, along with the value of the firm at each level of debt.

Table 4.12: WACC, Firm Value, and Debt RatiosD/(D+E) Cost of

EquityAfter tax Cost of Debt

Cost of Capital

Firm Value

0 10.50 % 4.80 % 10.50 % TZS2,74710 % 11.00 % 5.10 % 10.41 % TZS2,78020 % 11.60 % 5.40 % 10.36 % TZS2,79930 % 12.30 % 5.52 % 10.27 % TZS2,83540 % 13.10 % 5.70 % 10.14 % TZS2,88550 % 14.00 % 6.10 % 10.05 % TZS2,92260 % 15.00 % 7.20 % 10.32 % TZS2,81470 % 16.10 % 8.10 % 10.50 % TZS2,747

80 % 17.20 % 9.00 % 10.64 % TZS2,69690 % 18.40 % 10.20 % 11.02 % TZS2,569100 % 19.70 % 11.40 % 11.40 % TZS2,452

The value of the firm increases (decreases) as the WACC decreases (increases), as illustrated in Figure 4.4.

Figure 4.4: Cost of Capital and Firm Value as a Function of Leverage

This illustration makes the choice of an optimal financing mix seem trivial and it obscures some real problems that may arise in its applications. First, we typically do not have the benefit of having the entire schedule of costs of financing, prior to an analysis. In most cases, the only level of debt about which there is any certainty about the cost of financing is the current level. Second, the analysis assumes implicitly that the level of cash flows to the firm is unaffected by the financing mix of the firm and consequently by the default risk (or bond rating) for the firm. Although this may be reasonable in some cases, it might not in others. For instance, a firm that manufactures consumer durables (cars, televisions, etc.) might find that its sales and operating income drop if its default risk increases for the reason that investors are reluctant to buy its products. We will deal with the computational component of estimating costs of debt, equity and capital first in the standard cost of capital approach and then follow up by examining how to bring in changes in expected cash flows into the analysis in the enhanced cost of capital approach.The Standard Cost of Capital ApproachIn the standard cost of capital approach, we keep the operating income and cash flows fixed, while changing the cost of capital. Not surprisingly, the optimal debt ratio is the one that minimizes the cost of capital. While the assumptions seem heroic, it is a good starting point for the discussion.Steps in computing cost of capital

We need three basic inputs to compute the cost of capital – the cost of equity, the after-tax cost of debt, and the weights on debt and equity. The costs of equity and debt change as the debt ratio changes, and the primary challenge of this approach is in estimating each of these inputs. Let us begin with the cost of equity.

Cost of Equity = Risk-Free Rate + βlevered (Risk Premium)The cost of debt for a firm is a function of the firm’s default risk. As firms borrow more, their default risk will increase and so will the cost of debt. If we use bond ratings as the measure of default risk, we can estimate the cost of debt in three steps. First, we estimate a firm’s money debt and interest expenses at each debt ratio; as firms increase their debt ratio, both money debt and interest expenses will rise. Second, at each debt level, we compute a financial ratio or ratios that measure default risk and use the ratio(s) to estimate a rating for the firm; again, as firms borrow more, this rating will decline. Third, a default spread, based on the estimated rating, is added on to the risk-free rate to arrive at the pretax cost of debt. Applying the marginal tax rate to this pretax cost yields an after-tax cost of debt.Once we estimate the costs of equity and debt at each debt level, we weight them based on the proportions used of each to estimate the cost of capital. Although we have not explicitly allowed for a preferred stock component in this process, we can have preferred stock as part of capital. Nevertheless, we have to keep the preferred stock portion fixed while changing the weights on debt and equity. The debt ratio at which the cost of capital is minimized is the optimal debt ratio.In this approach, the effect of changing the capital structure, on firm value, is isolated by keeping the operating income fixed, and varying only the cost of capital. In practical terms, this requires us to make two assumptions. First, the debt ratio is decreased by raising new equity and retiring debt; conversely, the debt ratio is increased by borrowing money and buying back stock. This process is called recapitalization. Second, the pretax operating income is assumed to be unaffected by the firm’s financing mix and, by extension, its bond rating. If the operating income changes with a firm’s default risk, the basic analysis will not change, but minimizing the cost of capital may not be the optimal course of action, for the reason that the value of the firm is determined by both the cash flows and the cost of capital. The value of the firm will have to be computed at each debt level and the optimal debt ratio will be that which maximizes firm value.Sensitivity AnalysisThe optimal debt ratio we estimate for a firm is a function of all the inputs that go into the cost of capital computation – the beta of the firm, the risk-free rate, the risk premium, and the default spread. It is also indirectly a function of the firm’s operating income, for the reason that interest coverage ratios are based on this income, and these ratios are used to compute ratings and interest rates.The determinants of the optimal debt ratio for a firm can be divided into variables specific to the firm, and the macroeconomic variables. Among the variables specific to the firm that affect its optimal debt ratio are the tax rate, the firm’s capacity to generate operating income, and its cash flows. In general, the tax benefits from debt increases as the tax rate goes up. In relative terms, firms with higher tax rates will have higher optimal debt ratios than will firms with lower tax rates, other things being equal. It also follows that a firm’s optimal debt ratio will increase as its tax rate increases. Firms that generate higher operating income and cash flows as a percent of firm market value also can sustain much more debt as a proportion of the market value of the firm, for the reason that debt payments can be covered much more easily by prevailing cash flows.The macroeconomic determinants of optimal debt ratios include the level of interest rates and default spreads. As interest rates rise, the costs of debt and equity both increase. Nevertheless, optimal debt ratios tend to be lower when interest rates are higher, perhaps for the reason that interest coverage ratios drop at higher rates. The default spreads commanded by different ratings classes tend to increase during recessions and decrease during recoveries. Keeping other things constant, as the spreads

increase, optimal debt ratios decrease for the simple reason that higher default spreads result in higher costs of debt.How does sensitivity analysis allow a firm to choose an optimal debt ratio? After computing the optimal debt ratio with existing inputs, firms may put it to the test by changing both firm-specific inputs (such as operating income) and macroeconomic inputs (such as default spreads). The debt ratio the firm chooses as its optimal then reflects the volatility of the underlying variables and the risk aversion of the firm’s management.Enhanced Cost of Capital ApproachA key limitation of the standard cost of capital approach is that it keeps operating income fixed, while bond ratings vary. In effect, we are ignoring indirect bankruptcy costs, when computing the optimal debt ratio. In the enhanced cost of capital approach, we bring these indirect bankruptcy costs into the expected operating income. As the rating of the company declines, the operating income is adjusted to reflect the loss in operating income that will occur when customers, suppliers, and investors react.To quantify the distress costs, we have tie the operating income to a company’s bond rating. Put another way, we have to quantify how much we would expect the operating income to decline if a firm’s bond rating drops from AA to A or from A to BBB. This will clearly vary across sectors and across time.

Across sectors, the different effects of distress on operating income will reflect how much customers, suppliers and employees in that sector react to the perception of default risk in a company. For example, indirect bankruptcy costs are likely to be highest for firms that produce long-lived assets, where customers are dependent upon the firm for parts and service.

Across time, the indirect costs of distress will vary depending how easy it is to access financial markets and sell assets. In buoyant markets (in 1999 or 2006), the effect of a ratings downgrade on operating income are likely to be much smaller than in a market in crisis.

While getting agreement on these broad principles is easy, we are still faced with the practical question of how best to estimate the impact of declining ratings on operating income. We would suggest looking at the track record of other firms in the same sector that have been down graded by ratings agencies in the past, and the effects that the down grading has had on operating income in subsequent years.Once we link operating income to the bond rating, we can then modify the cost of capital approach to deliver the optimal debt ratio. Rather than look for the debt ratio that delivers the lowest cost of capital (the decision rule in the standard approach), we look for the debt ratio that delivers the highest firm value, through a combination of high earnings and low cost of capital.Extensions of Cost of Capital ApproachThe cost of capital approach, which works so well for manufacturing firms that are publicly traded, can be adapted to compute optimal debt ratios for cyclical firms, family group companies, private firms or even for financial service firms, such as banks and insurance companies.Cyclical and Commodity FirmsA key input that drives the optimal structure is the current operating income. If this income is depressed, either for the reason that the firm is a cyclical firm or for the reason that there are firm-specific factors that are expected to be temporary, the optimal debt ratio that will emerge from the analysis will be much lower than the firm’s true optimal. For example, automobile manufacturing firms will have very low debt ratios if the optimal debt ratios had been computed based on the operating income in 2008, which was a recession year for these firm, and oil companies would have had very high optimal debt ratios, with 2008 earnings, for the reason that high oil prices during the year inflated earnings.When evaluating a firm with depressed current operating income, we must first decide whether the drop in income is temporary or permanent. If the drop is temporary, we must estimate the normalized operating income for the firm, i.e., the income that the firm would generate in a normal year, rather than what it made in the most recent years. Most analysts normalize earnings by taking the average

earnings over a period of time (usually five years). For the reason that this holds the scale of the firm fixed, it may not be appropriate for firms that have changed in size over time. The right way to normalize income will vary across firms:

For cyclical firms, whose current operating income may be overstated (if the economy is booming) or understated (if the economy is in recession), the operating income can be estimated using the average operating margin over an entire economic cycle (usually 5 to 10 years).Normalized Operating Income = Average Operating Margin (Cycle) * Current Sales

For commodity firms, we can also estimate the normalized operating income by making an assumption about the normalized price of the commodity. With an oil company, for instance, this would translate into making a judgment about the normal oil price per barrel. This normalized commodity price can then be used, in conjunction with production, to generate normalized revenues and earnings.

For firms that have had a bad year in terms of operating income due to firm-specific factors such as the loss of a contract, the operating margin for the industry in which the firm operates can be used to calculate the normalized operating income:Normalized Operating Income = Average Operating Margin (Industry)*Current Sales

The normalized operating income can also be estimated using returns on capital across an economic cycle (for cyclical firms) or an industry (for firms with firm-specific problems), but returns on capital are much more likely to be skewed by mismeasurement of capital than operating margins.Companies that are part of a groupWhen a company is part of a family group, the logic of minimizing cost of capital does not change but the mechanics can be skewed by two factors.

The first is that the cost of debt may be more reflective of the credit standing of the group to which the firm belongs, rather than its own financial strength. Put another way, a distressed company that is part of a healthy family group of companies may be able to borrow more money at a lower rate than an otherwise similar stand-alone company. This can, at least artificially, increase its optimal debt ratio. Conversely, a healthy company that is part of distressed group may find its cost of debt and capital affected by perceptions about the group; in this case, the optimal debt ratio will be lower for this company than for an independent company.

The second is that rather than optimizing the mix of debt and equity for individual companies, the controllers of the family group of companies may view their objective as finding a mix of debt and equity that maximizes the value of the group of companies. Thus, our assessments of the capital structures of individual companies may not be particularly meaningful.

There is one final factor to consider. The consolidated operating income of the entire family group should be more stable than the earnings of the individual companies that comprise the group, reflecting diversification over multiple businesses. Consequently, the optimal debt computed for the family group will be higher than the aggregate of the optimal debt for individual companies in the group.Private FirmsThere are three major differences between public and private firms in terms of analyzing optimal debt ratios. One is that unlike the case of publicly traded firms, we do not have a direct estimate of the market value of a private firm. Consequently, we have to estimate firm value before we move to subsequent stages in the analysis. The second difference relates to the cost of equity and how we arrive at that cost. Although we use betas to estimate the cost of equity for a public firm, that usage might not be appropriate when we are computing the optimal debt ratio for a private firm, since the owner may not be well diversified. Finally, whereas publicly traded firms tend to think of their cost of debt in terms

of bond ratings and default spreads, private firms tend to borrow from banks. Banks assess default risk and charge the appropriate interest rates.To analyze the optimal debt ratio for a private firm, we make the following adjustments. First, we estimate the value of the private firm by looking at how publicly traded firms in the same business are priced by the market. Thus, if publicly traded firms in the business have market values that are roughly three times revenues, we would multiply the revenues of the private firm by this number to arrive at an estimated value. Second, we continue to estimate the costs of debt for a private firm using a synthetic bond rating, based on interest coverage ratios, but we will require much higher interest coverage ratios to arrive at the same thing, to reflect the fact that banks are likely to be more conservative in assessing default risk at small, private firms. Finally, we will use total betas to capture total risk, rather than just market risk, to estimate the cost of equity.

In practice: Optimal Debt Ratios for Private FirmsAlthough the trade-off between the costs and benefits of borrowing remain the same for private and publicly traded firms, there are differences between the two kinds of firms that may result in private firms borrowing less money.

Increasing debt increases default risk and expected bankruptcy costs much more substantially for small private firms than for larger publicly traded firms. This is partly for the reason that the owners of private firms may be exposed to unlimited liability, and partly for the reason that the perception of financial trouble on the part of customers and suppliers can be much more damaging to small, private firms.

Increasing debt yields a much smaller advantage in terms of disciplining managers in the case of privately run firms, for the reason that the owners of the firm tend to be the top managers as well.

Increasing debt generally exposes small private firms to far more restrictive bond covenants and higher agency costs than it does large publicly traded firms.

The loss of flexibility associated with using excess debt capacity is likely to weigh much more heavily on small, private firms than on large, publicly traded firms, due to the former’s lack of access to public markets.

All these factors would lead us to expect much lower debt ratios at small private firms.

Financial Service FirmsThere are several problems in applying the cost of capital approach to financial service firms, such as banks and insurance companies. The first is that the interest coverage ratio spreads, which are critical in determining the bond ratings, have to be estimated separately for financial service firms; applying manufacturing company spreads will result in absurdly low ratings for even the safest banks and very low optimal debt ratios. Furthermore, the relationship between interest coverage ratios and ratings tend to be much weaker for financial service firms than it is for manufacturing firms. The second is a measurement problem that arises partly from the difficulty in estimating the debt on a financial service company’s balance sheet. Given the mix of deposits, repurchase agreements, short-term financing, and other liabilities that may appear on a financial service firm’s balance sheet, one solution is to focus only on long-term debt, defined tightly, and to use interest coverage ratios defined using only long-term interest expenses. The third problem is that financial service firms are regulated and have to meet capital ratios that are defined in terms of book value. If, in the process of moving to an optimal market value debt ratio, these firms violate the book capital ratios, they could put themselves in jeopardy.While we could try to adapt the cost of capital approach to come up with optimal debt ratios for banks and other financial service companies, the results are very sensitive to how we define debt and the relationship we assume between bond ratings and operating income. An alternative and more effective approach is to use the regulatory capital ratios, usually determined in terms of book equity, as the basis

of regulatory capital ratios, usually determined in terms of book equity, as the basis of determining how much equity a financial service firm needs to raise to not only continue operating, but to do so without putting itself at peril. As a simple example, consider a bank with TZS 100 million in loans outstanding and a book value of equity of TZS6 million. Furthermore, assume that the regulatory requirement is that equity capital be maintained at 5 percent loans outstanding. Finally, assume that this bank wants to increase its loan base by TZS50 million to TZS 150 million and to augment its equity capital ratio to 7 percent of loans outstanding. The amount of equity that the bank will have to raise to fund its expansion is computed below:Loans outstanding after Expansion = TZS150 millionEquity/Capital ratio desired = 7 percentEquity after expansion = TZS10.5 millionExisting Equity = TZS6.0 millionNew Equity needed = TZS4.5 millionAs we look at more complex financial service firms that operate in multiple businesses with different risk levels, there are two challenges that we will face in putting this approach into practice:

a. Different regulatory capital requirements for different businesses: When a firm operates in different businesses, the regulatory capital restrictions can vary across businesses. In general, the capital requirements will be higher in riskier businesses and lower in safer businesses. Hence, the equity that a firm has to raise to fund expansion will depend in large part of which business are being expanded.

b. Regulatory vs Risk-based Capital Ratios: The regulatory ratios represent a floor on what a firm has to invest in equity, to keep its operations going and not a ceiling. It is possible that the firm’s own assessment of risk in a business can lead it to hold more equity than required by the regulatory authorities.As a final twist, it is worth nothing that banking regulators consider preferred stock as part of equity, when computing regulatory ratios. In general, there are three strategies that a financial service firm can follow when it comes to the use of leverage:

The regulatory minimum strategy: In this strategy, financial service firms try to stay with the bare minimum equity capital, as required by the regulatory ratios. In the most aggressive versions of this strategy, firms exploit loopholes in the regulatory framework to invest in those businesses where regulatory capital ratios are set too low (relative to the risk of these businesses). The upside of this strategy is that the returns on equity in good times will exceptionally high, since the equity capital is kept low. The downside of this strategy is that the risk in the investment ultimately will manifest itself and the absence of equity to cover losses will put the firm’s existence in jeopardy.

The Self-regulatory strategy: The objective for a bank raising equity is not to meet regulatory capital ratios but to ensure that losses from the business can be covered by the existing equity. In effect, financial service firms can assess how much equity they need to hold by evaluating the riskiness of their businesses and the potential for losses. Having done so, they can then check to also make sure that they meet the regulatory requirements for capital. The upside of this strategy is that it forces the firm to both assess risk in its businesses and to make the trade off between risk and return, when entering new business. The downside is that it is more data intensive, and errors in assessing risk will affect the firm’s value.

Combination strategy: In this strategy, the regulatory capital ratios operate as a floor for established business, with the firm adding buffers for safety where needed. In new or evolving businesses, the firm makes its own assessments of risk that may be very different from those made by the regulatory authorities.

We would argue that the responsibility for maintaining enough equity has to rest ultimately with the management of the firm and not with the regulatory authorities. A bank that blames the laxness of regulatory oversight for its failures is not a well-managed bank.Determinants of Optimal Debt RatioThe preceding analysis highlights some of the determinants of the optimal debt ratio. We can then divide these determinants into firm-specific and macroeconomic factors.Firm-Specific FactorsThe optimal debt ratios that we compute will vary across firms. There are three firm specific factors that contribute to these differences – the tax rate of the firm, its capacity to generate cash flows to cover debt payments and uncertainty about future income.

Firm’s Tax Rate: In general, the tax benefits from debt increase as the tax rate goes up. In relative terms, firms with higher tax rates will have higher optimal debt ratios than do firms with lower tax rates, other things being equal. It also follows that a firm’s optimal debt ratio will increase as its tax rate increases.

Pretax Returns on the Firm (in Cash Flow Terms): The most significant determinant of the optimal debt ratio is a firm’s earnings capacity. In fact, the operating income as a percentage of the market value of the firm (debt plus equity) is usually good indicator of the optimal debt ratio. When this number is high (low), the optimal debt ratio will also be high (low). A firm with higher pretax earnings can sustain much more debt as a proportions of the market value of the firm, for the reason that debt payments can be met much more easily from prevailing earnings.

Variance in Operating Income: The variance in operating income enters the base case analysis in two ways. First, it plays a role in determining the current beta: Firms with high (low) variance in operating income tend to have high (low) betas. Second, the volatility in operating income can be one of the factors determining bond ratings at different levels of debt: Ratings drop off much more dramatically for higher variance firms as debt levels are increased. It follows that firms with higher (lower) variance in operating income will have lower (higher) optimal debt ratios. The variance in operating also pays a role in the constrained analysis, for the reason that higher-variance firms are much more likely to register significant drops in operating income. Consequently, the decision to increase debt should be made much more cautiously for these firms.

Macroeconomic Factors Should macroeconomic conditions affect optimal debt ratios? In purely mechanical terms, the answer is yes. In good economic times, firms will generate higher earnings and be able to service more debt. In recessions, earnings will decline and with it the capacity to service debt. That is why prudent firms borrow based on normalized earnings rather than current earnings. Holding operating income constant, macroeconomic variables can still affect optimal debt ratios. In fact, both the level of risk-free rate and the magnitude of default spreads can affect optimal debt ratios.

Level of Rates: As interest rates decline, the conventional wisdom is that debt should become cheaper and more attractive for firms. Though this may seem intuitive, the effect is muted by the fact that lower interest rates also reduce the cost of equity. In fact, changing the risk-free rate has a surprisingly small effect on the optimal debt ratio as long as interest rates move within a normal range. When interest rates exceed normal levels, optimal debt ratios do decline partly for the reason that we keep operating income fixed. The higher interest payments at every debt ratio reduce bond ratings and affect the capacity of firms to borrow more.

Default Spreads: The default spreads for different ratings classes tend to increase during recessions and decrease during economic booms. Keeping other things constant, as the spreads increase (decrease) optimal debt ratios decrease (increase), for the simple reason that higher spreads penalize firms that borrow more money and have lower ratings. In fact, the default

spreads on corporate bonds declined between 2002 and 2007, leading to higher optimal debt ratios for all firms. In 2008, as the economy slowed and the market entered crisis mode, default spreads widened again, leading to lower optimal debt ratios.There is another factor to consider. The same factors that cause default spreads to increase and decrease also play a role in determining equity risk premiums. Hence, the question of how much changing default spreads affect optimal debt ratios cannot be answered without looking at how much equity risk premiums also change. If equity risk premiums increase more than default spreads do, debt will become a more attractive choice relatively to equity.Adjusted Present Value ApproachIn the adjusted present value (APV) approach, we begin with the value of the firm without debt. As we add debt to the firm, we consider the net effect on value by considering both the benefits and the costs of borrowing. The value of the levered firm can then be estimated at different levels of the debt, and the debt level that maximizes firm value is the optimal debt ratio.Steps in the APV ApproachIn the APV approach, we assume that the primary benefit of borrowing is a tax benefit and that the most significant cost of borrowing is the added risk of bankruptcy. To estimate the value of the firm with these assumptions, we proceed in three steps. We begin by estimating the value of the firm with no leverage. We then consider the present value of the interest tax savings generated by borrowing a given amount money. Finally, we evaluate the effect of borrowing the amount on the probability that the firm will go bankrupt and the expected cost of bankruptcy.Step 1: Estimate the value of the firm with no debt: The first step in this approach is the estimation of the value of the unlevered firm. This can be accomplished by valuing the firm as if it had no debt, that is, by discounting the expected after-tax operating cash flows at the unlevered cost of equity. In the special case where cash flows grow at a constant rate in perpetuity,

where FCFF1 is the expected after-tax operating cash flow to the firm in the next period, ρu is the unlevered cost of equity, and g is the expected growth rate. The inputs needed for this valuation are the expected cash flows, growth rates, and the unlevered cost of equity. To estimate the latter, we can draw on our earlier analysis and compute the unlevered beta of the firm:

Where βunlevered = unlevered beta of the firm, βcurrent = current equity beta of the firm, t = tax rate for the firm, and D/E = current debt/equity ratio. This unlevered beta can then be used to arrive at the unlevered cost of equity. Alternatively, we can take the current market value of the firm as a given and back out the value of the unlevered firm by subtracting out the tax benefits and adding back the expected bankruptcy cost from the existing debt.Current Firm Value = Value of Unlevered firm + PV of Tax Benefits – Expected Bankruptcy CostsValue of Unlevered Firm = Current Firm Value – PV of Tax Benefits + Expected Bankruptcy Costs

Step 2: Estimate the present value of tax benefits from debt: The second step in this approach is the calculation of the expected tax benefit from a given level of debt. This tax benefit is a function of the tax of the firm and is discounted at the cost of debt to reflect the riskiness of this cash flow. If the tax savings are viewed as a perpetuity.Value of Tax Benefits = [Tax Rate * Costs of Debt * Debt]/Cost of Debt

= Tax Rate * Debt

= tcDThe tax rate used here is the firm’s marginal tax rate, and it is assumed to stay constant over time. If we anticipate the tax rate changing over time, we can still compute the present value of tax benefits over time, but we cannot use the perpetual growth equation. Step 3: Estimate the expected bankruptcy costs as a result of the debt: The third step is to evaluate the effect of he given level of debt is to evaluate the effect of the given level of debt on the default risk of the firm and on expected bankruptcy costs. In theory, at least, this requires the estimation of the probability of default with the additional debt and the direct and indirect cost of bankruptcy. If π a is the probability of default after the additional debt and BC is the present value of the bankruptcy cost the present value of the expected bankruptcy cost can be estimated. —

= πaBCThis step of the APV approach poses the most significant estimation problem, for the reason that neither the probability of bankruptcy nor the bankruptcy cost can be estimated directly. There are two ways the probability of bankruptcy can be estimated indirectly. One is to estimate a bond rating, as we did in the cost of capital approach, at each level of debt and use the empirical estimates of default probabilities for each rating. For instance, Table 4.13, extracted from an annually updated study by Altman, summarizes the probability of default over ten years by bond rating class.

Table 4.13: Default Rates by Bond Rating ClassesRating Likelihood of DefaultAAA 0.07 %AA 0.51 %A+ 0.60 %A 0.66 %

A– 2.50 %tBBB 7.54 %BB 16.63 %B+ 25.00 %B 36.80 %

B– 45.00 %CCC 59.01 %CC 70.00 %C 85.00 %D 100.00 %

The other is to use a statistical approach, such as a profit to estimate the probability of default, based on the firm’s observable characteristics, at each level of debt.The bankruptcy cost can be estimated, albeit with considerable error, from studies that have looked at the magnitude of this cost in actual bankruptcies. Studies that have looked at the direct cost of bankruptcy conclude that they are small relative to firm value. The indirect costs of bankruptcy can be substantial, but the costs vary widely across firms. Shapiro and Titman speculate that the indirect costs could be as large as 25 to 30 percent of firm value but provide no direct evidence of the costs.The net effect of adding debt can be calculated by aggregating the cost and the benefits at each level of debt.

Value of Levered Firm = FCFF1/(ρu – g) + tcD – πaBCWe compute the value of the levered firm at different levels of debt. The debt level that maximizes the value of the levered firm is the optimal debt ratio.Benefits and Limitations of the APV Approach

The advantage of the APV approach is that it separates the effects of debt into different components and allows an analyst to use different discount rates for each component. In this approach, we do not assume that the debt ratio stays unchanged forever, which is an implicit assumption in the cost of capital approach. Instead, we have the flexibility to keep the money value of debt fixed and to calculate the benefits and costs of the fixed money debt.These advantages have to be weighed against the difficulty of estimating probabilities of default and the cost of bankruptcy. In fact, many analyses that use the APV approach ignore the expected bankruptcy costs, leading them to the conclusion that firm value increases as firms borrow money. Not surprisingly, they conclude that the optimal debt ratio for a firm is 100 percent debt.In general, with the same assumptions, the APV and the cost of capital conclusions give identical answers. Nevertheless, the APV approach is more practical when firms are evaluating the feasibility of adding an amount of debt, whereas the cost of capital approach is easier when firms are analyzing debt proportions.Comparative AnalysisThe most common approach to analyzing the debt ratio of a firm is to compare its leverage to that of similar firms. A simple way to perform this analysis is to compare a firm’s debt ratio to the average debt ratio for the industry in which the firm operates. A more complete analysis would consider the differences between a firm and the rest of the industry, when determining debt ratios. We will consider both ways below.Comparing to Industry AverageFirms sometimes choose their financing mixes by looking at the average debt ratio of other firms in the industry in which they operate. The underlying assumptions in this comparison are that firms within the same industry are comparable and that, on average, these firms are operating at or close to their optimal. Both assumptions can be questioned, nevertheless. Firms within the same industry can have different product mixes, different amounts of operating risk, different tax rates, and different project returns. In fact, most do. Controlling for Differences between FirmsFirms within the same industry can exhibit wide differences on tax rates, capacity to generate operating income and cash flows, and variance in operating income. Consequently, it can be dangerous to compare a firm’s debt ratio to the industry and draw conclusions about the optimal financing mix. The simplest way to control for differences across firms, while using the maximum information available in the market, is to run a regression, regressing debt ratios against these variables, across the firms in an industry:

Debt Ratio = α0 + α1 Tax Rate + α2 Pretax Returns + α3 Variance in Operating IncomeThere are several advantages to the cross-sectional approach. Once the regression has been run and the basic relationship established (i.e., the intercept and coefficients have been estimated), the predicted debt ratio for any firm can be computed quickly using the measures of the independent variables for this firm. If a task involves calculating the optimal debt ratio for a large number of firms in a short time period, this may be the only practical way of approaching the problem, for the reason that the other approaches described in this chapter are time-intensive.There are also limitations to this approach. The coefficients tend to shift over time. Besides some standard statistical problems and errors in measuring the variables, these regressions also tend to explain only a portion of the differences in debt ratios between firms. Nevertheless, the regressions provide significantly more information than a naive comparison of a firm’s debt ratio to the industry average.

EMPIRICAL EVIDENCE OF CAPITAL STRUCTURE

The “capital structure puzzle” i.e., the firm’s choice of an optimal capital structure, remains one of the large unresolved issues in the financial economics literature. The origin for five decades of capital structure research is provided by Modigliani and Miller (1958). They argue that in a neoclassical world capital structure decisions are irrelevant for the market value of a firm (MM theory). Nevertheless, the MM-theory builds on the restrictive assumptions of perfect and complete capital markets with rational investors.Consequently, further research subsequently enhanced the field of capital structure theory by accounting for several market imperfections. Modigiliani and Miller (1963) themselves started to extend their MM-theory by introducing taxes and costs of financial distress. The static “trade-of theory” of capital structure assumes the existence of a target debt ratio where the marginal cost of an additional unit of debt, i.e., the costs of financial distress, equal the marginal benefits of an additional unit of debt, i.e., the tax shield. In other words, according to this theory management aims to establish an optimal capital structure which is determined by a trade-off between the costs and benefits of borrowing debt. In contrast, the second major capital structure theory is based on a dynamic perspective of investment opportunities and information asymmetries. The “pecking order theory” of capital structure assumes that firms prefer to finance growth opportunities with internal funds, debt, preferred equity and common equity, in that order. Behind the pecking order theory is the rationale that information asymmetries between the informed firm insiders and uninformed outside investors lead to a mispricing of equity issues. Hence, the decisions of the management are driven by the desire to minimize transaction costs. Hence, the decisions of the management are driven by the desire to minimize transaction costs. Despite the dominance of those two paradigms in the discussion of capital structure theory, several authors have proposed further determinants of capital structure decisions: Among them are signaling aspects, corporate control considerations, market timing issues and firm history.In this discussion, we focus on another major determinant for capital structure decisions: agency costs. In particular, we compare two distinct groups of firms that are considered to be unequal in terms of agency costs: family firms and non-family firms. Based on the widespread assumption that agency costs are lower in family firms, we would expect that there is less need for the disciplining role of debt in family firms and that they hence have lower leverage ratios. Nevertheless, existing economies, such as the United States. While Mishra and McConaughy (1999) conclude that U.S. family firms use less debt to avoid a loss of control and decrease the likelihood of bankruptcy, Anderson and Reeb (2003b) find no systematic difference between U.S. family and non-family firms in terms of leverage. Finally, just recently Ellul (2008) finds a positive relationship between leverage and family blockholdings based on a cross-country analysis. Nevertheless, to our best knowledge detailed empirical evidence from a bank-based economy is missing so far. The importance of the institutional setting for capital structure decisions is stressed by Antoniou et al. (2008) who argue that “the capital structure of a firm is heavily influenced by the economic environment and its institutions, corporate governance practices, tax systems, the borrower-lender relation, exposure to capital markets, and the level of investor protection in the country in which the firm operates.”This discussion builds on this research gap and aims to shed more light on hitherto conflicting empirical results by analyzing capital structure decisions in family firms within a bank-based economy in greater detail. The country of our choice – Tanzania – seems to provide a very fruitful research environment since it is characterized by the following stylized facts: (i) a different legal and institutional setting and underdeveloped stock markets in comparison to anglo-saxon countries, (ii) a bank-orientated financial system with widespread relationship lending, (iii) a tradition where family firms are considered to be the backbone of the economy and (iv) still concentrated ownership patterns with a large amount of family firms even among listed companies.Starting from these conflicting observations, it is by far not clear whether family firms in Tanzania use more or less dent and what factors drive their capital structure decisions. Moreover, it is an open

question how the different components of a family firm, namely founding family ownership, supervisory and management board participation, affect those capital structure decisions. The focus of our discussion is to analyse these in greater detail. Thereby, we contribute to the literature in several important dimensions: First, to our best knowledge our empirical study is the first to analyse capital structure decisions of family firms for a bank-based economy. This is interesting against the background that Tanzania differs largely from the U.S. in terms of institutional setting. We complement recent albeit inconsistent empirical evidence on capital structure decisions of family firms that is so far largely focusing on the U.S. Second, in terms of methodology our analysis is more advanced that previous research on capital structure decisions within family firms. We do not only exploit cross-sectional heterogeneity with “between” estimates and pooled OLS regressions but also time variation based on “within” estimates. Those firm fixed effects models allow us to control for unobserved firm heterogeneity. Finally, we employ a battery of robustness tests including a matching estimator that allows us to control for issues of endogeneity. Our results are highly robust and not subject to any special kind of methodology. Third, we investigate an aspect that goes beyond existing research on family firms: We carefully analyse the impact of different family firm characteristics on capital structure decisions. In particular, we distinguish between three separate components of a family firm, i.e., family ownership, family supervisory and management board participation. In fact, that distinction allows us to show that firm leverage heavily depends on agency costs. From these three aspects the convergence-of-interest effect of family management seems to have the strongest (negative) influence on leverage ratios. In the case of combined family ownership and management board participation, in other words if the founding family is a large shareholder with monitoring incentives and simultaneously involved in firm management with convergence-of-interest effects between management and outside shareholders, agency costs and hence firm leverage are the lowest. This outcome underlines the importance of agency costs theory in family firm research. Moreover, in accordance with several previous performance studies we detect a significant impact of founder CEO on capital structure decisions. Our fourth contribution is related to the analysis of capital structure in Tanzania. In general empirical evidence on this topic based on a large sample of listed firms in Tanzania is rather limited. Classification of family firmsFollowing the extant body of literature on listed family firms, our definition of a family firm is based on two components: the ownership and management component. In particular, we classify a firm within our sample as a family firm if at least one of the following three conditions is satisfied: (i) the founding family has voting rights of at least 25 percent (family ownership) and/or (ii) at least one member of the founding family is represented in the supervisory board (supervisory board participation) (iii) at least one member of the founding family is involved in top management (management board participation). The fact that we use 25 percent of voting rights as ownership threshold is related to the typically more concentrated ownership structures in continental Europe. Moreover, 25 percent compromises an important control threshold according to the German stock corporation act.13 Of course, if a company was founded by a team of entrepreneurs, the term founding family might refer to more than one family. Based on this definition we have created a dummy variable called Family Firm which is one if the firm qualifies as a family business according to our definition and zero otherwise. Overall, our sample consists of 660 firms and 5,135 firm year observations: 2,410 family firm year observations and 2,725 non-family firm year observations. For an overview of the sample composition over time cf. table 4.14.

Table 14.4: Composition of SampleYear Firms Family

FirmsNon-Family Firms

1995 230 65 1651996 235 68 167

1997 250 75 1751998 312 111 2011999 430 203 2272000 566 312 2542001 568 315 2532002 542 278 2642003 514 262 2522004 500 248 2522005 494 237 2572006 494 236 258

Note: This table shows the development of the sample composition over time. Column 1 presents the 12 sample years between 1995 and 2006, column 2 the number of firms in each year and column 3 and 4 the number of family and non-family firms in each year.

In a second step, we test whether differences in capital structure are driven by family ownership or family management. Therefore, we substitute the dummy variable family firm in all our regression models by three variables: (i) family ownership, (ii) supervisory board participation and (iii) management board participation. Family ownership is the cumulated ownership fraction of the founding family. Supervisory board participation is a dummy variable with unit value one if a member of the founding family is part of the supervisory board. The same dummy variable is constructed for management board participation. Finally we run all regressions with two other specifications: First, we analyse the impact of the firm’s CEO. Several other studies, e.g. Anderson and Reeb (2003a) or Villalonga and Amit (2006), argue that the identity of the CEO is of special importance for the firm’s corporate policy and performance. Following those studies, we construct a dummy variable called founder CEO that takes unit value one if the founder is still running the firm as CEO and zero otherwise. Second, we interact family ownership and management participation. The interaction term allows us to investigate the hypothesis that the effect of agency costs on capital structure is strongest if the founding family is simultaneously a large shareholder and involved in running the firm’s daily business.Measurement of leverageWe measure leverage in several ways:

(i) We start with a broad definition of book and market leverage. Book leverage is the ratio of total liabilities to total assets while the market leverage is the ratio of total liabilities to the market value of equity plus total liabilities. Thereby, we treat preferred equity as equity rather than debt. By applying such a broad definition of leverage we follow several other studies on capital structure. Moreover, just recently Elsas and Florysiak (2008) have applied similar definitions of leverage for a large sample study of capital structure in the German environment. This broad definition includes non-interest-bearing debt components, such as pension liabilities or accounts payable, and is likely to overestimate financial leverage. Although such a definition is not a very good indication for the future default probability, for many firms those non-interest-bearing debt components are important parts of their capital structure.

(ii) We run all regression models with a definition of long-term leverage. Long-term book leverage is defined as total liabilities minus current liabilities divided by total assets. Accordingly, long-term market leverage is defined as total liabilities minus current liabilities to market value of equity plus total liabilities.

(iii) Finally, we calculate a financial leverage that only considers interesting-bearing debt components. Our measure for the book value of financial leverage is calculated as total liabilities minus the sum of accounts payable, provisions for risks and charges (including

pension liabilities) and deferred taxes divided by total assets minus the sum of accounts payable, provisions for risks and charges (including pension liabilities) and deferred taxes. As in the two other measures of leverage, we replace the book value of equity with the market value of equity when we calculate the market value of financial leverage.

Definition of control variablesIn our analysis, we use a set of control variables (for a detailed overview of all variables cf. table 4.15). Frank and Goyal (forthcoming) show that there are six core factors that can explain firm leverage for publicly traded American companies over the period 1950 to 2003: Firm size, profitability, market-to-book ratio, tangible assets ratio, median industry leverage and expected inflation. We include all these factors, which are described below, in our analysis.Since we already use total assets to scale our dependent variable, we use the natural logarithm of the number of employees to control for firm size (FIRM SIZE). Firm size is included in all specifications to account for the fact that larger firms have a higher creditworthiness, easier access to debt markets and might be able to borrow at lower costs. Moreover, larger firms might use their financing-mix to maximize tax benefits. Overall, we anticipate a positive relation between firm size and leverage. Family firms might experience lower agency costs of free cash flow and depend more on internal financing. We use an operating profit margin calculated as earnings before interest, taxes, depreciation and amortization divided by total assets (PROFITABILITY) as a proxy for firm profitability. The pecking order theory suggests that firms prefer to finance new investment projects with retained earnings followed by new debt while issuing external equity is only the last resort of financing. Consequently, we expect an inverse relationship between the firm profitability and the leverage ratio. We control for the firm’s growth options by including the market-to-book ratio (MARKET-TO-BOOK) into our regressions. Information asymmetries may lead firms to issue equity instead of debt if they have NPV-positive projects (Myers 1977). Furthermore, firms may prefer to retain earnings instead of distributing them if they have valuable growth options. Hence, we expect market-to-book ratio to be negatively related to leverage. We include the ratio of tangible assets to totals assets (TANGIBLE ASSETS RATIO) in our analysis to account for the fact that tangible assets function as collateral and hence increase borrowing capacity. We expect the tangibility ratio to be positively correlated to the firm’s leverage.The median industry leverage (MEDIAN INDUSTRY LEVERAGE) is included as a control for industry originalities. Firms operating in highly levered industries are expected to exhibit higher leverage ratios. For example, Frank and Goyal (forthcoming) show that the industry median leverage ratio has the single largest explanatory power for the firm-level leverage in their long-term dataset on U.S. firms. Although we use industry dummies to control for industry effects in general, we therefore include industry median leverage in our regressions as an additional control variable. This measure is calculated for each industry and year, whereby the firm’s industry classification is based on its one-digit primary SIC-Code. Of course, we expect industry leverage to have a positive impact on firm leverage. The expected inflation rate (EXPECTED INFLATION RATE) is another variable with high explanatory power for leverage ratios. We anticipate firms to show higher levels of leverage if the expected inflation rate is high since debt becomes more attractive in these time periods. In our analysis, we use the next year’s realised inflation rate as a proxy for the expected inflation rate. In order to investigate if this adoption leads to biased results, we applied the one-year inflation rate forecast of the German “Sachverständigenrat” as an alternative measure of expected inflation (results not reported). Nevertheless, the results for these two measures are qualitatively the same. Besides these control factors proposed by Frank and Goyal (2009) we include several additional variables in our regressions, which are described below.The dividend payout ratio (PAYOUT RATIO) is likely to play an important role for the firm’s financing mix. For example, Rozeff (1982) predicts an inverse relationship between dividend payout and leverage due to agency costs and transaction costs arguments. Also if dividends are considered to be a signal for future earnings, firms with high dividend payout ratios face lower cost of equity. They might prefer

equity instead of debt and hence we expect a negative relationship between the firm’s dividend payout ratio and the firm’s leverage.Nevertheless, we adopt the payout ratio as suggested by Julio and Ikenberry (2004) and von Eije and Megginson (2008): We set the payout ratio to 1 if it is negative (for the reason that of negative income) or above one. Firm age (FIRM AGE) is the natural logarithm of the number of years since the firm’s incorporation. Thereby, the number of years since the firm’s incorporation is calculated as the current sample year minus the founding year of the firm. We expect younger firms ceteris paribus to have better growth options than older firms. Younger firms might be more reliable on equity instead of debt and prefer to retain earnings within the firm to finance their growth options. Simultaneously we hypothesize that older firms have more tangible assets, a better borrowing capacity and are more profitable. Hence, the expected relationship between firm age and leverage is positive.One potential concern is that founding family ownership is not randomly assigned to different industries. In particular, instead of applying risk-reducing strategies at the firm level, founder families might prefer to invest in low-risk businesses and industries. Consequently, we include a measure of firm-specific risk (FIRM SPECIFIC RISK). Firm-specific risk captures the part of stock prize volatility that is unique to an individual firm and thus related to specific operations or capital structure decisions. It is calculated as the residuals’ sum of squares (SSE) from a regression of the individual stock returns on the returns of the market (CDAX) over the preceding calendar year based on stock prizes from calendar year end.Since higher debt-to-equity ratios increase the firm’s risk of default, we expect a positive relationship between firm specific risk and leverage. Decisions about capital structure are dependent on the firm’s governance structure. Consequently, we include some corporate governance measures in our analysis. Monitoring by outside shareholders might be an alternative to incentive alignment as a corporate governance device in order to alleviate the classical shareholder-manager conflict. Hence, we include the cumulative corporate ownership of large outside shareholders with an ownership stake of at least 5 percent in our analysis (OUTSIDE BLOCKHOLDERS). Alternatively, as indicated in our section about robustness tests, we control for the presence of a second large shareholder besides the founding family. In Germany, the sample period 1995 to 2006 is characterised by a huge heterogeneity in terms of applied accounting standards. This is due to the introduction of the capital raising facilitating act (Kapitalaufnahmeerleichterungsgesetz – KapAEG) in 1998. According to this law, all listed German consolidated companies have the possibility to prepare annual consolidated financial statements either in IFRS/IAS (International Financial Reporting Standards or respectively International Accounting Standards, henceforth IFRS) or USGAAP. Simultaneously they face no requirement to prepare additional annual consolidated (not individual) financial statement in German GAAP if they apply IFRS or US-GAAP. From 2005 onwards, the usage of IFRS is mandatory for consolidated companies according to § 315a German GAAP. Hence, all firms change the applied accounting standard during the sample period. Since the valuation of assets and liabilities is largely dependent on the application of a true-and-fair-view accounting system (such as IFRS or US-GAAP) or a conservative accounting system (such as German GAAP),18 we control for accounting systems with a dummy variable for the application of German GAAP (DUMMY ACCOUNTING STANDARD). The dummy variable takes unit value one if the firm applies German GAAP and is zero if the firm applies either IFRS or US-GAAP. Due to the principle of prudence in German GAAP we expect a positive relationship between the usage of a conservative accounting system and the leverage ratio.Theory predicts that mature industries with less opportunity for asset substitution (Jensen and Meckling 1976) have higher leverage ratios. Hence, we use industry dummies based on one-digit SIC codes in all our regressions to control for such industry specifics. Capital structure decisions might be subject to macroeconomic and legal conditions. To control for such time effects we include year dummies in all our analysis.

Table 4.15: Definition of VariablesVariable group Variable name Description of variable

Corporate Governance Variable Dummy Family Firm Dummy which is one if (a) the cumulative ownership stake of the founding family is at least 25 percent and/or (b) a member of the founding family is represented in either the management or supervisory board

Founding Family Ownership Percentage of stock ownership held by all members of the founding family

Family Management [MB] Equals 1 if a member of the founding family is involved in the management board

Family Management [SB] Equals 1 if a member of the founding family is involved in the management board

Founder CEO Equals 1 if the CEO is the founder of the firm

Ownership*Management [MB] Interaction term of founding family ownership and family management [MB]

Control Variables Firm Size [Ln] Ln of the firm’s number of employees

Profitability Earnings before interest, taxes, depreciation and amortization (EBITDA)/total assets

Outside Blockholders Stock ownership of outside block owners (which have an ownership stake of at least 5 percent)

Firm specific Risk Residuals’ sum of squares from a regression of the individual stock returns on the returns of the market

Firm Age [Ln] Ln of the number of years since the firm’s incorporation

Tangible Assets Ratio Tangible assets/Total assetsMarket-to-Book Market value of the firm/book

value of the firmDummy Accounting Standard Equals 1 if the firm applies IFRS

or GAAP according to the accounting standard of the concerned country

Payout Ratio Common dividends/net income

available to common equity; Equals 1 if calculated payout ratio is below 0 or above 1.

Median Industry Leverage Median leverage in the firm’s industry indicated by the first number of the SIC code for each year (Leverage is defined as for the dependent variable in each model)

Expected inflation Rate Inflation rate of the following year

Dummy High-Tech Firm Equals 1 if the firm went public during 1998 and 2000

Capital Structure Variables Book Leverage Total liabilities/(Book value of equity + total liabilities)

Market Leverage Total liabilities/(Market value of equity + total liabilities)

Long-term Book Leverage (Total liabilities – current liabilities)/(Book value of equity + total liabilities)

Long-term Market Leverage (Total liabilities – current liabilities)/(Market value of equity + total liabilities)

Financial Book Leverage (Total liabilities – provisions – accounts payable – deferred taxes)/ (Book value of equity + total liabilities – provisions – accounts payable – deferred taxes)

Financial Market Leverage (Total liabilities – provisions – accounts payable – deferred taxes)/(market value of equity + total liabilities – provisions – accounts payable – deferred taxes)

Descriptive statisticsTable 14.6 presents descriptive statistics for all sample companies as well as the two subgroups family and non-family firms. As the t-test of differences in means indicates there seem to be huge differences among these two subgroups of firms. Family firms are smaller (in assets, sales and employees) and as a result have smaller management and supervisory boards. They are younger both in terms of years since incorporation and years since the Initial Public Offering. For example, family firms are on average 31 years old, in comparison to an average age of 72 years for non-family firms. Furthermore, several differences in accounting based figures or accounting standard can be found.Since our study focuses on differences in leverage, it is very interesting that the descriptive statistics indicate that family firms have lower levels of leverage than their nonfamily counterparts. For example, the mean (median) book leverage is 0.49 (0.5) for family firms in comparison to 0.62 (0.66) for non-

family firms. Similar differences occur for market leverage with 0.36 (0.39) for family firms in comparison to 0.54 (0.53) for non-family firms. Statistically significant differences in similar magnitude do also occur for long-term leverage and financial leverage indicating that there are large differences in terms of capital structure between the two firm groups.

Table 4.16: Descriptive StatisticsCorporate Governance Aspects

All Firms Family Firms Non-Family Firms

Mean Median Mean Median Mean Median t-testFounding Family Ownership [%]

17.90 0 37.71 40.05 0.63 0 29.25

Outside Blockholders [%]

33.73 20.3 15.23 5.50 50.0 51.0 –16.97

Size Management Board

3.16 3 2.94 3 3.34 3 –3.34

Size Supervisory Board

7.56 6 5.32 3 9.54 8 –11.58

Firm Size and ageAssets (in million TZS)

2,998.08 142.74 996.62 74.67 4,757.64 310.49 –3.30

Sales (in million TZS)

2,501.35 167.39 1,121.77 80.38 3,735.04 369.07 –3.24

Employees 11,373 1023 6,324 428 15,863 2159 –2.65Firm Age 52.97 28 31.18 15 72.42 74 –10.79IPO Age 14.62 6 5.91 4 22.38 11 –12.34Accounting figuresProbability –0.07 0.11 0.10 0.09 –0.27 0.11 –1.19Tangible Assets Ratio

0.23 0.20 0.20 0.15 0.26 0.24 –4.81

Market-to-Book

2.86 1.73 3.08 1.74 2.66 1.72 0.72

Dummy Accounting Standard

0.46 0.00 0.32 0.00 0.58 1.00 –8.85

Payout rate 0.27 0.00 0.22 0.00 0.32 0.19 –5.11Dependent VariablesBook Leverage

0.56 0.59 0.49 0.50 0.62 0.66 –0.27

Market 0.47 0.46 0.36 0.39 0.54 0.53 –8.32

LeverageLong-term Book Leverage

0.26 0.25 0.20 0.17 0.32 0.31 –9.53

Long-term Market Leverage

0.22 0.20 0.16 0.12 0.26 0.25 –9.59

Financial Book Leverage

0.48 0.48 0.43 0.41 0.51 0.53 –4.18

Financial Market Leverage

0.35 0.37 0.28 0.33 0.39 0.41 –3.88

Note: Accounting information is obtained from Thomson's Worldscope Database. Information on ownership structure is hand-collected from the Hoppenstedt Stock Guide. The sample consists of all non-financial firms in the broadest German stock Index (CDAX) between 1995 and 2006. Firms are classified as family firms if the founding family has an ownership stake of at least 25 percent and/or a member of the founding family participates in the management or supervisory board. ***, ** and * indicate significance on the 1 percent-, 5 percent- and 10 percent-level respectively. The t-statistics are corrected for serial correlation. A detailed definition of all variables can be found in table 4.15.

Empirical resultsMethodologyOur data structure is organised as an unbalanced panel of 660 firms that are tracked over the 1995 to 2006 period. The panel structure of our data allows us to present three types of regression estimates: pooled OLS, “between” estimates and “within” estimates. From an econometric point of view, all three estimates have advantages and disadvantages. “Between” estimates are OLS estimates of firm means across time. “Within” estimates are OLS estimates of deviations from the firm means across time (also called fixed effects model since they include firm-fixed effects). While the “between” estimates only employs cross-sectional variation, the “within”-estimates only uses variation over time within each section. The pooled OLS estimator combines both aspects as it is a weighted average of both the ”between” and ”within” estimators. By reporting all three models, we follow earlier work on capital structure by Berger et al. (1997) in terms of methodology and try to show that our results are robust against several different estimation techniques.Thereby, the fixed-effects estimator has one strong advantage: It offers the possibility to control for unobserved, time-invariant firm heterogeneity. A recent study by Lemmon et al. (2009) indicates that the adjusted R-squares of leverage regressions with firm fixed effects are much higher than the adjusted R-squares from traditional leverage regressions. Hence, such firm fixed effects seem to have a high explanatory power for capital structure decisions. Nevertheless, in our context the results of the fixed-effects estimator have to be interpreted with caution since the ownership and board structures among listed German firms are rather stable and thus offer little potential to exploit variation over time. As a consequence, the results for these estimations may be driven by variations in few firm-year observations. Consequently, it is useful to exploit cross-sectional variance by the “between” and pooled-OLS estimates as well. In addition, “between” estimates allow to mitigate concerns that observations drawn repeatedly from the same sample are independent from each other. Contrary to the “within” estimates, pooled OLS and “between” estimates may be biased if unobservable, time-invariant firm-specific factors exist, leading to a correlation of the error term with the independent variables. This happens if our models fail to include all relevant explanatory variables that are correlated with both the regressors and the dependent variable. Since no single model combines all advantages, we decide to report the estimates of all three models (OLS, “between” and “within” estimates).

In the context of panel datasets it is essential to estimate the standard errors in a correct way as indicated by Petersen (2009). As suggested by him, we calculate the standard errors in the pooled-OLS-specifications and the “within”-estimates using the cluster-robust VCE estimator (this is not necessary for the “between” estimates since there is only one observation per firm and hence no time-series correlation). Our calculation includes adjustment for non-i.i.d. distributed standard errors, resulting both from heteroskedasticity (Huber-White standard errors, cf. White 1980) and time-series correlation. Finally, we calculate variance inflation factors (VIFs) to detect any multicollinearity problem. Nevertheless, the calculated (not reported) VIFs indicate that our variables are not subject to any issues of multicollinearity.COST OF CAPITAL: APPLICATIONSThe cost of capital is the expected rate of return that the market participants require in order to attract funds to a particular investment. In economic terms, the cost of capital for a particular investment is an opportunity cost — the cost of forgoing the next best alternative investment. In this sense, it relates to the economic principle of substitution; that is, an investor will not invest in a particular asset if there is a more attractive substitute.The term market refers to the universe of investors who are reasonable candidates to fund a particular investment. Capital or funds are usually provided in the form of cash, although in some instances capital may be provided in the form of other assets. The cost of capital usually is expressed in percentage terms, that is, the annual amount of dollars that the investor requires or expects to realize, expressed as a percentage of the amount of money invested.Put another way:Since the cost of anything can be defined as the price one must pay to get it, the cost of capital is the return a company must promise in order to get capital from the market, either debt or equity. A company does not set its own cost of capital; it must go into the market to discover it. Yet meeting this cost is the financial market’s one basic yardstick for determining whether a company’s performance is adequate.As the quote suggests, most of the information for estimating the cost of capital for a business, security, or project comes from the investment market. The cost of capital is always an expected (or forward looking) return. Thus, analysis and would-be investors never actually observe the market’s view as to expected returns at the time of their investment. Nevertheless, we often form our views of the future by analyzing historical market data.As Roger Ibbotson put it, “The opportunity Cost of Capital is equal to the return that could have been earned on alternative investments at a specific level of risk.” In other words, it is the competitive return available in the market on a comparable investment, with risk being the most important component of comparability.The valuation process is one of analysis of expected returns and pricing of risk. The cost of capital is the return appropriate for the expected level of risk in the expected returns. It is the price of risk. But often observed returns do not match expected returns. That is the essence or risk.The cost of capital is the expected rate of return on some base value. That base value is measured as the market value of an asset, not its book value, par value, or carrying value. For example, the yield to maturity shown in the bond quotations in the financial press is based on the closing market price of a bond, not on its face value. Similarly, the implied cost of equity for a company’s stock is based on the market price per share at which it trades, not on the company’s book value per share of stock. The cost of capital is estimated from market data. These data refer to expected returns relative to market prices. By applying the cost of capital derived from market expectations to the expected net cash flows (or other measure of economic income) from the investment or project under consideration, the market value can be estimated.

Keep in mind that we have talked about expectations including inflation. Assuming inflationary expectations, the return an investor requires includes compensation for reduced purchasing power of the currency over the life of the investment. Therefore, when the analyst or investor applies the cost of capital to expected returns in order to estimate value, he or she must also include expected inflation in those expected returns.This obviously assumes that investors have reasonable consensus expectations regarding inflation. For countries subject to unpredictable hyperinflation, it is sometimes more practical to estimate the cost of capital in real terms rather than in nominal terms and apply it to expected net cash expressed in real terms.The essence of the cost of capital is that it is the percentage return that equates expected economic income with present value. The expected rate of return in this context is called a discount rate. By discount rate, the financial community means an annually compounded rate at which each increment of expected economic income is discounted back to its present value. A discount rate reflects both the time value of money and risk. Therefore, in its totality it represents the cost of capital. The sum of the discounted values of each future period’s net cash flow or other measure of return equals the present value of the investment, reflecting the expected amounts of return over the life of the investment. The terms discount rate, cost of capital, and required rate of return are often used interchangeably.The economic income referenced here represents total expected benefits. In other words, this economic income includes increments of cash flow realized by he investor while holding the investment, as well as proceeds to the investor upon liquidation of the investment. The rate at which these expected future total returns are reduced to present value is the discount rate, which is the cost of capital (required rate of return) for a particular investment.Discount rate and capitalization rate are two distinctly different concepts. Discount rate equates to cost of capital. It is a rate applied to all expected economic income to convert the expected economic income stream to a present value.A capitalization rate, nevertheless, is merely a divisor applied to one single element of the economic income stream to estimate a present value. The only instance in which the discount rate is equal to the capitalization rate is when each future period’s economic income is equal (i.e., no growth), and the economic income is expected to continue into perpetuity. One of the few examples would be preferred stock paying a fixed dividend amount per share into perpetuity.SUMMARYThe hypothesis of capital market efficiency has attracted a great deal of interest and critical comment. This is somewhat surprising for the reason that capital market efficiency is a fairly limited concept. It says that the prices of securities instantaneously and fully reflect all available relevant information. It does not imply that product markets are perfectly competitive or that information is costless.Capital market efficiency relies on the ability of arbitrageurs to recognize that prices are out of line and to make a profit by driving them back to an equilibrium value consistent with available information. Given this type of behavioral paradigm, one often hears the following sort of questions: If capital market efficiency implies that no one can beat the market (i.e., make an abnormal profit), then how can analysts be expected to exist since they, too, cannot beat the market? If capital markets are efficient, how can we explain the existence of a multibillion dollar security analysis industry? The answer of course, is that neither of these questions is inconsistent with efficient capital markets. First, analysts can and do make profits. Nevertheless, they compete with each other to do so. If the point to analysis becomes abnormally large, then new individuals will enter the analysis business until, on average, the return from analysis equals the cost (which by the way, includes a fair return to the resources that are employed). As shown by Cornell and Roll (1981), it is reasonable to have efficient markets where people earn different gross rates of return for the reason that they pay differing costs for information. Nevertheless, net of costs their abnormal rates of return will be equal (to zero).

The concept of capital market efficiency is important in a wide range of applied topics, such as accounting information, new issues of securities, and portfolio performance measurement. By and large the evidence seems to indicate that capital markets are efficient in the weak and semistrong forms but not in the strong form.REVIEW QUESTIONS

1. Suppose you know with certainty that the Clark Capital Corporation will pay a dividend of TZS10 per share on every January 1 forever. The continuously compounded risk-free rate is 5 percent (also forever).(a) Graph the price path of Clark Capital common stock over time.(b) Is this (highly artificial) example a random walk? A martingale? A submartingale? (Why)?

2. Given the following situations, determine in each case whether or not the hypothesis of an efficient capital market (semistrong form) is contradicted.(a) Through the introduction of a complex computer program into the analysis of past stock

price changes, a brokerage firm is able to predict price movements well enough to earn a consistent 3 percent profit, adjusted for risk, above normal market returns.

(b) On the average, investors in the stock market this year are expected to earn a positive return (profit) on their investment. Some investors will earn considerably more than others.

(c) You have discovered that the square root of any given stock price multiplied by the day of the month provides an indication of the direction in price movement of that particular stock with a probability of .7.

(d) A securities and Exchange Commission (SEC) suit was failed against ABC Limited Company in 1965 for the reason that its corporate employees had made unusually high profits on company stock that they had purchased after exploratory drilling had started in Dar es Salam (in 1979) and before stock prices rose dramatically (in 1984) with the announcement of the discovery or large mineral deposits in Dar es Salam.

3. The first national bank has been losing money on automobile consumer loans and is considering the implementation of a new loan procedure that requires a credit check on loan applications. Experience indicates that 82 percent of the loans were paid off, whereas the remainder defaulted. Nevertheless, if the credit check is run, the probabilities can be revised as follows:

Favorable Credit Check

Unfavorable Credit Check

Loan is paid .9 .5Loan is defaulted .1 .5

An estimated 80 percent of the loan applications receive a favorable credit check. Assume that the bank earns 18 percent on successful loans, loses 100 percent on defaulted loans, and suffers an opportunity cost of 0 percent when the loan is not granted and would have defaulted. If the cost of a credit check is 5 percent of the value of the loan and the bank is risk neutral, should the bank go ahead with the new policy?

4. The Western Foods, one of the nation’s largest consumer products firms, is trying to decide whether it should spend TZS5 million to test market a new ready to eat product to proceed directly to a nationwide marketing effort, or to cancel the product. The expected payoffs (in million TZS) from cancellation versus nationwide marketing are given below:

ActionMarket Conditions Cancel Go

Nationwide

No acceptance 0 –10Marginal 0 10Success 0 80

Prior experience with nationwide marketing efforts has been:

Market Conditions ProbabilityNo acceptance .6Marginal .3Success .1

If the firm decides to test market the product, the following information becomes available:Probability

No Acceptance

Marginal Success

Outcome Predicted by the Test Market

No Acceptance .9 .1 0Marginal .1 .7 .2Success .1 .3 .6

For example, if the test market results predict a success, there is a 60 percent chance that the nationwide marketing effort really will be a success but a 30 percent chance it will be marginal and a 10 percent chance it will have no acceptance.

(a) If the firm is risk neutral, should it test market the product or not?(b) If the firm is risk averse with a utility function U(W) = in(W+11), should it test market the

product or not?5. The efficient market hypothesis implies that abnormal returns are expected to be zero. Yet in

order for markets to be efficient, arbitrageurs must be able to force prices back into equilibrium. If they earn profits in doing so, is this fact inconsistent with market efficiency?

6. (a) In a poker game with six players, you can expect to lose 83 percent of the time. How can this still be a martingale?

b. In the options market, call options expire unexercised over 80 percent of the time. Thus the option holders frequently lose all their investment. Does this imply that the options market is not a fair game? Not a martingale? Not a submartingale?

7. If securities markets are efficient, what is the NPV of any security, regardless of risk?8. State the assumptions inherent in this statement: A condition for market efficiency is that there

be no second-order stochastic dominance.

CASE STUDYThe Optimal Financing Mix

Objective: To estimate the optimal mix of debt and equity for your firm and to evaluate the effect on firm value of moving to that mix.Key Questions

• Based on the cost of capital approach, what is the optimal debt ratio for your firm?Bringing in reasonable constraints into the decision process, what would your recommended debt ratio be for this firm?

• Does your firm have too much or too little debt- relative to the industry in which they operate?- relative to the market?

Framework for Analysis1. Cost of Capital Approach

• What is the current cost of capital for the firm?• What happens to the cost of capital as the debt ratio is changed?• At what debt ratio is the cost of capital minimized and firm value maximized? (If they

are different, explain.)• What will happen to the firm value if the firm moves to its optimal?• What will happen to the stock price if the firm moves to the optimal and stockholders

are rational?2. Building Constraints into the Process

• What rating does the company have at the optimal debt ratio? If you were to impose a rating constraint, what would it be? Why? What is the optimal debt ratio with this rating constraint?

• How volatile is the operating income? What is the “normalized” operating income of this firm, and what is the optimal debt ratio of the firm at this level of income?

3. Relative Analysis• Relative to the industry to which this firm belongs, does it have too much or too little in

debt? (Do a regression, if necessary.)• Relative to the rest of the firms in the market, does it have too much or too little in

debt? (Use the market regression, if necessary.)Getting Information about Optimal Capital StructureTo get the inputs needed to estimate the optimal capital structure, examine regulatory filings and the annual report. The ratings and interest coverage ratios can be obtained from the ratings agencies (S&P, Moody’s), and default spreads can be estimated by finding traded bonds in each ratings class.Online sources of information:www.stern.nyu.edu/~adamodar/cfin2E/project/data.htm

Unit 5

Dividend Policy: Empirical Evidence and Applications

INTRODUCTION

In general firms are free to choose the level of dividend they desire to pay to holders of ordinary shares, even though factors such as legal requirements, debt covenants and the availability of cash resources enforce some limitations on this decision. It is thus not astonishing that the experiential literature has recorded regular variations in dividend behaviour in the firms, countries, time and type of dividend.

Variances between firms are noted, for instance, in Fama and French (2001). They bring substantiation to show that US dividend firms tend to be big and cost-effective, while non-payers are typically small, less profitable but with high investment opportunities. Variations throughout the countries comprise La Porta, Lopez-de-Silanes, Shleifer and Vishny (2000) who study the dividend policies of over 4000 firms from 33 countries around the world. It is established that dividend policies vary across legal regimes in a way that is regular with the idea that dividend payment is the outcome of effective pressure by minority shareholders to limit agency behavior. Therefore, firms in common law countries with good legal protection of investors tend to have higher payout ratios compared with firms in countries with infirmer legal protection. This is unfailing with Allen and Michaely (1995), who note that firms in the US, had payout ratios of about 60 percent during 1980s and early 1990s. Nevertheless, during the same period, Glen, Karmokolias, Miller and Shah (1995) observe an expenditure ratio of only about 40 percent, for a composite of emerging markets’ firms. Time trends in dividend behaviour is inquired by Fama and French (2001), who find that the percentage of US firms that Pay dividends fell from 66.5 in 1978 to 20.8 percent in 1999. The study also depicts a declining trend in the propensity to pay dividend by US corporations in the time period from the late 1970s to the late 1990s. Similarly DeAngelo and Skinner (2000) look at time trends in the type of dividends paid by US firms. They find that special dividends have gradually vanished in the period from the 1940s to the 1990s although incidents of very large special dividends have increased.

In light of the freedom over dividend policy and the experiential variations across firms, countries, time and type of dividends, the question of how dividend policy is determined has been the subject of numerous studies. This question is often referred to as the dividend puzzle, and the debate is usually believed to have been started by Miller and Modigliani’s (1961) irrelevancy theory. Miller & Modigliani (1961) show that in a perfect capital market with rational behaviour and perfect certainty and with investment and borrowing decisions given, dividend policy has no effect on the value of the firm.4

4 Miller & Modigliani (1961) specify what they mean by perfect capital market, rational behavior and perfect certainty. In a perfect capital market all buyers and sellers have equal access to information, and none receives preferential treatment. Thus all traders are price-takers and none can affect the market price. Further, trading does not entail any transaction costs and there are no tax differentials associated with paying dividends either for firms or for individuals. Rational behavior implies that more wealth is preferred to less, and investors are indifferent to whether wealth comes in the form of dividend or capital gains. Perfect certainty means that all investors are certain about the future investment and profits of all firms, thus there is no need to distinguish between debt and equity and an all equity firm is assumed.

The value of the firm at the commencement of the period can be articulated as the dividend to be received during the period plus the firm’s value at the end of the period, less the amount of external finance raised during the period, all expressed in present value terms. In turn, the amount of external finance is the amount of funds required to finance planned investments, less the firm’s earnings after deducting the amount of dividends paid. As the dividends for the period appear twice on the RHS of the equation with opposite signs, they sum to zero and hence eliminated.

Vt = PV[Dt + Vt+1 – (It –(Xt –Dt))] (5.1)

Where Vt is the value of the firm at time t; PV stands for present value; Dt is the total dividends paid during period t; It is investment; and Xt is the firm’s net profit for period t.

The firm’s value at the ending of the period can correspondingly be defined as the dividend paid during that period plus the value at that period end less external finance raised throughout the period, and so forth for all future periods. The present value of the firm can as a result be expressed as the inestimable sum of the present values of future earnings less investment expenditures. In view of the fact that profits, investment and the discount rate are all independent of dividend either by their nature or by assumption, the conclusion is that the dividend decisions has no effect on value.

When the suppositions underling the irrelevancy theory are unperturbed the question is whether it is still reasonable to conclude that dividends will have no effect on expected earnings, investment or on the firm’s risk and hence the discount rate. For instance, future earnings of a firm that pays dividends may be lower relative to a similar firm that does not pay dividends comprises incurring transaction costs or extra taxes. Indeed, much of the dividend literature has focused on the entailments of relaxing the Miller & Modigliani (1961) irrelevancy theory assumptions and of introducing market imperfections.

The text that deals with dividend policy in the presence of market imperfectnesses may be categorized under two basic views: for and against. On the ‘against’ camp are theories comprising the transaction cost theory of dividend and the tax hypothesis that suggest that dividend payments reduce shareholder wealth. On the ‘for’ cam are theories that suggest that dividend payments increase shareholders wealth, comprising the bird in hand argument, the signaling theory and the agency theory of dividend. All these theories have been broadly discussed and tested but to date there is no agreement on how firms determine their dividend policies.

DIVIDEND THEORIES

The Transaction Cost Theory

Firms may gain costs in distributing dividends while investors may incur costs in collecting and reinvesting these payments. In addition, both firms and investors may deserve costs when, due to paying dividends, the firm has to raise external finance in order to meet investment needs. Certainly, the transaction costs obtained in having to resort to external financing, is the cost of dividend in Bhattacharya’s (1979) model. In contrast, nevertheless, it may be argued that dividend are beneficial as they save the transactions costs associated with selling stocks for consumption purposes. Either way, if

there are additional transaction costs that are related with paying or not paying dividends, then dividend policy should impact earnings expectations and consequently share price and firm value.

Alternatively dividends may influence value if dividend policy has an influence on management’s investment decisions. For instance, managers may decided to forgo positive net present value investments for the reason that dividend payments exhausted internal finance and raising external funds comprises transaction or other costs. Indeed in Miller and Rock’s (1985) model the cost of dividends arise from cutting or distorting the investment decision. Nevertheless, more characteristically, the transaction cost theory of dividend retains the assumption of a given level of investment, and focuses on the costs of raising external funds when the firm increases its dividend payment. Transaction costs comprise flotation costs to the firm of raising additional external finance such as underwriter fees, administration costs, management time, and legal expenses. Further, when the firm pays dividend and then has to raise additional external finance, existing shareholders suffer dilution of control. Accordingly to maintain control or for other reasons, existing shareholders may subscribe to the new issue, incurring trading costs such as stamp duty and stockholders’ commissions. Finally all these transaction costs are reflected in the share price and firm value.

In addition to explicit transaction costs there are also less noticeable costs that are related with paying dividend and resorting to external finance, and which are due to information asymmetries and pecking order considerations. Chiefly, raising new equity can be costly if it comes at a time when the shares are temporarily under-valued or due to the signals this action send to the market regarding the value of the firm. In the same way, debt issues are also problematic for the reason that the announcement of the issue may be related with increased probability of default and with managers trying to issue debt before such bad news are revealed. Like explicit transaction costs, these less obvious costs should also influence earnings expectations and be reflected in the firm’s share price and value.

Consequently, due to the costs connected with raising external finance, the transaction cost theory of dividend suggests that firms should make use of retained earnings to the extent possible. Dividend should only be paid when this does not result in shortage of internal funds that are essential for investment. Thus Rozeff (1982) suggests that firms that have larger reliance on external finance would maximize shareholder wealth by adopting lower payout policies. Leverage, growth potential and instability are all factors that can increase dependency on costly external funds. High levels of leverage imply high fixed costs that the firm has to ensure it can meet. Growth potential means the firm is faced with good investment opportunities for which it require funds. Likewise earnings volatility suggests that dependency on external finance is higher for the reason that there is less certainty concerning earnings to be generated. This entails that highly leveraged, risky or growth firms should be related with traditional payout policies.

Another significant factor that has entailments for control consideration and for the transaction costs of raising external finance and thus for firms’ dividend policies, is size. Particularly, the ownership structure of small companies is likely to be less dispersed than that of larger firms. The more dispersed is ownership the less control is exercised by each shareholder and hence the problem of loosing control is more serious for smaller firms. Further, the cost of external finance is likely to be privileged for smaller

firms compared with larger, well-established firms with easier access to the capital markets. Ass to this the observation that growth firms are generally smaller and the conclusion is that small firms are likely to find the payment of dividends more costly compared with larger firms. This conclusion may explain the positive correlation often observed between firm size and the likelihood that the firm is a dividend.

Tax Theories

Another cost linked with dividend payments is taxes. The tax hypothesis proposes that corporate tax on distributions and taxes on dividends in the hand of investors are significant costs to be well thought-out when deciding on a dividend policy. More specifically, the difference between tax on dividends and on capital gains should be considered as well as the distinction between corporate tax on distributed and on retained earnings. For instance, if corporate tax on distributions is higher than that on retained earnings, this may reduce expected earnings of a firm that pays dividends relative to a firm that does not. In the same way, if dividends in the hands of shareholders are taxed higher than capital gains, investors should evaluate expected returns on an after tax basis and share prices will vary inversely with the firm’s payout level. Undeniably, the basic tax hypothesis proposes that additional taxes on dividends make capital gains a less costly way of returning wealth to shareholders. Thus, the basic tax hypothesis supports a conservative dividend policy, and proposes that if the firm wants to return cash to shareholders then this should be done through share purchases. It is thus puzzling to find that even though repurchases have increased since the 1980s, they have not substituted for dividends.

Nevertheless, Miller and Scholes (1978) show that under two provisions of the US Internal Revenue Code, taxable investors may still be indifferent to dividends even when the tax regime favours capital gains5. In addition, Miller and Modigliani (1961) argue that despite the presence of taxes, tax-induced clientele effect greatly reduces the tax costs of dividends. The idea is that there may be clienteles for both high and low dividend yields depending on tax positions. Institutions, which are often tax-exempt and individuals at low tax brackets may prefer companies with high payout policies. Other investors at high tax brackets for whom the relative tax cost of dividends is considerable will prefer firms with low payout policies. Shareholders select firms whose policies suit their preferences. As there are enough firms to satisfy all, no firm can increase its value by changing its dividend policy. In addition, by changing its dividend policy, a firm may trigger a change in clientele and this could be costly due to trading costs. Therefore, the clientele effect hypothesis supports the dividend irrelevancy conclusions.

The Bird in the Hand Argument

The conventional argument in favour of dividend is the idea that dividends reduce risk for the reason that they bring shareholders’ cash inflows forward. Although shareholders can create their own

5 The first provision used to illustrate the irrelevancy of taxes in Miller and Scholes (1978) is the ability to deduct interest payments from investment income received, in calculating tax liability. The second provision is that insurance companies pay no taxes on investment income. Thus, if the firm increases its dividend, a taxable investor can avoid the additional tax liability by increasing his interest liability to the point where it matches the increased level of dividends. To maintain the same level of risk, the investor can use the proceeds from the additional borrowings to buy insurance policy. This increases the investor’s level of assets so that his debt ratio is unchanged.

dividends by selling part of their holdings, this entails trading costs, which are saved when the firm pays dividends. The risk reduction or bird in the hand argument is connected with Graham and Dodd (1951) and with Gordon (1959) and it is often defended as follows. By paying dividends the firm brings forward cash inflows to shareholders, thus reducing the vagueness related with future cash flows. In terms of the discounted dividend equation of firm value, the idea is that the required rate of return demanded by investors (the discount rate) increases with the plough-back ratio. Even though the increased earnings retention brings higher expected future dividend, this additional dividend stream is more than offset by the increase in the discount rate.

This argument neglects the fact that the risk of the firm is decided by its investment decisions and not by how these are financed. The mandatory rate of return is influenced by the risk of the investments and should not change if these are financed from retained earnings rather than from the proceeds of new equity issues. As distinguished by Easterbrook (1984), despite of paying dividends the firm does not withdraw from risky investments, thus the risk is merely transferred to new investors.

The signaling theory

A more persuasive argument in favor of dividends is the signaling hypothesis, which is linked with propositions put forward in Bhattacharya (1979), Miller and Rock (1985), John and Williams (1985), and others. It is based on the idea of information asymmetries between the various participants in the market and in particular between managers and investors. Under such circumstances, the costly payment of dividend is used by managers, to signal information about the firm’s prospects to the market. For instance, in John and Williams’ (1985) model the firm may be temporarily under-valued when investors have to meet their liquidity requirements. If investors sell their holdings when the firm is undervalued, then there is a wealth transfer from old to new shareholders. Nevertheless, the firm can save losses to existing shareholders by paying dividends. Even though investors pay taxes on the dividends, the benefits from holding on to the undervalued firm more than offset these extra tax costs. A poor quality firm would not mimic the dividend behaviour of an undervalued firm for the reason that holding-on to over-valued shares does not enlarge wealth.

The signaling hypothesis can give details about the preference for dividends over stock repurchases in spite of the tax benefit of the latter. Predominantly, as suggested in Jagannathan, Stephens and Weisbach (2000) among others, the regular dividend signal an ongoing commitment to pay out cash. This signal is dependable with Lintner (1956) observation that managers are characteristically reluctant to decrease dividend levels. Nevertheless, unlike regular dividends, repurchases and special dividends can be used to signal prospects without long-term commitment to higher payouts. For that reason, announcements of increases in regular dividends signal permanent improvements in performance, and should be interpreted as confidence in the firm on behalf of managers thus triggering a price rise. On the other hand, announcements of dividend decreases should be interpreted as signaling poor performance and lack of managerial confidence and should for that reason trigger drops in prices.

If changes in the levels of dividend release information to the market, then firms can decrease price volatility and manipulate share prices by paying dividends. Nevertheless, it is only unexpected changes

which have an informative value and which can thus impact prices. Therefore, the value of the signal depends on the level of information asymmetries in the market. For instance, in developing countries where capital markets are typically less competent and where information is not as reliable as in more sophisticated markets, the signaling function of dividend may be more significant. Furthermore, it can be argued that information will finally be revealed whether or not the dividend signal is sent, therefore the dividend affect on prices is only temporary.

The Agency Theory of Dividend

One more argument in favour of generous dividend payments is that this shifts the reinvestment decision back to the owners. The primary assumption is that managers may not essentially always act as to maximize shareholders’ wealth. The problem here is the separation of ownership and control which gives rise to agency conflicts as defined in Jensen and Meckling (1976). Therefore when the levels of retained earnings are high managers are expected to channel funds into bad projects either in order to advance their own interests or due to incompetency. Therefore, generous dividend policy raises the firm’s value for the reason that it can be used to decrease the amount of free cash flows in the discretion of management and thus controls the over investment problem.

Another agency theory based explanation of how dividends augment value is described in Easterbrook (1984). While the transaction cost theory of dividend proposes that dividend payments reduce value for the reason that they lead to the raising of costly external finance, Easterbrook (1984) argues that it is this process which reduces agency problems. The idea is that the payment of dividends is one possible solution to the problem of collective action that tends to lead to under-monitoring of the firm and its management. Hence, the payment of dividends and the subsequent raising of external finance induce investigation of the firm by financial intermediaries such as investment banks, regulators of the securities exchange where the firm’s stock is traded, and potential investors. This capital market monitoring cuts down agency costs and lead to appreciation in the market value of the firm. Furthermore, total agency cost, as defined by Jensen and Meckling (1976), is the sum of the agency cost of equity and the agency cost of debt. The later is partly due to potential wealth transfer from bond to equity holders through assets transpositions. Thus Easterbrook (1984) note that by paying out dividends and then raising debt, new debt contracts can be managed to decrease the potential for wealth transfer.

Review of Selected Empirical Studies

The dividend theories discussed in the preceding section speak about the affect of dividend on value of transaction costs, taxes, risk, signaling and agency conflicts. Nevertheless, the chief empirical studies of the dividend policy puzzle focus in particular on the tax hypothesis, the signaling hypothesis and agency studies6.

Therefore, following the spirit in Prasad, Green and Murinde (2001), it is around these three theories that the following debate is organised. Transaction costs that are incurred due to changes in dividend policies are usually incorporated into each of these main hypotheses. These costs are usually assumed

6 Indeed these three theories (agency, asymmetric information and taxation) also commonly underline empirical work of the other financing decision, namely the capital structure choice.

to be a function of reliance on external finance and are controlled for by variables such as growth, size or profit. Relatively little empirical work has been conducted on the bird in hand argument consequently this branch of empirical work is discussed no further7.

Testing approaches depend to a large extent on the hypothesis under investigation. The clientele effect is often evaluated by an event study around the dividend payment days. Other tax studies look at the trading activity rather than the stock price behaviour around ex-dividend days. Some tax hypothesis studies take an another approach, and review the affect impact of tax reforms on relative prices while other regress the dividend policy on tax proxies to assess the significance of the latter in influencing the former.

Studies that investigate the signaling hypothesis often follow an event study around the dividend announcement period. Other signaling studies assess revisions in earnings forecasts following unexpected changes in dividends. A different approach to testing the validity of the signaling hypothesis is by looking at changes in firm characteristics, following changes in its dividend policy. A particular attention has often been paid to changes in earnings. Cross sectional comparisons between firms of varied characteristics are also used to assess how such distinctions may influence the value of the dividend signal.

Agency theory studies usually use regression analysis to evaluate the degree of substitutability among alternative mechanisms for controlling agency problems. Another approach, which is characteristically classified under the agency theory umbrella, is testing the fittingness of Rozeff’s (1982) cost minimization model. The cost minimization model actually combines transaction costs theory with agency theory, and proposes that the optimal payout ratio is that which minimizes the sum of costs of paying dividends. Thus, Rozeff (1982) and subsequent studies regress a proxy of the optimal payout ratio on proxies for agency costs that may be controlled by paying dividends and on proxies for transaction costs that are associated with dividend payment.

The literature review of this section will proceed by investigating a limited number of studies dealing with each of the above discusses theories in turn. Nevertheless, some researchers have attempted to model the management’s decision-making process that determines dividend changes. Some of these behavioral models, notably Lintner’s (1956), have significant implications in particular for the signaling theory and are therefore described first.

Behavioural Models — The Partial Adjustment Model

The main studies

One approach to dealing the dividend puzzle is to comprehend the management’s decision-making process that determines dividend changes. Undeniably, this is the approach in Lintner (1956), who carry

7 One exception is Allen and Rachim (1996) who investigated the relation between risk and the dividend policy using 173 Australian listed companies for the period 1972 to 1985. Stock price volatility is regressed on the dividend yield and on six control variables including earnings volatility, payout ratio, debt, growth, size and industry dummies. The study fails to find evidence that dividend yield is significantly correlated with stock price volatility, which suggests rejection of the bird in the hand (or duration) effect.

out a series of interviews with the managers of 28 US industrial firms about their firms’ dividend policies in the 7 years from 1947 to 1953. From the survey it comes forth that firms tend to establish dividend policies with target payout ratios that are applied to current earnings. It is also found that firms have adjustment rates that establish the percentage of the largest change by which dividend levels are actually changed. Lintner (1956) also reports that although the target payout ratios and speed of adjustments vary across firms, in most cases they stay sensibly stable over time.

Established on his findings, Lintner (1956) develops the partial adjustment model of the change in the dividend level from the previous to the current period. The model reflects management’s belief that investors dislike erratic patterns in dividend levels and hence the emphasis is one the change from the previous actual level:

∆Di,t = αi + Ci [D*i,t —Di(t–1)] + Ui,t (5.2)

Where,

∆Di,t = Di,t –Di(t-1) (5.3)

D*i,t = Ri(Pi,t) (5.4)

Thus, ∆Di,t is the change in the dividend payment; Di,t and Di(t-1) are the amounts of dividends paid in years t and t–1 respectively; D*t,i is the target divided amount where Ri is the target payout ratio and Pi,t

is current profits after tax; Ci is the speed of adjustment; αi is a constant which in general will be positive to reflect management’s reluctance to reduce dividends; Ui,t is an error term. Equation (5.2) can alternatively be expressed as follows:

Di,t = αi,t + βPi,t + γ Di (t-1) + Ui,t (5.5)

Where

β = Ci(Ri) and γ = 1 –Ci

As per Lintner (1956), current net earnings, Pt, play the most significant role in determining dividend changes. This is for the reason that current earnings are widely available and hence managers’ view is that investors expect dividends to reflect changes in this variable. Expanding (5.5), noting that Di(t-1) can be expressed as a function of that period’s profits and the previous period’s dividends, the dividend pattern is a smoothed pattern of earnings and is indicative of the time path of permanent earnings. The degree of smoothing depends on the speed of adjustment coefficient, Ci.

Therefore, the three key factors in the partial adjustment model are the speed of adjustment coefficient, Ci, the target payout ratio, Ri, and current earnings, Pt. Undeniably, the three questions that are usually raised about the Lintner model concern these factors. First, some researches have examined what decides the speed of adjustment and hence the degree to which smoothing takes place. Second, some researches try to establish whether firms have long-term target payout ratios towards which they move. Third, the question of whether current earnings are the key determinant of dividends has been

investigated. In general, nevertheless, empirical tests of the Lintner model have confirmed its validity. One of the earliest and widely quoted such study is by Fama and Babiak (1968). One more, which is going to be reviewed here for the reason explained below, is by Mookerjee (1992).

Mookerjee (1992) is unique in that it applies the Lintner model, which has been developed on the basis of a US survey, to a developing rather than a developed country. Chiefly, annual data for the aggregate Indian corporate sector for the period 1949 to 1981, before important reforms were introduced, is utilized to show that the basic Lintner model performs well in explaining dividend behaviour in India. Alteration of the basic model, by adding the availability of external finance as an explanatory variable, improves the fit of the model. Indeed, the lagged external finance enters with an important and positive estimated coefficient reflecting access to subsidized borrowing and hence tendency to use borrowing to finance higher dividends. Mookerjee (1992) also notes that the constant in the Lintner model is hypothesized to be important and positive, reflecting the fact that firms are more willing to raise rather than lower dividends. Even though the study finds the constant to be important under all specifications, it enters with a negative sign in all regressions. It is suggested that this could be reflection of the affect of taxes8.

Even if the study by Mookerjee (1992) is supportive of the Lintner’s model, it also addresses the third of the three questions discussed above, that are often raised with reference to this model. Specifically this is the question of whether management set the desired dividend level as a fraction of current earnings or as a fraction of permanent earnings. If the latter is the case and it is assumed that earnings follow a random walk with a drift, than the lagged profit after tax, should enter with a negative and important coefficient. Mookerjee (1992) finds that although the lagged earnings enter with a negative coefficient, in all cases it is also unimportant. In contrast, Lee (1996) finds stronger support for the view that it is permanent earnings as oppose to current earnings that determine dividend.The study by Lee (1996) assesses whether there is long-term relationship between different definitions of earnings and dividends. The study utilises a bivariate time-series model of earnings and dividend obtained from annual observations on the Standard & Poor's Index for the period 1871 to 1992. The model is adequately universal to allow various specification of target dividend to be nested within it. These restrictions are then tested, taking into account the non-stationarity of the dividend and earnings series and the cointegration between them. The results show that dividend behaviour is determined primarily by changes in permanent earnings and that the Lintner model performs better when the target payout ratio is a function of permanent rather than current earnings. This is supportive of the signalling hypothesis in the sense that current earnings are not a good pointer of the long-term financial position, therefore managers utilise dividends to signal this position.Shirvani and Wilbratte (1997) also use cointegration (albeit multivariate rather than bivariate) techniques to test the validity of the Lintner model. Nevertheless, their main aim is to address the second of the three questions mentioned above, namely whether firms have long-term payout ratios. Using quarterly observations on the Standard & Poor’s 500 index for the period 1948 to 1994, the first stage is to confirm the nonstationarity of the dividends, earnings and price index series. In addition, as these three series are found to cointegrate, tests of the coefficients in the cointegrated equation point to a long-run relationship between earnings and dividends. In general the hypothesis that the coefficients on the logs of the dividend and earnings variables, are equal and of the opposite signs is not rejected.

8 Mookerjee (1992) note that traditionally income tax had been very high in India, disadvantaging dividend payments over capital gains. It is also noted, however, that this trend began to reverse in the mid-1970s.

The Shirvani and Wilbratte's (1997) study further counts on the error correction model to capture short-run deviations from the long-run target payout ratio and the speed of adjustment. Thus the study also touches on the first of the three questions about the Lintner model, specifically the question of what determines the speed of adjustment. It is found that firms apply varied adjustment rates in raising and lowering dividends. When the payout ratio is below its long-run target, the firm will increase dividends. Nevertheless, when the payout ratio is above its target, the firm will hold the dividend level constant and wait for earnings to grow so that the target payout ratio is achieved. This ratchet effect is interpreted in terms of the signalling theory, and in particular as a way of avoiding the bad signals connected with dividend reductions.The idea that the speed of adjustment is determined by the signalling role of dividends is also supported in Dewenter and Warther (1998). The study reports the results from running the partial adjustment model for each of 180 Japanese firms and 313 US firms with at least five years of nonzero dividend during the period 1982 to 1993. It is found that the median speed of adjustment is higher for Japanese firms compared with US firms, and higher still for Keiretsu members. This pattern is explicated by the observation that the Japanese business environment is characterized by less information problems, thus there is less need for the dividend-smoothing device in the case of these firms.Turning back to the question about the existence of long-term payout ratios, Hines (1996) looks at possible reasons for the Lintner (1956) observation that payout ratios vary across firms. In particular, the payout rates of 505 US firms for the period 1984 to 1989 as well as the dividend patterns for the aggregate US corporate sector during the period 1950 to 1986 are investigated. Hines (1996) finds that the payout rates applied to profits from foreign sources are about three times higher than the payout rates applied to domestic profits. These findings support the signalling hypothesis in view of the fact that information asymmetries surrounding overseas operations are likely to be more acute than for domestic activities. Managers, thus, may feel a stronger need to send signals regarding the prospects of foreign operations.Conclusions on the empirical studies of the Partial Adjustment ModelTable 5.1 summarizes the pertinent key issues of each of the empirical studies reviewed in this section. Empirical findings appear to support the validity of the partial adjustment model, not only in respect to the behaviour of US firms, as shown in Fama and Babiak (1968), but also with respect to the behaviour in less developed countries as in Mookerjee (1992). The other studies reviewed above address the three questions that are related with the Lintner model regarding the existence of a target payout ratio, the determinants of the speed of adjustment coefficient, and the degree to which current earnings make clear dividend levels.The Lintner’s idea of a long-term payout ratio is supported in Shirvani and Wilbratte (1997), while Hines (1996) provides evidence supporting the notion that difference in the payout target is due to the signalling role of dividends and the degree of information asymmetries faced by the firm. The signalling role of dividends is likewise supported by Shirvani and Wilbratte (1997), who show that firms apply different adjustment rates to dividend increases and decreases. In addition evidence on what determines the speed of adjustment towards target dividend is given in Dewenter and Warther (1998). Finally, Lee (1996) shows that the partial adjustment model works better when the target dividend is modelled as a function of permanent as opposed to transitory earnings.

Table 5.1: Selected empirical studies of the Partial Adjustment ModelStudy Aim Methodology (Data and

Model) and resultsConclusions

Lintner (1956) To assess the process by which dividends are determined through interviews with

Data:US, 28 industrial firms selected to ensure a wide range of

Firms are primarily concerned with the stability of dividends. Managers

managers of US firms circumstances under which managers operate, 1947-1953Methodology: Survey – interviews

with managers to learn of the dividend policy of their firms over the previous seven years

Constructing a model that describes the dividend change behavior

Testing the model in predicting post-war dividend of all American corporations

Model:Change in dividend = α+ (speed of adjustment coefficient) (target dividend – actual lagged dividend)Where target dividend is the target payout ratio applied to the current year’s profits after taxes.

prefer a gradual upward trend in dividends.

Earnings are the most significant determinants of any change in dividends

Dividend policy is set first and other policies are then adjusted

The market reacts positively to dividend increase announcements and negativity to announcements of dividend decreases

Fama and Babiak (1968) Testing the Lintner dividend model

Data:US, 392 major industrial firms over the period 1946 to 1964Methodology:

1.Ordinary Least Squares time series of a number of specifications of the Lintner model for individual firms, assessing the statistical characteristics of the distribution of the estimated coefficients (using

Net income appears to explain the dividend change decision better than a cash flow measure

The Lintner model performs well relative to other models. Nevertheless, deleting the constant and adding the lagged earnings variable leads to a slight improvement in the predictive power of the model.

part of the sample to estimate the models and the other part to validate it)

2.Monte Carlo experiments – generating series for earnings, dividends and the error terms based on assumptions regarding the data generating process. Then using artificial samples to estimate the coefficients. The estimated coefficients are compared with the coefficients of the model actually used to generate the data.

Model:The Lintner model

Mookerjee (1992) Apply the Lintner partial adjustment model and modifications of this model, to a developing country, India

Data:Annual data for the aggregate Indian corporate sector, 1950-1981 (all variables expressed in real termed by deflating by the consumer price index).Methodology:Ordinary Least SquaresModel:∆(dividend) = α0 + α1 (Profit after tax) – α2 (Lagged profit after tax) – α3 (Lagged dividend) + α4 (Lagged external finance)Results: ∆(dividend) = –0.33 +

0.62 (Profit after tax) –

The basic Lintner model is good in explaining the dividend behaviour in India. All variables are significant at conventional levels and the model explains 61% of variation

Adding external finance in previous period as explanatory variable improves the model. This is explained by access to subsidized borrowing which encourage firms to use borrowings to finance dividends.

0.73 (Lagged dividend) ∆(dividend) = –1.04 +

0.28 (profit after tax) – 0.79 (Lagged dividend) + 0.37 (Lagged external finance)

∆(dividend) = –0.99 + 0.32 (Profit after tax) – 0.09 (Lagged profit after tax) – 0.72 (Lagged dividend) + 0.36 (Lagged external finance)

The constant is significant but enter with a negative sign, reflecting dividend tax disadvantage in India.

Lee (1966) Test the hypothesis that dividend changes are determined by changes in permanent earnings. Show that the Lintner model performs better when the target dividend is a function of permanent as opposed to current earnings.

Data:US, annually, aggregated (real) data on the S&P Composite Index, 1871-1992Methodology: Testing for

cointegration between various definitions of earnings and dividends.

Constructing a Bivariate Vector Autoregression (BVAR) of change in dividends and the spread between earnings, and dividends.

Testing nested models in the BVAR to assess whether permanent earnings, transitory earnings or current earnings, influence dividends.

Model: Permanent earnings

hypothesis: Change in dividend = α1 (change in the permanent component of earnings)

Changes in dividends are influenced by permanent earnings, but not by transitory earnings, but not by transitory earnings. Therefore, the spread between dividends and earnings is influenced by transitory earnings but not by permanent earnings.

The hypothesis that the target dividend in the partial adjustment model is a function of permanent earnings is not rejected while the hypothesis that the target dividend is a function of current earnings is rejected.

Partial adjustment model: Change in dividend = (speed of adjustment) × (target dividend – lagged dividend)

Results: Earnings and

dividends are non-stationary but are cointegrated.

Hence, the Bivariate Moving Average Representation of the first difference in dividends and the spread between earnings and dividends in invertible and can be specified as a BVAR.

Shivani and Wilbrate (1997)

Use cointegration analysis to check for existence of a long-term stable payout ratio. Use the error correction equations to assess the ratchet effect.

Data:US, quarterly, aggregated (real terms) data on the S&P500, 1948:1 to 1994:4Methodology:Multivariate cointegration testsModel:Long-run model: Error term = θ1 log of dividend – θ2 log of earnings + θ3 log of price level + θ4Results:Long-run model: Error term = log of dividend – 1.045 log of earnings + 0.097 log of price level + 0.185 LR test of H0:θ1 + θ2 = θ3 = 0; 0: x2

1 = 0.75 with probability value of 0.30.

Current earnings is the strongest proxy for ability to pay dividends.

The cointegration results suggest that firms pursue a long run payout ratio, consistent with the Lintner model.

Short run ratcher effect – firms apply different adjustment methods to raising and lowering payout ratios. This is due to signaling effects.

Dewenter and Warther (1998)

Compare the responsiveness of the dividends of US and

Data:313 US firms and 180 Japanese companies

The dividends of Japanese Keirestu firms are mor

Japanese firms to earnings changes

with at least 5 years of nonzero dividend and earnings data. 1982-1993Methodology:Run the Lintner model without the intercept term for each firm and obtain the medium speed of adjustment coefficient for the samples of US firms and Japanese firms. Also split the Japanese sample as per Keiretsu membership and compare the median speed of adjustment across these sub-samples.Model:Lintner model without the constantResults:The median speed of adjustment estimates are 0.055 for all US firms, 0.094 for all Japanese firms, and 0.117, 0.082, and 0.021 for keiretsu, hybrid, and independent firms, respectively.

responsive to earnings changes than those of both US firms and Japanese independent firms.

Thus US dividends are smoother than Japanese dividends and this is due to the observation that Japanese firms and keiretsu firms in particular, face less information asymmetry and fewer agency conflicts than US firms.

Hines (1996) Look at possible determinants of payout rates by exploring the connection between payout ratios and the share of firms’ profits from overseas sources.

Data:US, 505 firms for the period 1984-1989 and aggregate US data covering the time period 1950 to 1986Methodology: Ordinary least

squares regression using cross sectional firm-level data for each of the years 1984 to 1989

Ordinary least squares regression

Different payout ratios are applied to domestic and foreign profits – US corporations pay dividends out of their foreign profits at about three times the rate that they do out of their domestic profits

This may be due to the signaling function of dividends if foreign profits are

using time series on aggregate data

Model:Dividend payout = [after tax domestic profits + η(after tax foreign profits)] αResults:Firm-level for year 1986: Dividend payout = 0.2353 after tax domestic profits + 0.7290 after tax foreign profits

particularly difficult for investors to verify.

Before proceeding to the next section, it is significant to note that other behavioural model to that of Lintner (1956) have been suggested and tested. For instance, Cyert, Kang and Kumar (1996) develop a behavioural model where firms do not have a target long-term payout ratio for the reason that managers do not like predicting long term future events. Instead, the model is based on the notion that managers seek to avoid uncertainty and to optimise self welfare.Nevertheless, whether managers have long-term target payout ratios or whether they follow shorter-term goals, the behavioural models imply that managers’ intentions and information on the firm and its future can be inferred from the dividend decision. This is the notion underlying the signalling hypothesis, the empirical evidence on which is reviewed immediately after the following review of selected empirical studies of the tax hypothesis of dividends. Empirical studies of the tax hypothesisTax effectThe basic tax hypothesis suggests that for the reason that personal taxes on dividends tend to exceed those on capital gains, firms have an incentive to adopt a conservative payout policy and such policy should be value enhancing. A probable method to evaluate the validity of this hypothesis is to study stock price and dividend policy changes in respond to tax reforms. Hubbard and Michaely (1997) and Papaioannou and Savarese (1994) adopt this methodology. Alternatively the significance of taxes to the dividend decision may be assessed by regressing dividend policy on proxies for the tax cost of dividends. Gentry (1994) and Lasfer (1996) adopt this methodology. Using data on firms that are listed on either the NYSE or the AMEX for 1987 (65 firms) and 1988 (64 firms), Gentry (1994) finds support for the tax hypothesis. The study investigates the dividend policies of corporations versus Publicly Traded Partnerships (PTPs) in the oil and gas exploration industry. PTPs and corporations in the oil and gas industry are of similar size and this makes them comparable. The chief distinction between PTPs and corporations is that during the period studied PTPs were not taxed at the corporate level and hence escaped the US double taxation system. For that reason, if the tax hypothesis is valid, as PTPs have lower tax cost associated with the payment of dividends, their payout rates should be larger. Using cross sectional instrumental variable technique, the dividend payout is regressed on an organisational form dummy as well as on a number of other control variables. Results of the study point to that firms consider taxes when formulating their dividend policies and that, coherent with the tax hypothesis, PTP pay more dividends than corporations.In addition support for the tax hypothesis, is provided by Lasfer (1996) who uses 108 firms quoted on the LSE for the period 1973 to 1983. The study regards both personal and corporate taxes by running a regression of the partial adjustment model. The original partial adjustment model is adapted to

integrate the effects of both personal and corporate taxes on the determination of the long run target dividend level. Lasfer (1996) tests whether the target dividend (and therefore also the actual dividend) is a function of earnings, of a tax discrimination variable and of a tax exhaustion dummy. The tax unfairness variable, surrogating for the effects of personal taxes, varies inversely with the personal income tax rate. When the tax discrimination is better than one, income tax on dividends is cheaper than tax on capital gain and the firm is anticipated to prefer a high payout policy. The tax exhaustion dummy, surrounding for the effects of the firm’s tax position, is set to one if the taxable profit is lower than gross dividends and advanced corporation tax (ACT) is irrecoverable. When ACT is disregarded, the firm is anticipated to prefer a low dividend payout, and consequently the coefficient is expected to be negatively signed.Results in Lasfer (1996) show all variables to be important and to enter with the signs predicted by the tax hypothesis. In addition, results of an event study are also supportive of the tax hypothesis, rejecting the tax induced clientele effect. Particularly, important and positive abnormal returns are reported on the ex dividend day frequent running with the notion that the price drop on the ex dividend day is systematically less than the value of the dividends. The basic behind this is dividend taxation, which causes the value of the dividends to investors to be less than their nominal amount. The study concludes that taxes affects both the dividend policy and exdividend day returns, and that firms set their dividend policies so as to maximize the after tax returns to their shareholders as well as to minimize their own tax liabilities.Similarly Lasfer (1996), Papaioannou and Savarese (1994) also utilize an elaborated dividend partial adjustment model on a sample of 236 industrial and 40 utility US firms for the period 1983 to 1991. Nevertheless, they focus on firms’ reaction to the US Tax Reform Act of 1986 (TRA). Undeniably the study provides influence that firms adjust their dividend policies in response to changes in the tax system and this is interpreted as supportive of the tax hypothesis. It is reported that a total of 23.2% of the sample firms experience shifts important at the 10 percent level in their target payout ratios in the post-TRA period. This provides further support for the notion that firms consider taxes when setting their dividend policies.The Tax Reform Act, 1986 is also utilized by Hubbard and Michaely (1997) to evaluate the affects of taxes on dividend policy. Specifically, it is studied whether shifts in tax policies have led to shifts in the comparative values of two classes of common shares of a single firm, the Citizen Utilities Company (CU). It is noted that during the period studied holders of class A stock received stock-dividends while holders of class B stock received cash-dividends. The relative price of the shares should for that reason reflect preference or aversion to cash dividends. In addition, the TRA reduced the relative aversion of investors paying tax on dividends at the personal income tax rate to dividend. Therefore if taxes are significant than the relative value of class B should have increase after 1987.Hubbard and Michaely (1997) begin by obtaining the dividend-adjusted average relative price of CU for each of the periods 1982-1984 (pre-TRA), 1985-1986 (TRA implementation period) and 1987-1989 (post-TRA). The relative price is calculated as the ratio of the average price of class A to the average price of class B divided by the relevant dividend ratio. It is found that although the relative price declined significantly from 1.01 in the pre-TRA period to 0.91 in the accomplishment period, this relative increase in the value of class B shares did not have a lasting impact. The relative price in the post-TRA increased to its pre-TRA level. Further, over the period 1978 to 1993 the relative price is found to be around unity. This is in spite of the tax shortcoming of cash dividends during that period, and is therefore incompatible with the tax hypothesis.Probable explanations for the discrepancy between the tax hypothesis and the observed price behaviour of the two classes of CU’s shares are explored in Hubbard and Michaely (1997). One suggested explanation is that clientele influences may be the reason that value is not affected by the tax changes. Undeniably it is noted that clientele effects could also explain the temporary change in value during the

TRA implementation period in terms of shifts in clientele. Even though Hubbard and Michaely (1997) do not find evidence in support of clientele effect, other empirical studies do. Tax clientele effectThe tax clientele effect speaks about the preference of different categories of investors, on the basis of their tax position, for various types of stock. Accordingly firms adjust their dividend policies and investors move to satisfy their tax requirements, until, in equilibrium, no value can be added by changing dividend policy. One probable method of establishing whether a tax-induced clientele exists is to investigate the relationship between the dividend yield on stocks and the marginal income tax rate of investors. Specifically, the finding of an inverse relation between dividend yield and marginal tax rates is supportive of the presence of a clientele effect. Elton and Gruber (1970) suggest that clienteles’ marginal tax rates can be inferred from the ex-dividend day price behaviour. This point, for the case of preferential tax treatment for capital gains, is explained in Green and Rydqvist (1999) as follows.If stocks offer no ex-dividend-day compensations then investors will be unwilling to sell ex-dividend. Selling on ex-dividend days implies paying higher taxes on the dividends and this can be avoided by selling cum-dividend. On the cum-dividend day the price comprises the present value of the dividends to be paid but this is taxed at the (lower) capital gains rate. To ensure investors are willing to hold stock through the payment day and sell ex-dividend, the after tax receipts to the seller who trade on the cum-dividend day must, in equilibrium, be equal to the after tax receipts to the seller who trade on the ex-dividend day. This equilibrium position is shown in Elton and Gruber (1970) to be:

Pc – Tg (Pc – Po) = Pe – Tg [ Pe – Po] + D (1- Td) (5.6)Where Pc is the cum-dividend day stock price; Po is the price for which the stock was purchased; Pe is the ex-dividend day price; D is the amount of dividend; and Tg and Td are the personal tax rates on capital gains and dividends respectively.An expression for the ex-dividend day price drop is obtained from Equation (2.6) and is shown to reflect the marginal tax rate on dividend relative to capital gains of the clientele holding that stock:

(Pc – Pe) / D = (1- Td)/ (1- Tg) (5.7)If tax clientele exists than the ratio of the drop in price relative to the nominal dividend amount should be closer to unity for high-yield stock and less than harmony for low-yield stock. This is for the reason that high-yield stock is held by investors who face lower tax rates on dividends. In contrast, investors in low-yield stock are those facing high taxes on dividends. For these high tax payers, the after tax value of the dividend is substantially less than the amount actually received (D) and the required compensation for receiving the dividends is therefore higher. Elton and Gruber (1970) divide their sample, of 4148 stock listed on the NYSE which paid dividend in the 12 month period from 1 April 1966, into 10 groups as per the value of the dividend yield. They find that tax brackets are negatively related to firms’ dividend policies. This is supportive of the tax clientele effect and suggests that a change in dividend policy rather then the dividend policy itself could affect value. Nevertheless, the Elton and Gruber (1970) approach to inferring the existence of a tax induced clientele effect has been criticised on a number of points. First, it has been suggested that an observed ex-dividend-day-premium (i.e. a price drop, which is less than the dividend amount) could be the result of factors other than a reflection of marginal taxes. Second, it has been argued that short term trading could obscure tax clientele effect on ex-day returns even if tax clientele exists. Third, it is claimed that the volatility of equity prices invalidates inferences about tax effects from ex-dividend-day price behaviour.Frank and Jagannathan (1998) address the first point. Namely, that ex-dividend day price behaviour may not necessarily be the result of a tax clientele effect. It is shown that prices fall on ex-dividend days by less than the value of the dividend even in markets where there are no taxes on either dividends or capital gains. The ex-day premium therefore does not reflect the tax rate faced by the stock’s clientele but is explained by the costs associated with collecting and reinvesting the dividends. The study by Frank and Jagannathan (1998) examine 1,896 cash dividend payments by 351 firms listed on the Hong Kong

Stock Exchange between 1980 and 1993. The sample is split into a low-dividend group and a high-dividend group. The percentage price drop on the ex-dividend day is regressed on the dividend yield for the full sample as well as for the sub samples of low and high dividends. The regression is based on a model of the form

(Pc-Pe)/Pc= α + β (dividend yield) (5.8) α = -Π(SPREAD) (5.9)

Where Pc and Pe are the prices on the last cum dividend trading day and on the first day on which the stock is traded ex-dividend respectively. Π is the ratio of rational to total traders; SPREAD is the average bid-ask spread around the ex-dividend day expressed as a percentage of the cum-price. The slope coefficient, β, represents the value of the dividend to market makers.Frank and Jagannathan (1998) find support for the notion that rational traders try to avoid receiving dividends due to their lack of skill and experience in collecting and reinvesting these payments relative to market makers. Subsequently, on the last cum-day there is a selling pressure while on the ex-day there is a buying pressure. This results in a price drop that is smaller than the value of the dividend and this is reflected in the negatively signed constant. As the dividend amount increases, Π also increases for the reason that of the wealth implications of ignoring the dividends. As Π rises, the ex-day premiums increase and this is reflected in the observation that for the high-dividend sample the constant is larger in absolute value. Nevertheless, even for the low-dividend group, where Π can be expected to be at its lowest, the price drop is still lower than the value of the dividend as the constant is significantly different from zero. Finally, the slope β is significantly lower than one for the low-dividend sample but insignificantly so for the high-dividend sample. This indicates the ability of market makers to benefit from economies of scale in handling the dividends.While Frank and Jagannathan (1998) address the first criticism of Elton and Gruber (1970), namely that ex-dividend-day-premiums are not necessarily a reflection of marginal taxes, Koski and Scruggs (1998) address the second criticism. Namely this is the criticism that short term trading may reduce or eliminate (depending on the level of trading costs) the tax effect on ex-dividend-day prices. Thus the Koski and Scruggs (1998) approach is based on what Allen and Michaely (1995) term a dynamic taxclientele effect which comprises investigating trading volume around ex-dividend days. The argument put forward is that short-term trading, motivated by traders exploiting exdividend day premiums, results in abnormal trading volume. Therefore, even if the existence of a tax clientele can not be inferred from ex-dividend day premiums, it can still be inferred from abnormal trading volume around the dividend payment days. If taxes impact ex-dividend returns then security dealers, who are tax neutral, will increase their trading around ex-dividend days. Also, if low dividend-yield stock is held by dividend averse investors (as predicted by the clientele effect) then it should be associated with ex-day premiums. Under such circumstances security dealers are expected to take long positions to capture the dividends. (They will increase their cumdividend buying and sell at the ex-dividend price, which will drop by less than the value of the dividend they collected). Similarly for high-dividend yield stock, held by investors with preference for dividends, the ex-dividend price is expected to drop by more than the nominal amount of the dividends. In that case, dealers can be expected to take short positions. Furthermore, as US firms were exempt from taxes on 70% of their inter- corporate dividends received during the period investigated, they are also expected to engage in dividend capturing by establishing short positions. Koski and Scruggs (1998) collect data on trading volume by dealers and by individuals/firms for 70 ex-dividend days between November 1990 and January 1991.

The abnormal trading volumes on the ex-dividend day and on the previous day are based on an event window of 11 days centred on the ex-dividend date. Abnormal trading volume is obtained as actual volume less the average volume during normal trading period and is standardised by the standard deviation of the normal trading volume. The means of the standardised abnormal volumes provide

strong evidence that tax-neutral securities dealers engage in short selling of high yield stock around ex-dividend. The study further tests the hypothesis that abnormal trading volumes around ex dividend days are positively related to the dividend yield and negatively related to transaction costs. Results of the ordinary least squares regressions of the standardized abnormal volume on the last cum-dividend day indicate that securities dealers engage in short term trading on cum-dividend days. This is supportive of a dynamic tax clientele effect as per which tax-neutral dealers engage in trading around the ex-dividend date in order to capture tax-driven differences between the ex-dividend capital loss and the amount of dividend paid. Nevertheless, it is precisely this arbitrage activity by securities dealer that can eliminate tax clientele effect on ex-dividend day returns. Thus Koski and Scruggs (1998) address the second criticism of the Elton and Gruber (1970) approach by showing that short term trading could obscure tax clientele effect on ex-day returns. Green and Rdyqvist (1999) address the third criticism of the Elton and Gruber (1970) approach, namely the notion that equity price volatility invalidates inferences about tax effects from ex-dividend-day price behaviour. They do this by studying ex-day price behaviour of Swedish lottery bonds, which are more stable than equity shares, thus reducing noise from ex-distribution price behaviour.Nevertheless, while more stable than equity shares, lottery bonds are similar to equity shares as there are tax implications to whether the bond is sold cum-lottery or ex-lottery. Particularly, if the lottery bond is sold on the cum-lottery day, the extra payment to the seller forms part of the bond price and is treated as capital gains not as accrued coupon payment. Moreover, for the buyer the extra payment is treated as a capital loss and form part of his/her tax basis. The implication of this feature of the lottery bond is that, if tax differentials on capital gains and distributions matter, then ex-lottery returns, like in the case of equities, should reflect the marginal tax rates of their holders.Green and Rdyqvist (1999) note that another advantage of looking at the Swedish lottery bonds is that distributions are tax-exempt. In most cases, where the tax system favours capital gains, factors such as transaction costs of handling dividends can substitute for the effects of taxes, making ex-days behaviour difficult to assess. In the lottery bonds market such factors have an opposite effect to that of taxes for the reason that the tax system favours distributions. These non-tax factors, such as transaction costs, may therefore reduce the effects of taxes on lottery-bonds ex-day price behaviour, but they do not offer potential alternative explanation for them.The data in Green and Rdyqvist (1999) comprise 46 lottery bonds of two types (mixed and sequenced) with between 5-10 years to maturity, trading on the Stockholm Stock Exchange in the period 1986 to 1997. There are 455 lottery payments with 287 lottery days (due to lotteries that pay their coupons on the same days). The sample is sub-divided as per the tax regime at the time when they were issued. The oldest sample, issues pre-1981, has the largest tax advantage as capital losses on these bonds can be fully used to offset tax due on any other income. Cumulative abnormal trading volumes of 10 days around the ex-distribution day are calculated for each tax regime sample. Abnormal volume is calculated as that day’s volume divided by the average daily volume over the period beginning 6 trading days after the previous distribution and ending 6 days before the current distribution. In a similar manner, 20 days cumulative abnormal trading volumes are also obtained. Green and Rdyqvist (1999) find evidence of high abnormal trading volumes for lottery bonds issued under all tax regimes, but the highest abnormal volumes are recorded for the pre-1981 sample. This evidence is supportive of the dynamic tax clientele. There is an increase in trading activity around distribution days due to tax differentials amongst market participants, and this abnormal activity is stronger, the higher the tax benefits attached to the bonds. To further investigate how taxes influence the returns around lottery days, these returns are regressed on the coupon yield as in the following model:P(t+k) - Pt = Pt ( γ0 k) + C ( γ1-1) (5.10)γ0 = E(r) / (1-T) (5.11)

γ1 = -T / (1-T) (2.12)

Where P(t+k) is the price on day (t+k) and Pt is the price on day t. Thus the LHS of Equation (5.10) is the change in price over the period k.Similarly, on the RHS, k is the number of days in the trading period and the coefficient γ 0 is defined by Equation (5.11), where E(r) is the expected after-tax daily return and T is the tax rate faced by the marginal investor. So the first expression on the RHS of (5.10) is the pre-tax expected return over the period k. The second expression on the RHS of (5.10) is the pre-tax change in price that is due to the coupon payment, C. The coefficient γ1 is defined in Equation (5.12), thus the implied tax rate, T, can be obtained from the estimates of γ1.When (γ1-1) is equal to –1 (i.e. γ1 is equal to 0), the pre-tax price on the ex-coupon day falls by C, the value of the coupon, providing evidence of tax irrelevancy. When (γ1-1) is greater than –1 (i.e. γ1 is greater than 0), then the pre-tax price on the ex-coupon day falls by less than C, the value of the coupon. This provides evidence in support of the view that taxes drive coupon payment to be worth less than their nominal amount.When (γ1-1) is less than –1 (i.e. γ1 is less than 0), then the pre-tax price on the ex-coupon day falls by more than C, the value of the coupon. This provides evidence in support of the tax advantage of coupon payments, and is also the hypothesis put forward for empirical testing. To empirically test for the value of γ1, C is added to both sides of Equation (5.10), which is then dividend through by Pt:

R = γ0 k + γ1 C/Pt (5.13)R = [P(t+k) + C – Pt]/ Pt (5.14)

Where R is the pre-tax return over the trading period, k, as defined in Equation (5.14), and C/Pt is the coupon yield. Green and Rdyqvist (1999) run the Weighted Least Squares regressions on the two types of lottery bonds. The findings of a negative γ1 imply that the price on the ex-lottery day falls by more than the value of the coupon. This is consistent with the tax clientele effect in markets where distributions have tax advantages. Thus, to summarise the evidence on the tax clientele effect, Green and Rdyqvist (1999) find support for Elton and Gruber (1970) proposition that clientele effects can be observed from ex-days price behaviour. Nevertheless, the implications from Koski and Scruggs (1998) is that short term trading may eliminate the tax clientele effects on ex-day returns even if investors consider taxes when choosing stock. In contrast Frank and Jagannathan (1998) note that price behaviour around ex-days may be driven by factors other than tax clientele effects.Empirical studies of the signalling hypothesisFrom theory to empiricsThe signalling hypothesis is based on the notion of asymmetric information particularly between managers and investors. Under this assumption dividend changes are valuable in that they convey information about the firm’s prospects. Indeed, Lintner (1956) observes that managers are more willing to raise rather than reduce dividend levels, and this has been widely interpreted as indicating that dividend decreases are associated with negative signals while dividend increases signal positive news. But what precisely is the nature of the information contained in dividend changes?The risk-information hypothesis claims that dividend increases signal risk reduction. Alternatively, as per the cash flow signalling hypothesis, dividend changes contain information about future cash flows. Another opinion is that dividend changes signal permanent shifts in current earnings. In any event, as noted by Allen and Michaely (1995), regardless of the precise information contained in the dividend signal, there are principally two conditions that have to be met in order for the signaling hypothesis to be valid. The first condition requires that market participants understand dividends as signals. For instance, if the unexpected dividend change signals future earnings changes, then market participants should revise

their future earnings forecasts following the dividend announcement. More generally, if unexpected dividend changes are interpreted as signals of new developments in the firm’s prospects, then a price reaction should be observed in the same direction as the unexpected dividend change announced. The second condition that have to be met to validate the signalling hypothesis, is that the dividend change is followed by a change in the same direction in earnings or other firm’s characteristics that the dividend change is assumed to predict. Empirical methods in studies of the signalling hypothesis have therefore focused on assessing the extent to which these conditions are met.The following empirical review looks at each of these two conditions in turn. It begins with studies that are concerned with assessing market interpretation of the dividend signal. This first condition is commonly tested by event studies around the announcement period and by studying analysts’ earnings forecasts revisions. The validity of the second condition is then assessed, by reviewing empirical evidence on actual changes in firm’s characteristics following dividend change announcements.Finally, some empirical studies of the conditional signalling hypothesis are discussed. The conditional signalling hypothesis proposes that the dividend signal is conditional on firm specific characteristics. This implies that both the interpretation of the dividend signal, and actual long-term changes in the firm following the signal, are conditional on firm characteristics. Thus empirical evidence on the conditional signaling hypothesis looks at cross sectional variations in the immediate reaction to dividend announcements. It further looks at variations in long term changes in performance and other characteristics following dividend changes. These cross sectional variations are tested either by comparison analyses or by regression analyses.Interpretation of the dividend signalInterpretation of the dividend signal is typically assessed by event studies around the dividend change announcement period as has been done by numerous papers. Nevertheless, Laux, Starks and Yoon (1998) and Howe and Shen (1998) innovate by studying price reaction of rivals of firms that announce dividend changes. Both these studies use US firms and define the event window as the two days including the day of the dividend change announcement and the previous day. Both also utilise the market model, estimated post event, to generate abnormal returns. Laux, Starks and Yoon (1998) study dividend announcements in the period 1969-1988, but restrict observations to dividend changes of at least 25 percent. They calculate the averages of the cumulative two-day abnormal returns across the sub samples of firms declaring dividend increases and firms declaring reductions. Consistent with the signalling hypothesis, it is found that firms experience significant abnormal reactions at the time of the announcement and in the same direction. Particularly, the mean two-day abnormal reaction of the 217 firms declaring dividend increases is significant at 1.01 percent. In contrast, the mean abnormal reaction of the 105 firms announcing dividend decreases, is significant at –6.35 percent. Further, the findings that the reaction to dividend reductions is stronger than for dividend increases confirm Lintner’s (1956) observation of managers’ particular dislike for dividend cuts. Laux, Starks and Yoon (1998) also try to determine the information that market participants perceive to be contained in the dividend change announcements. They do this by looking at the price reaction of non-announcing firms to dividend change announcements by firms in the same industry. Specifically, it is proposed that rivals’ price reaction should be of the same direction as that of the announcing firm if the dividend change announcement is interpreted as indicating industry-wide information (contagion effect). In contrast, if the announcement is interpreted as signalling a shift in the competitive position of the announcer then the price reaction of rivals should be in the opposite direction (competition effects). The two-day average abnormal return, for 1,243 firms, rivals to dividend-increasing firms, is recorded at 0.05 percent while for 667 firms, rivals to dividend-decreasing firms, the corresponding figure is –0.32 percent. As the price reaction of rivals is in the same direction as that of the announcer, this indicates that announcements are interpreted as containing information about common factors for the industry as a whole.

Nevertheless, Laux, Starks and Yoon (1998) note that dividend change announcements may also contain information about shifts in the competitive balance in the industry. First, the average reactions of rivals to dividend change announcements are not significant, which may be the result of contagion and competition effects offsetting each other. Second, it is found that rivals’ same-direction reactions are strongest for those rivals least likely to be affected by changes in the competitive position of the announcer. Specifically, the most competitive rivals display a significant positive reaction to increase announcements while the least competitive rivals display a significant negative reaction to decrease announcements.Howe and Shen (1998) also investigate market’s interpretation of the nature of the dividend signal by studying non-announcing rivals. The sample used in the study consist of rivals of dividend initiating firms traded on the NYSE/AMEX, in the period 1968 to 1992. The price reaction and analysts’ forecast earnings revisions, following dividend initiation announcements by rival firms in the same industry, are analysed. The average announcement period’s two-day cumulative abnormal price reaction of 3540 rivals is recorded as insignificant at –0.07 percent. Similarly analysts do not revise their earnings forecasts for rivals of announcing firms. The mean, unadjusted, earnings forecast revision across 345 rivals of dividend initiating firms is insignificant at 0.1677, while the mean abnormal forecast revision is insignificant at 0.167121. It is thus concluded that the information contained in the initiation announcement is interpreted as firm specific without any evidence of intra-industry contagion or competition effects. Thus the findings in Howe and Shen (1998) are not fully consistent with those in Laux, Starks and Yoon (1998). Particularly, while Laux, Starks and Yoon (1998) show that dividend change announcements signal information about rivals of the announcers, Howe and Shen (1998) do not support this view. Nevertheless, evidence in Laux, Starks and Yoon (1998) supports the validity of the first condition for the signalling hypothesis to hold, namely that dividend changes are interpreted as signals. As the first condition is shown to be valid, the question is whether the second condition is also met. Particularly for the second condition to be valid, the dividend change announcement should be followed by actual changes in the firm characteristics, which the dividend change is predicted to signal. This is the subject to be discussed next.Actual changes following the dividend change announcementsAssessing actual changes in firms’ characteristics, following dividend change announcements is the subject of various empirical studies. This is for the reason that the findings that dividend change announcements are followed by particular changes in the firm may help in establishing two things. First, it confirms the validity of the signalling hypothesis for the reason that it makes sense to interpret the dividend change as a signal of an unexpected change if indeed it is followed by such a change. Second, it can shed light on the precise nature of the information contained in the announcement. DeAngelo DeAngelo and Skinner (1996) investigate whether dividend change announcements are followed by changes in earnings that are in the same direction. In order to isolate the effects of the signalling hypothesis from other effects that may influence firms’ dividend policy, DeAngelo DeAngelo and Skinner (1996) select firms experiencing a sudden earnings decline after a long period of stable growth. In particular, the sample contains 145 US firms experiencing a decline in annual earnings between 1980 and 1987 after consistent earnings growth over at least nine years. This selection method ensures that the dividend change is a signal of future rather than past changes. The selection method also implies greater need for signalling for the reason that firms that expect the current decline to be corrected in the near future, have to convey this information to market participants.The initial results in DeAngelo DeAngelo and Skinner (1996) do not support the signalling hypothesis of dividends, as there is no indication that dividend increases represent reliable signals that the current earning problem is only temporary. Specifically, it is found that the 99 firms that increased their dividends in the first year of the earnings decline experienced no positive abnormal earnings in the subsequent three years. To further investigate the robustness of these results, DeAngelo DeAngelo and

Skinner (1996) use 135 firms with complete earnings data, and regress the abnormal future earnings on a dividend signal and a number of control variables assumed to help in predicting future earnings. They find the coefficient on the dividend signal to be insignificantly different from zero and this result holds when alternative proxies for the dividend-signal are used. Thus the results from the regression analysis confirm the earlier findings that dividend increases are not a reliable signal of improved future earnings performance.Two possible explanations are offered in DeAngelo DeAngelo and Skinner (1996) for the unreliability of the dividend signal. The first is mangers’ tendency to send overoptimistic signals either naively or deliberately. Second, it is suggested that the cash commitment associated with the dividend increase is relatively small. The median firm’s dividend increase in the year of the earnings decline, amounts to only 3.5 percent of earnings, hence weaker firms can afford to send misleading signals. A similar figure of 5 percent of earnings is recorded in Lipson, Maquieira and Meggison (1998) with respect to the dividend initiations of newly listed firms.Lipson, Maquieira and Meggison (1998) compare the performance of 99 newly public US firms that initiated dividends in the period 1980 to 1986, with similar firms that did not. The argument for the choice of newly listed firms is similar to that of DeAngelo DeAngelo and Skinner (1996) for choosing firms experiencing earnings decline after a long period of earnings growth. Specifically, it is noted that the need for signaling as a way of distinguishing quality may be more significant for these firms. Indeed, it is found that earnings surprises in the first and second years following dividend initiations are significantly greater compared with similar newly listed firms that did not initiate dividends. Furthermore, the dividend cash commitment represents about 5 percent of earnings of the newly listed firms that initiated dividends. This is significantly lower than 8.5 percent of earnings that similar non-initiating firms would have had to commit to, if they wanted to match the dividend yield, dividend to sales ratio or dividend to assets ratio of initiating firms.Thus, Lipson, Maquieira and Meggison (1998) provide support for the view that dividend initiations signal future earnings prospects, as they distinguish one newly listed public firm from another newly listed firms. Nevertheless, it is also shown that dividend initiations do not distinguish newly listed firms from established firms in the same industry. This provides partial support for the signaling theory and for the second condition, namely that the dividend changes are followed by actual changes in the firm’s characteristics. It is, nevertheless, inconsistent with DeAngelo DeAngelo and Skinner (1996), as is the conclusion that dividend initiations are a valid signal of future performance, for the reason that weaker firms would find the implied resource commitment, required in order to match the actions of quality firms, too costly to mimic. Thus more evidence is needed on the question of whether dividend changes are a reliable signal, and this is provided in Benartzi, Michaely and Thaler (1997).Benartzi, Michaely and Thaler (1997) take an empirical approach similar to DeAngelo DeAngelo and Skinner (1996), comparing the unexpected earnings of firms that changed their dividends with those that did not. The sample contains 7186 firm-year observations of 1025 US firms that trade on the NYSE or the AMEX for at least two years during the period 1979 to 1991 and which meet various other requirements. The hypothesis is that firms that increase their dividends in a given year should enjoy positive unexpected earnings in years that follow. Similarly, firms that decrease their dividends in a given year should experience negative unexpected earnings in years that follow. Benartzi, Michaely and Thaler (1997) also investigate variation in the unexpected earnings across dividends increasing firms. The hypothesis is that if signaling is costly, then the larger the dividend-increase, the greater should the unexpected earnings in the following year be.Results in Benartzi, Michaely and Thaler (1997) show a strong contemporaneous correlation between dividend changes and earnings changes. Firms that increase their dividends in year 0, experience earnings increases in that year, which are significantly higher than the mean earnings change of the group of firms that did not change their dividends. Similarly, firms decreasing their dividends, experience

significantly more severe earnings decreases in the same year compared with the group of firms that did not change their dividends. Nevertheless contrary to the signaling hypothesis no correlation is found between the sign and size of dividend increases in a given year, and earnings changes in future years. Furthermore firms that cut dividends in a given year, experience significant earnings increases in the following year. Thus the results in Benartzi, Michaely and Thaler (1997) are supportive of DeAngelo DeAngelo and Skinner (1996), as they also reject the link between dividend changes and unexpected future earnings growth. This rejection of the traditional interpretation of the signaling hypothesis is also the conclusion in Jensen and Johnson (1995). Nevertheless, what set Jensen and Johnson (1995) apart from these studies is that they concentrate specifically on dividend decrease announcements rather than on dividend changes. The study investigates whether firms reducing dividend by at least 20 percent after twelve consecutive quarters of positive, non-decreasing dividends, also experience a decline in earnings.The sample in Jensen and Johnson (1995) consists of 268 observations of 218 reductions and 50 omissions by 242 different US firms in the period 1974 to 1989. It is found that while earnings decline significantly in the period before the dividend cut they rise significantly afterwards. The stock price, nevertheless, is found to drop at the time of the dividend reduction announcement and this is explained by the observation that although the cut marks a turning point in earnings pattern, there are still lingering problems. For instance, it is observed that the earnings level at the end of the second year after the dividend cut is still below its level three years before the cut. The study thus proceeds to assess the nature of the problem more closely.To do that, and to investigate the precise information the reduction announcements contain, Jensen and Johnson (1995) look at a range of changes in various other firm’s financial variables. The patterns of these variables over the three years before and three years after a dividend reduction are examined. Findings indicate that firms use the funds, saved from the dividend cut, to improve their positions. The dividend cuts thus lead to improvements in liquidity position and to reduction in the level of debt. The conclusion is that dividend reductions do not necessarily signal a decline in earnings. Rather such cuts appear to signal the beginning of restructuring activities and a turn around in financial decline.Thus the implications of Jensen and Johnson (1995) are similar to those of Lipson, Maquieira and Meggison (1998) in the sense that both provide some evidence to support the notion that dividend changes are followed by actual changes in the firm. Nevertheless, both also illustrate that dividend changes should not be simply interpreted as signalling future earnings increases or decreases. Furthermore, such a blanket view on the nature of the dividend signal is strongly rejected in DeAngelo DeAngelo and Skinner (1996), and in Benartzi, Michaely and Thaler (1997). Results from these studies clearly call for further investigation into the precise nature of the information contained in dividend change announcements. Alternative hypotheses, which have been put forward, comprise the permanent earnings hypothesis, the cash flow hypothesis and the risk information hypothesis. Some of the relevant empirical work in these areas is discussed below.Permanent earnings, cash flow and risk information hypotheses of dividendsThe permanent earnings hypothesis proposes that changes in dividends do not necessarily signal future growth or contraction in the levels of current earnings. Instead, announcements of dividend changes contained information about permanent as oppose to temporary shifts in current earnings. This view is consistent with the survey findings by Lintner (1956) and with Lee (1996) who finds that the partial adjustment model performs better when the target dividend is expressed as a function of permanent earnings.

Lintner (1956) finds that mangers tend to smooth dividends, and this tendency is reflected in the partial adjustment model. Indeed, the model can be manipulated so as to express dividends in terms of a weighted-average of current and past earnings. Thus as per this model, dividend trends reflect the

smoothed pattern of current earnings, eliminating transitory fluctuations. A signalling theory interpretation of this is that by smoothing out temporary fluctuations in the factors that determine dividends, the dividend pattern reflects the stable pattern of those factors. As it is current earnings, which determine dividend levels in the partial adjustment model, a dividend change has to be the result of a permanent shift in current earnings. The second part of Benartzi, Michaely and Thaler (1997) assesses the hypothesis that dividend changes signal earnings stability rather than future earnings growth. The study compares the likelihood of a dividend-increasing firm experiencing a decline in its following year’s earnings with the probability of a firm that does not change its dividends to experience such an event. The results indicate that compared with firms that maintain their dividend levels, dividend-increasing firms are less likely to experience unexpected declines in earnings at least in the first year after the dividend change. As no correlation is found between dividend increases and future earnings changes, the conclusion arrived at is that dividends do not signal unexpected future earnings increases.Instead, it is concluded that, consistent with Lintner (1956), dividend increases signal that current earnings levels are permanent. This distinction between permanent and temporary changes is also explored in Brook, Charlton, and Hendershott (1998). That study, nevertheless, is based on the hypothesis that dividend changes contain information about cash flow rather than about earnings. This is the cash flow signaling hypothesis, which proposes that dividend changes signal changes in expected cash flows.Brook, Charlton, and Hendershott (1998) investigate this hypothesis and in particular whether dividend changes signal permanent as opposed to temporary changes in firms’ cash flows. For that purpose a sample of non-regulated, US firms is divided into three groups on the basis of expectations regarding changes in cash flows in years 1 through 4 where 1992 is year 029. Classification into groups is then carried out as follows. The first group, the permanent-increase group, contains 101 firms whose cash flows remain at least 30% above year 0 in each of the subsequent four years. The second group, the temporary-increase group, contains 45 firms whose cash flows increase by at least 40% in year 1 but then fall to less than 20% above year 0 level in either of the subsequent two years. The third group, the no-increase group, consists of 34 firms whose cash flows increase by less than 30% over the four-year period and by less than 15% in each year.Results from the comparison analysis in Brook, Charlton, and Hendershott (1998) are consistent with the notion that firms use dividends to signal a permanent increase in cash inflows. Specifically, it is reported that the permanent-increase group’s average dividend per share changed by 16.5 percent in year 0, before the cash flow increase. This is significantly larger than the 6.8 percent change experienced by the temporary-increase group. Furthermore, comparing annual abnormal stock returns, across the three groups, indicates that the dividend signal is understood by market participants. The permanent increase group experiences an average annual stock return, net of the CRSP value weighted index, of 17.5 percent in year 0. This is statistically different from zero, and statistically different from the 6.5 percent experienced by the temporary-increase group. Thus consistent with the cash flow signalling hypothesis, Brook, Charlton, and Hendershott (1998) find a positive link between increases in permanent cash flows, dividend rises and stock price reaction. Firms expecting a permanent improvement in their cash flows, signal this information by increasing their dividends. The market understands the signals and the stock price rises before the actual cash flow increase occurs.Thus while Benartzi, Michaely and Thaler (1997) suggest that dividend changes signal changes in permanent earnings, Brook, Charlton, and Hendershott (1998) find it is permanent cash flows that dividend changes signal. In both cases, nevertheless, dividends are used to signal changes in the pattern of long-term performance. An alternative explanation is that dividend changes signal information about changes in the firm’s risk. This is the risk information hypothesis, which is investigated in Dyl and Weigand (1998). In particular, Dyl and Weigand (1998) distinguish the risk information effect by

investigating whether dividend initiation announcements are followed by reduction in earnings volatility and risk or by earnings increases. The sample in Dyl and Weigand (1998) consists of 240 firms listed on the NYSE/AMEX, and which initiated dividends during the period 1972 to 1993. In order to assess the change in risk following dividend initiations, the total risk of returns, market risk of returns and earnings-per-share volatility, before and after the dividend initiation, are compared. Thus proxies for these variables are calculated for each firm in respect of the period before the dividend initiation and in respect of the period after the initiation.The means and medians for each of these proxies are obtained and the significance of the change from the pre-initiation period to the post-initiation period is assessed. Dyl and Weigand (1998) find that 70 percent of the sample firms have lower variances in the post dividend initiation period. Furthermore, the hypothesis of equal mean variances before and after the dividend initiation is rejected. Likewise, 68 percent of the sample firms have lower market risk as measured by β after the dividend initiation and the difference in the mean β pre and post initiation is statistically significant. There is also evidence to show that earnings volatility declines in the period following the dividend announcement, as the post-initiation earnings volatility is significantly lower compared with the pre-initiation period. In contrast, nevertheless, there is no significant difference in the mean of the standardised earnings per share in the pre- and postinitiation periods. Thus, it appears that announcements of dividend initiations are not followed by increases in future profitability.Based on their findings Dyl and Weigand (1998) conclude that announcements of dividend initiations do not signal enhanced profitability, but instead they are signals of stability. This risk-information hypothesis of dividend signalling is particularly interesting as it highlights a weakness in the bird in the hand argument in favour of generous dividends. Accordingly, the reason that stock price reaction to dividend increases is typically positive, this is not for the reason that dividend cause risk reduction, but for the reason that they are signals of risk reduction and future stability. From the discussion in this sub section it emerges that the nature of the information, conveyed from the dividend change announcement, is ambiguous. The studies by Benartzi, Michaely and Thaler (1997) and Brook, Charlton, and Hendershott (1998) conclude that dividend signal shifts in permanent as opposed to transitory performance. Although the emphasis in the former study is on earnings performance while in the latter it is cash flow, over the long-term these are essentially the same. These conclusions tie-in well with Lintner’s (1956) observation that managers seek to achieve a gradual upward progression in dividends that reflect long-term, permanent changes in performance. In contrast, Dyl and Weigand (1998) find that dividend changes indicate shifts in risk and earnings volatility rather than changes in performance. A possible resolution for this confusion is the idea that dividend changes convey different information to different firms. The reaction to dividend-change announcements therefore depends on particular characteristics of the announcing firm and its circumstances. This hypothesis is termed the conditional signalling hypothesis and is typically investigated by cross sectional comparisons, or by regression analysis where firm characteristics are entered as explanatory variables.The conditional signalling hypothesisResearches have investigated three main factors that may cause variations in the signalling function of dividends across firms or even over time for the same firm. First, such variations may be due to the combination of activities with which the firm engages prior to the dividend change announcement. Second, variations in the meaning and interpretation of the dividend signal may be caused by differences in the environment in which the firm operates. Third, cross sectional differences in the meaning of the dividend signal may be the result of differences in firms’ characteristics. The discussion that follows presents some of the empirical work on each of these three factors, namely, prior activities, the environment and firm’s characteristics.Related to the first factor, the effect of prior activities on the dividend signal, is the idea that the value of the dividend signal depends on the surprise with which it is met by market participants. For instance, a

dividend change announcement that comes after certain activities, such as the publication of earnings data, may be less informative than if such prior activity did not occur. Similarly, a dividend change announcement that follows a particular activity may contain different information than if it came after a different activity. The first issue is dealt with in Balachandran, Cadle and Theobald (1999), while the second issue is the subject of Born and Rimbey (1993).Born and Rimbey (1993) argue that financing activity, undertaken prior to dividend increase announcements, can distinguish dividend increasing firms with future growth prospects from those firms that disinvest. The study investigates this hypothesis by regression analysis methodology of the price reactions to surprise dividend increase announcements. To ensure only surprise and hence informative increases, enter the sample, the selection procedure imposes the restriction that only firms initiating or resuming dividends after at least ten years of omissions are comprised. This selection procedure results in a sample of 490 US firms that have initiated or resumed dividend in the period 1962 to 1989. For these firms the whole of the dividend is taken as unexpected. The sample is then partitioned on the basis of whether the firm has been engaged in financing activity in the twelve months prior to the dividend change announcement. This provides a sub sample of 102 firms that were engaged in prior financing activity, and a sub sample of 388 firms that were not.Born and Rimbey (1993) begins their empirical investigation by running separate regressions for each sub sample, of the reaction to the dividend increase announcement on a constant and the dividend yield35. Results indicate that the intercept is lower for the sub sample of prior-financing firms compared with the sub sample of non-financing firms. This is consistent with the notion that prior financing activity leads to partial anticipation of a dividend increase, which impacts the share price prior to the actual increase announcement. Results also show that for the sub sample of prior-financing firms, the abnormal return per unit of dividend yield is much larger than for the nonfinancing firms. (2.800 as oppose to 1.745). This is consistent with the notion that prior financing activity alters market reaction to dividend increase announcements as it distinguishes firms with good growth prospects from those with poor growth opportunities.Based on these results Born and Rimbey (1993) proceed to assess the effect of the size of the prior financing activity on the reaction to the dividend announcement. Utilising the sub sample of prior-financing firms only, the price reaction to the dividend announcement is regressed on the dividend yield and on the financing yield. The estimated coefficient on the financing yield is shown to be positive and significant. This is taken to indicate that the larger the amount of finance, raised prior to the dividend increase announcement, the stronger is the positive price reaction to the announcement. Thus Born and Rimbey (1993) conclude that a dividend increase announcement that follows prior financing activity, is interpreted as a stronger indication of growth compared with announcements that do not follow such activity. Nevertheless, the prior activity also reveals information and results in anticipation of a dividend change, therefore the actual dividend change announcement has less informative value. This last point is further taken in Balachandran, Cadle and Theobald (1999), who also investigate the effects of prior activities (albeit not financing) on the dividend signalling function. Balachandran, Cadle and Theobald (1999) look at the extent to which interim dividend reductions (IDR) announcements can be anticipated from prior dividend cuts or other factor. It is proposed that anticipations of IDR should lead to weaker price reaction on the announcement date compared to situations where these announcements come as a surprise. To test this proposition, price reactions to IDR announced by 242 non-financial UK firms, in the period 1986 to 1993 are studied. The study, nevertheless, begins by testing the traditional signalling hypothesis assumption consistent with which the IDR announcements should lead to negative price reactions. Indeed, the unadjusted mean abnormal return in the event window around the IDR announcement is found to be negative and significantly different from zero across the five return generating processes. In the next stage of the investigation, Balachandran, Cadle and Theobald (1999) look at differences in price reaction between IDR that follow previous dividend reductions and IDR that

do not. To do this the sample of IDR announcements is divided into 142 First Interim Dividend Reductions (FIDR) and 100 Subsequent Interim Dividend Reductions (SIDR). The hypothesis put forward is that FIDR lead to more negative price reaction than SIDR for the reason that the former provide more information to the market while SIDR are to some extent anticipated. Using the market model, the mean unadjusted price reaction to FIDR announcements is recorded as significant at –0.094 while reaction to SIDR announcements is significant at –0.053. As the difference between these reactions is significant, the results support the hypothesis that price reaction to dividend signals are weaker the more they are anticipated. In the third stage of the investigation, Balachandran, Cadle and Theobald (1999) focus on SIDR announcements. It is proposed that when the subsequent interim dividend reduction is greater than the Prior Final Dividend Reduction (PFDR), the price reaction should be stronger compared with when the SIDR is less than the PFDR. This may be the case if the increased dividend reduction at the interim stage provide further information and is tested by splitting the SIDR sample into two groups. The first group consists of 39 SIDR where the percentage dividend reduction is greater than the percentage PFDR, and the second group consists of 61 SIDR where the percentage dividend reduction is less or equal to the percentage PFDR. Using the market model, the mean unadjusted reaction when the SIDR is greater than the PFDR is –0.098, while when the SIDR is not greater than the PFDR the mean reaction declines to –0.025. Thus the results are supportive the proposition that price reaction should be stronger, the stronger the surprise.In the final stage of the investigation, Balachandran, Cadle and Theobald (1999) search for factors that could explain cross sectional differences in price reactions to IDR. For this purpose, the cumulative abnormal returns around the IDR, generated from the market model, is regressed on various variables that are hypothesised to impact the surprise in the IDR. Results indicate that, consistent with signalling hypothesis, the price reaction to the IDR is significantly related to the size of the reduction. Furthermore, there are mixed results about the significance of changes in interim earnings in influencing price reaction. This is consistent with the view that the dividend signal is valuable for the reason that the information in the earnings change is a noisy signal of future performance. The regression results also support the conditional signalling hypothesis and the notion that cross sectional differences may result in variations in the signalling function of dividends. Particularly, the price reaction is significantly influenced by whether the firm has previously reduced its dividends and by the gearing ratio. The environment in which the firm operates also appears significant as the surprise in the IDR and thus the price reaction to it, are influenced by prior dividend reductions by other firms in the industry.The impact of the environment in which the firm operates on market interpretation of the dividend signal is also the subject in Impson (1997). Nevertheless, the emphasis there is on differences between regulated and unregulated firms. The hypothesis is that, due to the particular circumstances faced by utilities, dividend reduction announcements by these firms result in stronger reaction than in the case of other firms. The study uses 262 regulated and unregulated US firms declaring dividend reductions/omissions between 1974 and 1993 and the Weighted Least Squares cross sectional regression approach. Thus the price reaction to the dividend announcement is regressed on a regulated-firm dummy and on control variables.Results in Impson (1997) indicate that regulated utilities experience significantly more severe reaction to dividend reduction announcements compared with unregulated firms. It is suggested that the surprise in the dividend reduction announcements may be greater for regulated utilities as these firms have traditionally been associated with high yields and stable pattern of dividends. Furthermore, the coefficients on the Tobin’s Q and size control variables are positive and significant. The former indicates that the higher the over-investment, the more negative is the price reaction to the dividend cut announcement. The latter indicates that the smaller the firm, the more severe is the reaction. Thus,

consistent with the conditional signalling theory, firm’s characteristic and environmental factors are found to be significant in explaining price reaction to dividend change announcements.The significance of Tobin’s Q and firm size recorded in Impson (1997), as well as the significance of the gearing level recorded in Balachandran, Cadle and Theobald (1999), are suggestive of a link between firms’ characteristics and the dividend signal. Akhigbe and Madura (1996) investigate this issue further, by assessing cross sectional variation in long-term price performance following dividend change announcements. The study is based on a sample of US firms announcing dividend changes during the period 1972 to 1990. Nevertheless, prior to assessing cross sectional variations in long-term performance following dividend change announcement, a basic signalling hypothesis’ prediction is tested. This is the prediction that dividend increase signals should be realised by improvements in long-term performance while decreases should be realized by future decline in performance. For this end, the mean long-term price performance of 128 firms initiating dividends is compared to the mean for the 299 firms omitting dividends. Indeed, it is found that firms tend to experience a favourable share price performance, over the longer term, following dividend initiations and unfavourable performance following omissions.To assess cross sectional variations in long-term price performance following dividend initiations and omissions Akhigbe and Madura (1996) regress the 36-month cumulative abnormal return on firm characteristics and the size of the dividend change. Results of this procedure indicate that larger cuts in dividends are associated with more severe long-term price performance. Further, the coefficient on the past profitability measure is negative and significant in the initiation sample, suggesting that firms with inefficient management improve their performance following dividend initiations. With regards firm size it is found that smaller firms tend to perform significantly better in the three years following dividend initiations while large firms tend to perform significantly worse following dividend omissions. Finally, long term reaction to dividend initiations is influenced by the Tobin’s Q measure, implying that firms that over-invest perform significantly better following dividend initiations. These findings are consistent with the conditional signalling hypothesis and with the findings in Gombola and Liu (1999). Gombola and Liu (1999) explore the link between Tobin’s Q and the short-term price reaction to dividend increase announcements. In particular, the study analyses the price reaction to 196 Special Designated Dividend announcements made by US firms between 1977 and 1989. It is hypothesised that firms facing low investment opportunities, with low Tobin’s Q, should experience stronger price reaction to the announcement of Special Designated Dividend. This is consistent with the signaling hypothesis, for the reason that the surprise in the special dividend announcement should be greater for firms with little investment opportunities. Indeed the event study methodology finds that the mean three-day cumulative abnormal return around the Special Designated Dividend announcement for the low Tobin’s Q sample is positive and significant. Nevertheless, the mean price reaction for the high Tobin’s Q sample is insignificantly different from zero, while the mean difference between the two groups is significant.The approach in Gombola and Liu (1999) is based on an earlier study by Lang and Litzenberger (1989). Both studies investigate the validity of the conditional signalling hypothesis with respect to Tobin’s Q, but while the former focuses on price reaction to special dividend, the focus of the latter in on substantial changes in regular dividends. As per the conditional signalling hypothesis the reaction to substantial dividend change announcements should be larger for firms with low investment opportunities. The rationale for this is explained as follows. Investors expect an increase in cash flows for firms with good investment opportunities and they also expect these firms to announce dividend increases to signal this. Therefore the reaction to dividend increase announcements should not be strong for high Tobin’s Q firms while the reaction to announcements of substantial dividend cuts should be strong. In contrast firms without high investment opportunities are not expected to enjoy an increase in cash flows, thus large dividend increases or decreases are not expected for low Tobin’s Q firms. If such dividend changes are announced, market reaction should be strong. Lang and Litzenberger (1989)

therefore predict that if the price reaction is measured as the average to all dividend changes (increases and decreases), the average reaction in the case of low Tobin’s Q firms should be stronger than for high Tobin’s Q firms. To test this prediction the study utilises 429 substantial dividend change announcements made between 1979 to 1984 by US firms. To ensure the dividend change is substantial, a restriction is imposed where the absolute value of the percentage dividend change of each observation must be greater then 10%. The average reaction on the day of the announcement for firms with Tobin’s Q less than unity is recorded as 0.011, as opposed to 0.003 in the case of high Tobin’s Q firms. Further, the 0.008 difference in the mean reaction between low and high Tobin’s Q firms is highly significant. Nevertheless, Lang and Litzenberger (1989) point out that a stronger price reaction to substantial dividend changes by low Tobin’s Q firms compared with high Tobin’s Q firms is also consistent with the over-investment hypothesis. Large dividend increases by firms with low investment opportunities reduce the potential for over-investment by these firms. Such announcements should therefore be received with more positive reaction compared with reaction to similar changes by firms that face many investment opportunities. Similarly, large dividend decrease announcements by firms with low investment opportunities indicate an increase in the probability of over investment by management. Such announcements by low Tobin’s Q firms should therefore be met by more severe reaction. Thus similar to the conditional signalling hypothesis, the overinvestment hypothesis also predicts a stronger reaction to substantial dividend increases or decreases by low Tobin’s Q firms. To distinguish between the conditional signalling hypothesis and the over- investment hypothesis, Lang and Litzenberger (1989) further partition their sample of low and high Tobin’s Q groups into dividend increase and dividend decrease announcements. The reaction to announcements of substantial dividend decreases is the key to distinguishing between the two hypotheses. Particularly, the conditional signaling hypothesis predicts strong reactions to dividend decreases regardless of the firm’s Tobin’s Q, which is due to the negative information such announcements contain regarding future expected cash flows. In contrast, the over-investment hypothesis predicts that the reaction to dividend changes will always be greater for low Tobin’s Q firms for the reason that the potential for over-investment in the case of firms with little investment opportunities is greater.Based on this distinguishing feature between the two alternative hypotheses, Lang and Litzenberger (1989) compare the mean price reaction to dividend decrease announcements. They find that the average reaction to dividend decreases by high Tobin’s Q firms is insignificant, at –0.003, while for low Tobin’s Q firms the reaction is significant at –0.027. These results indicate that there is a significant difference in the price reaction to dividend cut announcements between firms with high and low investment opportunities. These findings are consistent with the over-investment hypothesis and inconsistent with the conditional signalling hypothesis. Similar results in support of the over-investment hypothesis over the conditional signalling hypothesis are also obtained from comparing the post-announcement revisions of analysts’ current earnings forecasts.Thus the results in Lang and Litzenberger (1989) in favour of an agency theory based explanation for market reaction to dividend changes contradict the conclusions in Gombola and Liu (1999) in favour of conditional signalling theory. Nevertheless, the results in Akhigbe and Madura (1996), Impson (1997), and Balachandran, Cadle and Theobald (1999) are consistent with Gombola and Liu (1999). These studies show that firms’ characteristics, and in particular investment opportunities (Tobin’s Q), are significant in determining how the dividend signal is interpreted. Impson (1997) and Balachandran, Cadle and Theobald (1999) further illustrate that environmental factors, such as whether the firm is regulated or the dividend behaviour of other firms in the industry, also influence the price reaction to the dividend signal. Finally, Balachandran, Cadle and Theobald (1999) and Born and Rimbey (1993) show that activities undertaken by the firm prior to declaring dividend changes have implications for how this signal is interpreted. In particular, Balachandran, Cadle and Theobald (1999) show that prior activities

such as past dividend announcements by the firm, influence the amount of surprise and hence the value of the dividend signal. Similarly, Born and Rimbey (1993) show that past activities such as prior financing can distinguish dividend initiating firms that signal quality from dividend initiating firms that disinvest.

Empirical studies of the agency theory of dividendsFrom theory to empiricsAgency theory predicts that managers abuse their position as agents of the firm to appropriate benefits to themselves. A number of studies have investigated the validity of this assumption. Opler, Pinkowitz, Stulz and Williamson (1999) show that managers tend to accumulate excess cash when they have the opportunity to do so. Nevertheless, they also find that firms with excess cash do not use it to over-invest as predicted by agency theory. There is also no evidence of reluctance by managers to return cash to shareholders in the form of dividends when investment opportunities are low.Similar conclusions are also reported in Long, Malitz and Sefcik (1994) who investigate the validity of the agency cost of debt. In particular, the study investigates the under investment problem, which predicts that firms will increase dividends following the issuance of debt as a means of expropriating wealth from debt holders to equity holders. Nevertheless, Long, Malitz and Sefcik (1994) find no evidence to support the view that firms act in a manner consistent with the wealth expropriation hypothesis. It is therefore concluded that reputation has greater value to the firm and its management than the value of the benefits to be obtained by a one off wealth expropriation.If reputation is significant to managers and acting as predicted by agency theory can harm their reputation, then it may be in managers’ interests to show that the firm is free of potential agency problems. One way for managers to create reputation, particularly in countries with poor protection for minority shareholders, is by paying dividends which signals decent treatment of minority shareholders. This idea is developed in La Porta, Lopez-de-Silanes, Shleifer and Vishny (2000) who term it the substitute model of dividends. Nevertheless, La Porta, Lopez-de-Silanes, Shleifer and Vishny (2000) reject the substitute model in favour of what they term the outcome model of dividends. In the outcome model dividends are the outcome of effective pressure by minority shareholders and therefore higher payout ratios tend to be observed in countries with good protection for minority shareholders.Whether the motivation to pay dividends is due to the need by insiders to create reputation for good treatment of minority shareholders, or is the outcome of pressure by minority shareholders, dividends derive their value from reducing agency problems. Dividends can reduce agency problems by reducing the free cash flows (Jensen, 1986) or by forcing the firm to the capital market thus inducing capital market monitoring of the firm and its management (Easterbrook, 1984). Rozeff (1982) incorporates the agency related value of dividends into a model, which he calls the cost minimization model and which allows for empirical testing of the agency theory of dividends.There are, nevertheless, other ways to control agency costs which may be less costly than the dividend device. For instance, growing firms are likely to resort to external financing on a regular basis, and thus subject themselves to external monitoring even without using dividends. Similarly Jensen (1986) proposes that agency costs may be controlled by debt. Other alternatives to dividends in controlling agency costs comprise managerial ownership and management compensation schemes that are designed to align the interests of managers and outside shareholders. Indeed, Fama and French (2001) propose that the declining trends in dividends by US firms may be due to growing use of stock options by managers, which lower the benefits of dividends in controlling agency costs. Thus the availability and cost of non-dividend monitoring mechanisms may impact the degree to which the dividend device is used and thus the validity of the cost minimization model.In light of the above discussion, the following selective review of empirical studies of the agency theory of dividends is dividend into two sub sections. The first sub section describes some studies of the cost

minimization model. The second sub section is devoted to some of the studies that seek to assess the degree of substitutability amongst the various methods by which firms can control agency costs.The Cost Minimization ModelRozeff (1982) introduces the cost minimization model as per which the optimal dividend payout is at the level that minimizes the sum of transaction costs and agency costs. Transaction costs are incurred when external finance is raised, which may be necessary when internal funds are paid out as dividends. Agency costs are the costs associated with the agency problem. These costs can be reduced by the payment of dividends as suggested by Easterbrook (1984).Rozeff (1982) tests the cost minimization model using Ordinary Least Squares cross sectional regression and 1000 US firms with data relating to 1981. Transactions costs faced by the firm are measured by past and forecasted growth rates in revenues and by the firm’s beta, which represents risk. An agency cost variable is taken as the natural logarithm of the number of outside shareholders, which measures ownership dispersion. It is expected to be positively related to the payout ratio for the reason that the more dispersed is the ownership structure, the more difficult monitoring becomes. An inverse agency cost variables is the fraction of the firm owned by insiders. It is expected to be negatively related to the payout ratio, for the reason that by increasing their holdings in the firm, managers align their interests with that of outside investors. Rozeff (1982) shows the estimated coefficients on the five explanatory variables to be significant and to bear the signs predicted by the cost minimization model.Innovations on the Rozeff’s (1982) model can basically be split into three types, including adding new variables, improving the empirical technique or focusing on particular types of firms. Llyod, Jahera and Page (1985) innovate by adding a new variable, namely firm size, and by refining the empirical approach. The empirical approach follows Rozeff (1982) by employing the Ordinary Least Squares method and data on 957 US firms for 1984. Nevertheless, innovation comes in an attempt to reduce correlation between the explanatory variables by regressing the agency variables on the size variable, and using the residuals obtained in place of the original agency variables. Results indicate that after multicollinearity is properly controlled for, the cost minimization model is still valid. All the explanatory variables appear significant and enter the model with the expected signs. Further, the study concludes that size is also an significant explanatory variable.Schooley and Barney (1994) also innovate on the Rozeff’s (1982) model by adding a new variable, namely the squared of insider holding, and by attempting to improve the technique. In particular they relax the linearity assumption with regard insider holdings and assess whether the relationship between this variable and the optimal payout ratio may be more complex than originally assumed. Rozeff (1982) suggests that the optimal payout ratio should decline monotonically with rises in insider ownership. As insider ownership increases, insiders’ interests are more aligned with that of shareholders, hence agency costs are reduced and the need for the dividend tool to control these costs is lessened. Schooley and Barney (1994) suggest that at low level of ownership the relationship between dividends and insider ownership is as predicted by Rozeff (1982). Nevertheless, when the level of insider ownership reaches a certain level, further increases cause agency costs to start rising and the need for the dividend control tool becomes necessary. This occurs due to two reasons. First, when high proportion of their wealth is invested in the firm, insiders become less diversified. They then tend to evaluate projects based on total risk and may reject projects even when these are justified based on systematic risk. Second, when insiders hold substantial percentage of voting rights they achieve a sufficient level of control that diminishes their risk of being replaced. The results of the Ordinary Least Squares regression, using 1980 data of 235 industrial US firms provide support for the cost minimization model. Further, the relationship between insider ownership and the firm’s dividend policy appears to confirm to expectation.Moh’d, Perry and Rimbey (1995) innovate on the cost minimization model by adding a number of new variables, and by using Weighted Least Square methodology and panel data for 341 US firms over 18

years from 1972 to 1989. The aim is to test whether variation in payout ratios across time can be explained by changes in the agency cost/transaction cost structure. To capture the dynamics in the dividend process, variables are not aggregated and the previous period’s dividend payout ratio is added to the RHS of the model. The study also decomposes beta, the systematic risk, into its components to assess more directly the separate effects of financial leverage, operating leverage and the intrinsic business risk. Further, institutional ownership is added as an explanatory variable. As per agency theory, the presence of institutional investors should reduce payout ratios due to their role in monitoring managers’ activities as suggested in Shleifer and Vishny (1986). Nevertheless, if investors consider taxes on dividends and on capital gains, then the presence of institutional ownership should increase the firm’s payout ratio as suggested in Redding (1997). Indeed, results with respect to institutional holdings indicate support for the tax hypothesis of dividends. The study also provides supports for the dynamic nature of the dividend process as per which firms adjust their dividend each year, as new information becomes available. Holder, Langrehr and Hexter (1998) add two new variables to the cost minimization model. Free cash flow is added as an agency proxy and the firm’s focus is added to test stakeholder theory. Stakeholder theory proposes that non-investors that have implicit contracts with the firm, such as employees, customers, suppliers and others, also influence the firm’s decisions including its dividend policy decisions. Particularly, dividend policy can create value for the reason that by reducing its payout ratio, the firm signals to implicit claimants an increase in its ability to meet implicit claims. Using panel data for 477 US firms each with eight years of observations from 1983 to 1990, the study provides support for both the agency model and stakeholder theory.All the innovations to Rozeff (1982) reviewed above focused on adding new variables to variants of the cost minimization model. In contrast the innovation in Hansen, Kumar and Shome (1994) is with respect to the type of firms for which the model is applied, namely the regulated electric utility industry. It is proposed that the agency theory of dividend should fit particularly well to the behaviour of regulated firms for two main reasons. First, agency conflicts in regulated firms are predicted to be particularly severe as they comprise conflicts between shareholders and regulators. Nevertheless, by paying dividends the regulated firm exposes its managers and its regulators to capital market monitoring, which in turn contributes to reducing agency costs. Second, it is proposed that the costs associated with dividend-induced capital monitoring are lower for utilities for the reason that direct flotation costs of issuing new equity can be passed on, at least in part, to ratepayers. The study begins by comparing the mean payout ratios of utilities and S&P400 industrial firms over the period 1981-1985 and over the period 1986-1990. Results are consistent with the prediction that utilities have higher payout rates as in both periods the averages payout ratio of utilities are significantly greater than that of non-regulated firms. Further, results of cross sectional Ordinary Least Squares regressions offer support for the monitoring rationale for dividends in the case of regulated firms. Indeed, it is concluded that the monitoring rationale for dividends could be the answer to the puzzle of why firms often issue new equity while at the same time paying large dividends.The innovation in Hansen, Kumar and Shome (1994) of applying the cost minimization model to a particular type of firms, is also the approach in Rao and White (1994). Nevertheless, while the former study applies the model to firms for which the monitoring rationale for dividend is predicted to be particularly suited, the latter study applies the model to the opposite type of firms. Thus, Rao and White (1994) apply the cost minimization model to private firms for which the monitoring rationale for dividend is predicted to be particularly unsuitable. Indeed, it is noted that the motivation to use dividend as an agency-cost controlling device may be less significant for private firms due to less agency problems. Moreover, as private firms do not participate in the capital market, the rationale for dividends as inducing further equity issues leading to capital market monitoring also loses some of its momentum.Another innovation in Hansen, Kumar and Shome (1994) is that they incorporate taxes into the model, as it is argued that tax savings considerations may contribute to private firms’ preference for low payout

policies. This is explained as follows. Owners/managers of private firms receive returns in the form of either salary, which is a tax deductible business expense, or dividends. There is therefore incentive to minimize the dividends and to increase the salary component. Although limits are imposed on the amount of salary that can be paid, owners may still have incentive to minimize dividends due to tax differentials between dividends and capital gains. Nevertheless, it is noted that if the Internal Revenue Service suspects that a firm is retaining its earnings for the purpose of avoiding taxes, it may take steps to impose Accumulated Earnings Tax on that firm. Thus AET is added to the model to proxy for the tax cost of not paying dividends. The third innovation in Rao and White (1994) is the empirical technique, which is the limited dependent variable regression as opposed to Ordinary Least Squares. The rationale for this is as follows. The sample comprises 66 private US firms that had been challenged in court by the Internal Revenue Service for Accumulated Earnings Tax liability between 1928 and 1988. Nevertheless, as the Internal Revenue Service is unlikely to challenge firms with high payout ratios, the sample excludes firms with payout ratios greater than some high latent level. This implies that the dependent variable of the firms comprised in the sample is not normally distributed but truncated from above, and Ordinary Least Squares method is inappropriate.Results in Rao and White (1994) show that the agency cost variables, namely the fraction of shares held by insiders, and the dispersion of ownership as reflected in the number of shareholders, influence the payout ratio in the manner predicted. This suggests that the agency cost argument for dividends appears applicable even for private firms that do not participate in the capital market. It is suggested that by paying dividends private firms can still induce monitoring by bankers, accountants and tax authorities. The proxy for the Accumulated Earnings Tax cost is also found to be significant and to enter the model with the predicted sign. This suggests that tax cost considerations influence the dividend decisions of private firms. Thus firms that are likely to be challenged and charged by the Internal Revenue Service for Accumulated Earnings Tax, reduce the probability of facing such costs by increasing their payouts. Rao and White (1994) demonstrate the relevance of agency considerations to dividend decisions not merely in the most likely cases as shown in Hansen, Kumar and Shome (1994) for regulated firms, but rather in the least likely cases such as private firms. Hansen, Kumar and Shome (1994) contribute to the discussion by emphasizing the use of dividend to control conflicts beyond shareholders and managers, such as conflicts with regulators. In the same trend, Holder, Langrehr and Hexter (1998) discuss how dividend can be used to control conflicts relating to non-investor stakeholders in the firm.The complexity of agency behaviour, and in particular how insider holdings influences agency costs, is emphasized in Schooley and Barney (1994), while Moh’d, Perry and Rimbey (1995) address the dynamic nature of the agency/transaction cost structure. The latter study also illustrates the significance of tax considerations in determining the payout ratio of firms as reflected in the positive and significant impact of institutional investors on payout levels. The significance of incorporating tax into the model is also picked-up in Rao and White (1994), while the significance of firm size is shown in Holder, Langrehr and Hexter (1998), Moh’d, Perry and Rimbey (1995), and Llyod, Jahera and Page (1985). One thread, nevertheless, common to all the above-mentioned studies is that they provide support for the monitoring rationale of dividend and for Rozeff’s (1982) cost minimization model. Nevertheless, as predicted by tax and transaction cost theories, and indeed as incorporated in the cost minimization model, using the dividend monitoring device is not costless. It has therefore been suggested by a number of studies, that the extent to which the dividend-monitoring device is used to control agency cost should display sensitivity to the availability of alternative mechanisms. Some of the empirical work in this area is reviewed in the next sub section.The partiality of the monitoring rationale for dividends (or substitutability among dividend and non-dividend mechanisms for controlling agency costs)

Easterbrook (1984) points to two significant implications of the monitoring rationale for dividends. First, it is noted that dividends must influence the firm’s financing policies, if the reason that they are paid is to drive the firm to the capital market. Second, as the dividend-monitoring device is costly, the presence of alternative mechanisms that limit agency problems, or conditions that force the firm to the capital market, should reduce the use of the dividend device. The implications of these two points are that the dividend rationale is applicable only in some cases. Further, in these cases, the dividend and capital structure decisions are endogenous variables and should be modelled as a pair of simultaneous equations in a signal model. Noronha, Shome and Morgan (1996) test whether the presence of growth-induced monitoring or other non-dividend devices that limit agency problems, lessen the monitoring role of dividends and the simultaneity of dividend and capital structure decisions. For that purpose a sample of 341 US industrial firms is stratified as per the presence of growth opportunities as measured by Tobin’s Q. A firm with Tobin’s Q value above the sample average is classified as high on growth opportunities. The sample is then further stratified as per the presence of alternative non-dividend monitoring mechanisms. A firm is classified as possessing alternative non-dividend monitoring mechanism if it satisfies two conditions. First, the firm has to have an above average incentive component in its managerial compensation package, which serves to align managers-shareholder interests. Second, the firm has to have a single large outside shareholder holding at least 5% of the firm’s equity, for the reason that a large outside shareholder serves as an external monitor and a potential take-over threat.The stratification procedure, in Noronha, Shome and Morgan (1996), results in two sub-samples. Sample A consists of 131 firms with high alternative control mechanisms and/or growth-induced capital market monitoring. Sample B consists of 210 firms with low alternative control mechanisms and low growth opportunities. The sample data is pooled from the period 1986 to 1988 following a Chow test that fails to reject the null of stability. The monitoring rationale for dividends is tested by an Ordinary Least Squares regression of a variant of the cost minimization model, where firm size is taken as a proxy for transaction costs. Results for group A are weak, as none of the agency cost/transaction cost structure variables are insignificant. In contrast, results for group B support the cost minimization model as the coefficients on all the variables bear the expected signs, and all but the coefficient on firm size, are significant.In the second part of the study, Noronha, Shome and Morgan (1996) test for simultaneity between dividend and capital structure decisions. It is predicted that simultaneity should be evident only in cases where the dividend monitoring rationale applies. Indeed, results of a Three Stage Least Squares tests show no evidence of simultaneity of dividend and capital structure decisions for group A. The equity ratio explanatory variable in the payout equation, and the payout variable in the equity ratio equation are both not significant. In contrast, for group B both, the equity ratio variable in the payout equation and the payout variable in the equity ratio equation are negative and significant. These results as well as the results from testing the validity of the cost minimization model, support the partial explanation of the monitoring rationale for dividends.More support for the partial explanation of the monitoring rationale for dividends is provided by Johnson (1995), who studies 129 straight debt offerings by NYSE/AMEX industrial firms between 1977 and 1983. Nevertheless, while in Noronha, Shome and Morgan (1996) alternative agency cost controlling devices comprise growth opportunities, management incentive schemes and a large outside shareholder, in Johnson (1995) it is debt. In particular it is shown that dividends and debt are alternative devices to reduce the agency problem associated with free cash flow (Jensen, 1986). This is for the reason that both debt and dividend signal a commitment to pay out cash and both may increase visits to the capital market thus inducing capital market monitoring of management’s actions.Indeed, the study finds the average two-day excess return around the debt issue announcement day is positive and significantly different from zero for low payout firms (0.78 percent). This is interpreted as

implying that the issue is indicative of reduced agency problems. Nevertheless, for high payout firms, the average excess return on the debt issue announcement is insignificantly different from zero (-0.18 percent). Johnson (1995) further utilizes a Weighted Least Squares methodology to regress the two-day excess return around the debt issue announcement day on a payout variable. The results from this procedure also support the notion that dividends and debt are alternative mechanisms for controlling agency problems. Particularly, the intercept is positive and significant suggesting that the price reaction to debt issue announcement is generally positive. Nevertheless, the coefficient on the payout variable is negative and significant suggesting that the reaction to the debt issue announcement is weaker for dividends paying firms. These effects are stronger when the regression is run on a sub sample of 64 low growth firms but weaker when run on a sub sample of 65 high growth firms. This is consistent with the view that the potential for wasting free cash flow is greater when profitable investment opportunities are low. Thus Johnson (1995) provides further support for the conclusions in Noronha, Shome and Morgan (1996) regarding the substitutability of growth opportunities and dividend as agency costs controlling devices. Johnson (1995) illustrates the significance of debt as an alternative to dividends in controlling agency costs. Crutchley and Hansen (1989) make the same point and suggest that leverage can achieve these results for the reason that debt finance reduces equity financing and hence manager-shareholder conflicts. Crutchley and Hansen (1989) further note that manager-shareholder conflicts may also be reduced by increasing insider holdings. Nevertheless, the crucial point in the study is the realization that each of the three agency control devices, namely managerial ownership, leverage and dividends, is costly. For instance, while increasing management’s ownership helps to align manger-shareholder interests, it also increases the proportion of the manager’s total personal wealth, which is invested in the firm. As the manager suffers increasing lack of diversification, she will be more risk averse even when this is not in line with shareholder interests.To test the agency theory of managerial ownership, leverage and dividends, Crutchley and Hansen (1989) use 603, US industrial firms for the period 1981 to 1985, and Ordinary Least Squares analysis. Particularly, each of the three policy decisions is regressed on five firm’s characteristics that are hypothesized to influence the levels of the costs associated with each policy. These explanatory variables comprise firm diversification, earnings volatility, flotation costs, advertising and R&D expenditure, and firm size. The results support the notion that managers employ a mix of policies including leverage policy, dividend policy and managerial ownership incentives in an effort to control for agency costs in the most efficient manner. The precise combination of policies varies across firms and is determined by firm’s characteristics.First Crutchley and Hansen (1989) find that managers of diversified firms bear relatively lower costs in increasing the percentage of their wealth invested in the firm’s equity. Thus diversified firms tend to use more of the managerial ownership device and less of the debt and dividend devices to control agency costs. Second firms with volatile earnings face higher bankruptcy risk thus managers reduce leverage and increase dependency on managerial ownership and dividends. Third firms with volatile stock expect to pay higher underwriting fees when issuing new equity thus they tend to increase the use of managerial ownership and leverage, but avoid using dividends. Forth, firms with high R&D expenditure have more freedom to engage in wealth expropriation from both debt and share holders, thus these firms tend to use less debt and dividends and more managerial ownership. Fifth, large firms face lower bankruptcy costs and lower flotation costs on the issue of new equity, while managers of these firms find it more expensive in terms of diversification costs to increase their percentage holdings. Thus large firms tend to rely more on the debt and dividend policy devices and less on managerial ownership. Similar to Crutchley and Hansen (1989), Agrawal and Jayaraman (1994) also investigate the substitutability between leverage, dividends and management ownership in controlling agency costs. The study utilizes 71 industry-size matched pairs of all equity and levered firms for the year 1981, and an

Ordinary Least Squares regression analysis. Specifically, proxies for dividend policy are regressed on leverage, managerial ownership, an interaction term, and on two control variables including free cash flow and growth. Results show the coefficient on the leverage dummy, to be negative and significant, which is consistent with the prediction that all-equity firms follow a policy of higher payout ratios than levered firms. Similarly, consistent with the prediction that firms with lower insider ownership adopt higher payout ratios, the coefficient on this variable is reported as negative and significant. This negative correlation between dividends and insider ownership is stronger in all-equity firms as observed by the positive and significant coefficient on the interaction term between leverage and insider ownership.Bathala and Rao (1995) introduce Board composition as a possible agency-cost controlling device. They investigate the interrelation between Board composition, insider ownership, dividends and leverage as alternative mechanisms for reducing manager-shareholder conflicts. It is argued that outside directors on the Board can reduce conflicts due to their independence and due to the need to maintain reputation in the market for their services. To test this, 261 non-regulated, US firms are used in a cross sectional Ordinary Least Squares regression of a measure of Board composition on alternative agency-cost control devices and on a set of control variables. The results from this procedure show that the alternative agency-cost control devices, including insider holdings, dividends and leverage, have a negative and significant impact on the fraction of outside directors on the Board. These findings are consistent with the notion of firms relying on a mix of alternative mechanisms, including Board composition, to control agency conflicts.Bathala and Rao (1995) note that alternative mechanisms may control different aspects of agency conflicts and that each mechanism may have other, non-agency-related benefits, associated with its use. Empirical work appears to confirm the presence of substitutability amongst various mechanisms including dividends, leverage, managerial ownership and incentive schemes, the presence of a large shareholder, growth, and outside directors on the Board. In the face of these many alternatives, the agency related value of dividends is still unclear. CONCLUSIONS AND PROMISING RESEARCH IDEASThe empirical literature has recorded systematic variations in dividend behaviour across firms, countries and time, as well as in the type of dividend paid. Such systematic variations are inconsistent with Miller and Modigliani’s (1961) irrelevancy theory, but can be expected in imperfect markets. Indeed, once the assumptions underling the irrelevancy theory are relaxed, it may be unreasonable to expect that dividends will have no effect on expected earnings, investment decisions or on the firm’s risk. If dividend policy influences any of these factors, then it is also likely to affect value. Precisely how dividend policy affects value, in the presence of market imperfections, is the subject of various dividend theories, which together form the dividend controversy. The aim of this chapter was to take stock of the generic theories that have evolved under market imperfections such as transaction costs, taxes, information asymmetries and agency conflicts. It was also intended to review the main empirical methodologies and evidence collected so far, in an endeavor of clarifying where the dividend controversy stands today, after four decades of debate.The generic dividend theories comprise the transaction costs theory, the tax hypothesis, the bird in the hand argument, and the signaling and agency theories. The transaction cost theory of dividends is based on transaction costs and control considerations that are associated with paying dividends and then resorting to external finance to fund investments. Also incorporated under this theory are pecking order considerations, which are based on information asymmetries and which become relevant if dividends are paid and external finance raised. Thus, the transaction cost theory of dividends basically suggests that firms should utilize retained earnings to the extent possible before paying out dividends.The tax hypothesis proposes that government distortions by way of taxes have significant implications for dividend policy and firm value. Thus the tax hypothesis generally states that due to differences between taxes on dividends and on capital gains, generous dividends reduce wealth. Accordingly, the

share prices of firms that adopt high payout policies will reflect this tax disadvantage. The underlying assumption here is that all investors are taxed the same and that dividend income is taxed more heavily than capital gains. Alternatively, if there exist tax-based clienteles for low and high dividend policies, or if transaction costs are not too high as to prohibit active trading, then tax effects on prices should disappear. The bird in the hand argument is the traditional rationale for generous dividends, and is based on the idea that dividends reduce risk for the reason that they bring shareholders’ cash inflows forward. This argument, nevertheless, is commonly repudiated by the assertion that the risk of the firm comes from the investments in which it is comprised, not from how the proceeds from these investments are distributed. A more credible argument in favor of dividends is the signaling theory, which is based on information asymmetries between managers and outside shareholders. Thus as per the signaling theory, unexpected dividend changes convey valuable information to market participants that relate to managers’ expectations regarding the prospects of the firm. The last dividend theory is the agency theory of dividends which, like the signaling theory, proposes that dividends are value enhancing.Nevertheless, while the signaling theory is based on the assumption that managers always act in the interests of existing shareholders, the agency theory relaxes this assumption and allows for agency conflicts. The agency theory of dividends is different from the signaling theory in another crucial respect. Particularly, as per the signaling theory dividends have no value in themselves, but their value is derived from the information they contain about the firm’s fundamentals. In contrast, the agency theory of dividends states that the payment of dividends is in itself valuable for the reason that it controls agency costs in two ways. First, the payment of dividends reduces the free cash flows under managers’ discretion. Second, the payment of dividends forces the firm to the capital market inducing external monitoring of the firm and its management, which is valuable due the free rider problem of collective monitoring.The discussion on the theoretical themes that have developed to explain the dividend puzzle was followed by a review of some of the relevant empirical methodologies and evidence. Event studies around ex-dividend days are typically used to investigate tax clientele effects. Similarly, the market reaction to the dividend signal is often investigated by event studies around dividend announcement dates, while other methodologies comprise comparison and regression analyses. Nevertheless, one unique approach to understanding dividend policy, whose findings have been central to the dividend debate is the Lintner’s (1956) survey of US managers. The main conclusions from this study are that managers concern themselves primarily with the stability of dividends, believing that the market reacts favorably to dividend increases and unfavorably to decreases. Furthermore, the level of earnings is the most significant determinant of the dividend level, and the dividend decision is taken before other decisions such as investment decisions, which are then adjusted.Lintner’s (1956) study is consistent with the signaling rationale for dividends. Nevertheless, evidence from empirical studies of the signaling hypothesis is mixed. In general it appears that the market reacts strongly to unexpected dividend changes, and that the reaction is typically in the same direction as the dividend change. Nevertheless, evidence is weaker on actual changes in performance that follow the dividend change announcement. Similarly no consensus has been achieved on the effects of taxes, particularly on whether taxes have permanent or only temporary impact on prices, although the general conclusion is that taxes enter the dividend decision. The transaction/agency costs structure faced by the firm appears significant in determining its dividend policy. Nevertheless, there is also evidence of substitutability amongst dividends and other agency cost control devices such as leverage, managerial ownership, incentive schemes, investment opportunities and others. Thus the general conclusion is that after four decades of debate, the jury is still out on the dividend puzzle.For this matter, further research is required to sustain the spotlight on the dividend puzzle. In particular, there are four promising research ideas (PRIs), which derive directly from the theoretical and empirical

literature surveyed in this chapter. The first PRI relates to the role of agency theory in explaining dividend policy for firms operating in emerging markets, where imperfections are the norm rather than the exception. In these markets agency conflicts and information problems can be expected to loom strong and the finance gap to be particularly wide. Thus models that incorporate these factors, such as the cost minimization model reviewed, should describe well the target payout process of firms in emerging markets. The second PRI is inspired by developments in areas of corporate governance, and therefore seeks to attain a synergy between corporate governance and the dividend policy puzzle. It may be that the failure to unravel the dividend puzzle has been amplified by failure to recognize interaction of the dividend policy practice with other business features. One idea is to bring together the literature on business groups and firm ownership in order to understand dividend policy, especially in the case of emerging markets.The third PRI recognizes that different theories may have the same practical implications thus making difficult the task of distinguishing amongst them. For instance, both agency theory and signaling theories predict a positive reaction to dividend increases and negative reaction to decreases. One possible method of distinguishing between these theories is by exploiting institutional differences across countries, as in Dewenter, and Warther (1998) with respect to the US and Japan. Furthermore, cross country comparisons can also assist in establishing fine distinctions between various under-themes within major theories. For instance, La Porta, Lopez-de-Silanes, Shleifer and Vishny (2000) use cross-country comparisons to distinguish between two competing agency models of dividends, namely the outcome and substitute models. Finally, the forth PRI is to use a system of equations instead of the single equation model of dividends. This idea is discussed in Prasad Green and Murinde (2001), and acknowledges the possibility that policy choices may be simultaneously determined. An example is Jensen, Solberg and Zorn (1992), where insider ownership, debt and dividend policies are modeled as a simultaneous equations system.REVIEW QUESTIONS

1. Explain the Transaction Cost Theory.2. What do you mean by “The Bird in Hand Argument”?3. What is Agency Theory of Dividend?4. Explain permanent earnings, cash flow and risk information hypothesis of dividend.5. Explain the Rozeff’s Cost Minimization Model.

CASE STUDY

Determinants of the Dividend Policy in Emerging Stock Exchanges: The Case of GCC Countries

IntroductionDividend policy is one of the most intriguing topics in financial research. Even now, economists provide considerable attention and thought to solving the dividend puzzle, resulting in a large number of conflicting hypotheses, theories and explanations. Researchers have primarily focused on developed markets; nevertheless, additional insight into the dividend policy debate can be gained by an examination of developing countries, which is currently lacking in the literature. Dividend policy in emerging markets is often different in its nature, characteristics, and efficiency, from that of developed markets. The purpose of this paper is to identify the factors that influence the dividend policy of firms listed in GCC countries, while focusing on agency and transaction cost theory.

Literature ReviewDividend policy has been the subject of considerable debate since Miller and Modigliani (1961) illustrated that under certain assumptions, dividends were irrelevant and had no influence on a firm’s share price. Since then, financial researchers and practitioners have disagreed with Miller and Modigliani’s proposition and have argued that they based their proposition on perfect capital market assumptions, assumptions that do not exist in the real world. Those in conflict with Miller and Modigliani’s ideas introduced competing theories and hypotheses to provide empirical evidence to illustrate that when the capital market is imperfect, dividends do matter. For instance, the bird in the hand theory (predating Miller and Modigliani’s paper) explains that investors prefer dividends (certain) to retained earnings (less certain): therefore, firms should set a large dividend payout ratio to maximise firm share price (Gordon, 1956; Lintner, 1956; Fisher, 1961; Walter, 1963; Brigham and Gordon, 1968).In the early 1970s and 1980s, several studies introduced tax preference theory (Brennan, 1970; Elton and Gruber, 1970; Litzenberger and Ramaswamy, 1979; Litzenberger and Ramaswamy, 1982; Kalay, 1982; John and Williams, 1985; Poterba and Summers, 1984; Miller and Rock, 1985; Ambarish et al., 1987). This theory suggests that dividends are subject to a higher tax cut than capital gains. This theory further argues that dividends are taxed directly, while capital gains tax is not realised until a stock is sold. Therefore, for tax-related reasons, investors prefer the retention of a firm’s profit over the distribution of cash dividends. The advantage of capital gains treatment, nevertheless, may lead investors to favour a low dividend payout, as opposed to a high payout. In the early 1980s, signalling theory was analysed. It revealed that information asymmetry between managers and outside shareholders allows managers to use dividends as a tool to signal private information about a firm’s performance to outsiders (Aharony and Swary, 1980; Asquith and Mullins, 1986; Kalay and Loewenstein, 1985; Healy and Palepu, 1988).Another explanation for dividend policy is based on the transaction cost and residual theory. This theory indicates that firms incurring large transaction costs, will be required to reduce dividend payouts to avoid the costs of external financing (Mueller, 1967; Higgins, 1972; Crutchley and Hansen, 1989; Alli et al., 1993; Holder et al., 1998). A different explanation, which received little consideration prior to the 1980s, relates dividend policy to the effect of agency costs (La Porta et al., 2000). Agency costs, in this case, are costs incurred in monitoring company management to prevent inappropriate behaviour. Large dividend payouts reduce internal cash flows, forcing managers to seek external financing, and thereby, making them liable to capital suppliers, thus, reducing agency costs (Rozeff, 1982; Easterbrook, 1984; Lloyd, 1985; Crutchley and Hansen, 1989; Dempsey and Laber, 1992; Alli et al., 1993; Moh’d et al., 1995; Glen et al., 1995; Holder et al., 1998; Saxena, 1999). Dividend policy has been analysed for many decades, but no universally accepted explanation for companies’ observed dividend behaviour has been established. Brealey and Myers (2005) described dividend policy as one of the top ten most difficult unsolved problems in financial economics. This description is consistent with Black (1976) who stated that “The harder we look at the dividend picture, the more it seems like a puzzle, with pieces that don’t fit together”. What might be significant to mention, is that researchers have primarily focused on developed markets, while little attention has been paid to dividend policy in emerging markets. As a result, this field is not well established in the financial literature. Dividend policy in emerging markets is often very different in its nature, characteristics, and efficiency, from that of developed markets. In particular, the case of the GCC countries has some interesting characteristics that make the study appropriate in terms of policy recommendations for the GCC countries and other emerging countries. First, the GCC environment is unique in that taxes are not paid on dividends, or capital gains. This leads investors, in these particular countries, to favour a large dividend payout. Second, the stock exchanges in these countries are more volatile and entail a certain degree of information asymmetry, in addition to an expectation that high agency costs will be incurred. Third, governments own a significant proportion of shares in the GCC listed firms, especially large-sized firms. Therefore, government

participation might create a complex setting of agency theory, whereby government comprisement may duplicate the agency problem and, at the same time, served managing. As a result of the GCC’s characteristics, there is considerable interest in identifying dividend policy determinants for these companies, especially the validity of the agency explanation in the listed GCC firms. Thus, this paper continues the debate over dividends in the emerging market area, by presenting new evidence from GCC countries. Therefore, it may be that additional insight into the dividend policy debate can be gained for the case of developing countries.3. Factors Influencing Dividend Decision and Research HypothesesThe previous section reviewed the framework of dividend policy and discussed several studies that tested dividend policy in emerging markets. This section formulates seven hypotheses to further examine the factors, which may affect corporate dividend policy. This section also explains the appropriate proxy variables used to measure the factors affecting dividend payouts.Research HypothesesOwnership Structure: In a modern corporate environment where there is a large separation between ownership and management, conflicts of interest can arise between managers, inside owners (controlling shareholders), and outside shareholders, such as minority shareholders. Referring to this problem, Jensen and Meckling (1976) describe the firm as a nexus of contracting relationships among individuals. Nevertheless, when the manager makes a decision, it tends to be in favour of the agent, rather than of the firm. La Porta et al. (2000) illustrated that managers may take advantage of their authority to benefit themselves by diverting firm assets to themselves through theft, excessive salaries or sales of assets at favourable prices to themselves. Accordingly, the ownership structure in large firms may influence dividends and other financial policies (Desmetz, 1983; Desmetz and Lehn, 1985; Shleifer and Vishny, 1986; Morck et al., 1988; Schooley and Barney, 1994; Fluck, 1999; La Porta 2000; Gugler and Yurtoglu, 2003). Several studies have suggested that dividend payouts can play a useful role in reducing the conflict between inside and outside owners. When insider owners pay cash dividends, they return corporate earnings to investors and can no longer use these earnings to benefit themselves (La Porta et al., 2000). Nevertheless, the percentage of earnings that can be used as dividends depends upon the ownership structure of the firm.Glen et al. (1995), Gul (1999a), Naser et al. (2004) and Al-Malkawi (2007) specified that in emerging markets, government ownership is a major determinant of the dividend decision-making process. Gul (1999a) suggested a positive association between government ownership and dividends, arguing that firms with high government ownership find it less difficult to finance investment projects, and hence, can afford to distribute more dividends. Conversely, firms with lower (or no) government ownership face difficulties in raising money, and instead consequently rely on retained earnings for investments, thereby paying small dividends. Glen et al. (1995) argued that investors need to be protected in countries with poor legal systems. In addition, since governments are powerful investors, they should act as a safeguard for the minority shareholders by monitoring the insider shareholders and forcing them to disgorge cash. Naser et al. (2004) added that in an emerging market, where legal protection is limited, governments have a strong desire to build up firm reputations and avoid the exploitation of minority shareholders by paying them large dividends. They further asserted that the need for such a reputation has significant effects on young stock exchanges where there is no history of the good treatment of minority outside shareholders. In addition, this need is greater when there is high uncertainty about the future cash flow of firms.In a recent study, Al-Malkawi (2007) found that, among large shareholders, the government is one of the most influential shareholders in affecting the dividend policy of firms listed on the Amman Stock Exchange. He explained that the government acted on behalf of the citizens, who did not control the firm directly.

Therefore, in such firms, “a double principal-agent” conflict existed. This conflict may occur between citizens and government representatives, who might not act in the citizens’ best interests and between government representatives and other managers. The solution to this problem is a larger payment of dividends, which reduces the cash flow available to managers, thus, reducing the agency problems of the firm. This explanation concurs with the findings of Gugler (2003), who examined the dividend policies of Australian firms. Considering all previously discussed arguments, the following hypothesis can be formulated:

H1: The dividend payout is positively associated with government ownershipIn which the percentage of shares held by the government can be used as an indicator of the firm ownership structure.Free Cash FlowThe percentage of shares owned by different types of shareholders may not be the sole determinant of the dividend-agency relationship; the free cash flow may also be significant. Jensen (1986) defined free cash flow as the cash flow in excess of the funds required for all projects with a positive net present value (NPV). He demonstrated that as the free cash flow increases, it raises the agency conflict between the interests of managerial and outside shareholders, leading to a decrease in the performance of the company. While shareholders desire for their managers to maximise the value of their shares, the managers may have a different interest and prefer to derive benefits for themselves. Jensen's free cash flow hypothesis has been supported by subsequent studies by Jensen et al. (1992) and Smith and Watts (1992). La Porta et al. (2000) added that when a firm has a free cash flow, its managers will engage in wasteful practices, even when the protection for inventors improves. A number of studies have suggested that firms with a greater “free cash flow” need to pay more dividends to reduce the agency costs of the free cash flow (Jensen, 1986; Holder et al., 1998; La Porta et al., 2000; and Mollah et al., 2002). Based on the findings of the above studies, it can be speculated that there is a positive relationship between the free cash flow and the dividend payout ratio. Therefore, the second hypothesis becomes:

H2: The dividend payout is positively associated with free cash flowThe free cash flow to total asset ratio can be used as a proxy for the free cash flow variable. Eddy and Seifert (1988), Jensen et al. (1992), Redding (1997), and Fama and French (2000) indicated that large firms distribute a higher amount of their net profits as cash dividends, than do small firms. Several studies have tested the impact of firm size on the dividend-agency relationship. Lloyd et al. (1985) were among the first to modify Rozeff's model by adding “firm size” as an additional variable. They considered it an significant explanatory variable, as large companies are more likely to increase their dividend payouts to decrease agency costs. Their findings support Jensen and Meckling’s (1976) argument, that agency costs are associated with firm size. They were of the view that for large firms, widely spread ownership has a greater bargaining control, which, in turn, increases agency costs. Furthermore, Sawicki (2005) illustrated that dividend payouts can help to indirectly monitor the performance of managers in large firms. That is, in large firms, information asymmetry increases due to ownership dispersion, decreasing the shareholders’ ability to monitor the internal and external activities of the firm, resulting in the inefficient control by management. Paying large dividends can be a solution for such a problem for the reason that large dividends lead to an increase in the need for external financing, and the need for external financing leads to an increase in the monitoring of large firms, for the reason that of the existence of creditors.Other studies related the positive association between dividends and firm size to transaction costs. For instance, Holder et al. (1998) revealed that larger firms have better access to capital markets and find it easier to raise funds at lower costs, allowing them to pay higher dividends to shareholders. This demonstrates a positive association between dividend payouts and firm size. The positive relationship between dividend payout policy and firm size is also supported by a growing number of other studies

(Eddy and Seifert, 1988; Jensen et al., 1992; Redding, 1997; Holder et al., 1998; Fama and French, 2000; Manos, 2002; Mollah 2002; Travlos et al., 2002; Al-Malkawi, 2007). Therefore, the hypothesis in regard to firm size is formulated as:

H3: The dividend payout is positively associated with firm size Growth OpportunitiesA review of the literature revealed several explanations for the relationship between growth opportunities and dividend policy. One explanation was that a firm tended to use internal funding sources to finance investment projects if it had large growth opportunities and large investment projects. Such a firm chooses to cut, or pay fewer dividends, to reduce its dependence on costly external financing. On the other hand, firms with slow growth and fewer investment opportunities pay higher dividends to prevent managers from over-investing company cash. As such, a dividend here would play an incentive role, by removing resources from the firm and decreasing the agency costs of free cash flows (Jensen, 1986; Lang and Litzenberger, 1989; Al-Malkawi, 2007). Consequently, dividends were found to be higher in firms with slow growth opportunities, compared to firms with high-growth opportunities, as firms with high-growth opportunities have lower free cash flows (Rozeff, 1982; Lloyd et al., 1985; Jensen et al., 1992; Dempsey and Laber, 1992; Alli et al., 1993; Moh'd et al., 1995; Holder et al., 1998).A number of other studies compared investment opportunity ratios to distinguish growth from non-growth firms (Murrali and Welch, 1989; Titman and Wessels, 1988; Gavers and Gavers, 1993; Moh'd et al., 1995). These studies revealed that growth firms, as compared to non-growth firms, exhibited a lower debt to reduce their dependence on external financing, which is costly. This explanation is consistent with Myers (1984), who stated that investment policy can be substituted for dividend payouts, therefore, reducing the agency problem, for the reason that it reduces the free cash flow. La Porta et al. (2000) investigated countries with high legal protection and revealed that fast-growth firms paid lower dividends, as the shareholders were legally protected, allowing them to wait to receive their dividends when the investment opportunities were good. On the other hand, in countries with low legal protection for shareholders, firms kept the dividend payout high, to develop and maintain a strong reputation, even when they had better investment opportunities.Several studies found that the sales/revenues growth rate was commonly used as a proxy variable for growth opportunities (Rozeff, 1982; Lloyd et al., 1985; Jensen et al, 1992; Alli et al., 1993; Moh’d et al., 1995; Holder et al., 1998; Chen et al., 1999;, Saxsena, 1999; Manos, 2002; Travlos, 2002). To retain comparability, this study also used this proxy for growth opportunities and tested the hypothesis that:

H4: The dividend payout is positively/negatively associated with growth opportunities.Financial LeverageA growing number of studies have found that the level of financial leverage negatively affects dividend policy ( Jensen et al., 1992; Agrawal and Jayaraman, 1994; Crutchley and Hansen, 1989; Faccio et al., 2001; Gugler and Yurtoglu, 2003; Al-Malkawi, 2005). Their studies inferred that highly levered firms look forward to maintaining their internal cash flow to fulfil duties, instead of distributing available cash to shareholders and protect their creditors. Nevertheless, Mollah et al. (2001) examined an emerging market and found a direct relationship between financial leverage and debt-burden level that increases transaction costs. Thus, firms with high leverage ratios have high transaction costs, and are in a weak position to pay higher dividends to avoid the cost of external financing. To analyze the extent to which debt can affect dividend payouts, this study employed the financial leverage ratio, or ratio of liabilities (total short-term and long-term debt) to total shareholders’ equity. Based on the above arguments, the following hypothesis was formulated for further investigation:

H5: The dividend payout is negatively associated with financial leverage.Business Risk

Several studies have been used to measure the beta value, as a proxy for the systematic risk where beta measures the stock's volatility in relation to the market (Rozeff, 1982; Lloyd et al., 1985; Alli et al., 1993; Moh’d et al., 1995; Casey and Dickens, 2000).In addition, it has been argued that high-risk firms tend to have a higher volatility in their cash flows, than low-risk firms. Consequently, the external financing requirement of such firms will increase, driving them to reduce the dividend payout to avoid costly external financing (Higgins, 1972; McCabe, 1979; Rozeff, 1982; Chang and Rahee, 1990; Chen and Steiner, 1999). Jensen et al. (1992) contended that greater systematic risk increased the uncertainty of the direct relationship between current and expected future profits. Hence, firms avoid commitment to pay large dividends, as the uncertainty about earnings increases. Moh'd et al. (1995) also reported an inverse relationship between the dividend ratio and intrinsic business risk, proxied by beta. They indicated that firms with unstable earnings paid lower dividends, in an attempt to keep the dividend payout stable and to avoid the high cost of external financing. In contrast, Mollah (2002) found that firms listed on the Dhaka Stock Exchange paid a large dividend, even though the beta for their stock was high. He then argued that in an emerging stock exchange, the dividend might not be the most appropriate tool to convey correct information about transaction costs to the market. This study uses beta as a common proxy for firm business risk, which represents a firm’s operating and financial risk (Rozeff, 1982; Loyed et al., 1985; Jensen et al., 1992; Alli et al., 1993; Moh’d et al., 1995; Holder et al., 1998; Chen et al., 1999; Saxsena, 1999; Manos, 2002).Based on the previous discussion, the following hypothesis was formulated:

H6: The dividend payout is negatively associated with systematic risk.ProfitabilityThe financial literature documents that a firm’s profitability is a significant and positive explanatory variable of dividend policy (Jensen et al., 1992; Han et al., 1999; Fama and French, 2000). Nevertheless, there is a significant difference between dividend policies in developed and developing countries. This difference has been reported by Glen et al. (1995), showing that dividend payout rates in developing countries are approximately two-thirds of those in developed countries. Moreover, emerging market corporations do not follow a stable dividend policy; dividend payment for a given year is based on firm profitability for the same year. La Porta et al. (2000) compared countries that had strong legal protection for shareholders with those that had poor shareholder legal protection, and related that to countries with inferior quality shareholder legal protection. Their conclusion was that shareholders will take whatever cash dividend they can get from firm profits, where a dividend is perceived as unstable. Wang et al. (2002) compared the dividend policy of Chinese and UK listed companies, and found that the former tended to vote for a higher dividend payout ratio, than the latter. Moreover, UK companies had a clear dividend policy in which annual dividend increases and all companies paid a cash dividend. In contrast, Chinese companies had unstable dividend payments and their dividend ratios were heavily based on firm earnings for the same year, not on any other factor. The latter finding was consistent with that of Adaoğlu (2000), who stated that the main determinant in the amount of cash dividends in the Istanbul Stock Exchange was earnings for the same year. Any variability in the earnings of corporations was directly reflected in the cash dividend level. A similar result was reported by Pandey (2001) for Malaysian firms. Al-Malkawi (2007) identified the profitability ratio as the key determinant of the corporate dividend policy in Jordan.As a proxy, this study measured firm profitability by the return on equity (ROE) (Aivazian et al., 2003, ap Gwilym et al., 2004). The following hypothesis was formulated to test the ROE:

H7: The dividend payout is positively associated with a firm’s current profitability4. Theoretical Framework and Measures of VariablesTo investigate the seven hypotheses created in this study associated with the impact of agency and transaction costs on dividend payment ratios of GCC listed companies, this study undertook an empirical testing of a model with the following framework:

DIV = f (GOV, FCF, SIZE, GROW, LEV, BETA, PROF)Where: the dividend payout ratio (DIV) is the dependent variable that is defined as:

DIV = (cash dividends/net profits)*100The dividend payout ratio indicates the percentage of profits distributed by the company among shareholders out of the net profits, or what remains after subtracting all costs (e.g., depreciation, interest, and taxes) from a company’s revenues. Most of the previous studies that investigated the impact of agency theory and transaction cost theory employed dividend payout ratios as a determinant of dividend in lieu of dividend per share and dividend yield (Rozeff, 1982; Lloyd,1985; Jensen et al., 1992; Dempsey and Laber, 1992; Alli et al., 1993; Moh’d et al., 1995; Holder et al., 1998; Chen et al., 1999; Saxena, 1999; Mollah et al., 2002; Manos, 2002; Travlos, 2002). The dividend payout ratio is also used in this research, rather than dividend per share and dividend yield, for two reasons. Firstly, the dividend payout ratio takes into consideration both dividend payout and dividend retention. Such a consideration is essential, for the reason that the hypotheses to be examined in this study are concerned with the relationship between the dividend payout and the amount of cash retained in the business, as well as how this may reduce agency costs and encourage future investment. Secondly, dividend per share and dividend yield were considered unsuitable, for the reason that neither takes into account the dividend paid in relation to the income level. It may also be true that the dividend yield model is considered a measure of firm value and a return to shareholders, and therefore, it may not necessarily be related to agency theory.To investigate whether the dividend payout ratio is affected by ownership structure, the model uses the percentage of shares owned by the government (GOV), as has been used in several existing studies (Gul, 1999a; Gugler, 2003). Free cash flow (FCF) is a measure of how much cash a company has for ongoing activities and growth after paying its bills. FCF is calculated as:

FCF = (net profit – changes in fixed assets- changes in net working capital)/total assetsFirm size (SIZE) is measured as a natural logarithm of market capitalisation. This is due to the fact that large firms will pay large dividends to reduce agency costs (Ghosh and Woolridge, 1988; Eddy and Seifert, 1988; Redding, 1997). Growth rate (GROW) is measured as the growth rate of sales (Rozeff, 1982; Lloyd et al., 1985; Jensen et al., 1992; Alli et al., 1993; Moh’d et al., 1995; Holder et al., 1998; Chen et al., 1999; Sexsena, 1999; Manos, 2002; Travlos, 2002).Leverage ratio (LEV) is measured as the debt to equity ratio as shown below:

LEV = total debt/shareholders’ equityDebt to equity ratio has also been used as a proxy by a number of existing studies (Jensen et al., 1992; Mollah, 2001; and Al-Malkawi, 2005).Business risk is denoted by BETA, a mathematical measure of the sensitivity of the rates of return on a given stock, compared with the rates of return on the market as a whole. It is used as a proxy for business risk (Rozeff, 1982; Lloyd et al., 1985; Jensen et al., 1992; Alli et al., 1993; Moh’d et al., 1995; Holder et al., 1998; Chen et al., 1999; Sexsena, 1999; Manos, 2002). Profitability (PROF) is the ratio of net profits to the amount of money that shareholders have put into the company. ROE has been used in several studies as a proxy for firm profitability (Aivazian et al., 2003, ap Gwilym et al., 2004.) and is calculated as follows:

PROF = (Net profit/shareholder’s equity)*100This creates the assumption that the dividend ratio per year is based on firm earnings for the same year.5. DataIn order to test the seven hypotheses related to dividend policies of the firms listed on the GCC countries’ stock exchanges, the factors representing the characteristics of the firms need to be collected. As discussed previously, the dependent variable of the proposed dividend policy model is the annual dividend ratio paid by a firm, and the explanatory variables are percentages of shares of the firm held by the government, free cash flow, firm size (i.e., market capitalisation), firm growth rate, leverage ratio,

business risk and firm profitability. The primary idea was to test the dividend policies of the firms listed on the GCC stock exchanges. The intention was to assemble a large sample (cross-sectional and time-series data) to obtain a good result, collecting data of the above factors for both financial and non-financial firms, for as many years as possible. At the same time, it was essential that the time period in which the factors were observed be the same for all firms. Due to limited information on financial firms, and the problem of missing data, it was not possible to collect the required data related to financial firms for the same time period. For similar reasons, the required data for non-financial firms were only available for the five years from 1999 to 2003. Although 245 non-financial firms were listed on the stock exchanges of GCC countries, the required data were available for 191 nonfinancial firms from 1999 to 2003.Table 1 shows the total number of non-financial firms listed on each stock exchange of GCC countries. This table also reveals the number of firms for which the required data were available. It can be seen that the Muscat Stock Exchange (MSE) had the highest number of non-financial firms (i.e., 75 firms), while the Doha Stock Exchange (DSE) had the lowest number of non-financial firms (i.e., 10 firms) for which the required data were available. Table 1 also illustrates that data were available for only 53% of the firms listed on the DSE. 9 out of 19 firms did not publish the required data prior to 2003, and therefore, these firms were excluded from the sample.

Table 1: Non-financial firms within GCC countries’ stock exchangesMarket Name Total

number of Listed Firms

Total number of firms for which required data were available

% of available firms

Kuwait Stock Exchange (Kuwait) 59 37 63%Saudi Arabia Stock Exchange (Saudi Arabia) 62 57 92%Muscat Stock Exchange (Oman) 92 75 82%Doha Stock Exchange (Qatar) 19 10 53%Bahrain Stock Exchange (Bahrain) 13 12 92%Total 245 191 78%

Both the dividend payout ratio and the factors affecting the dividends for 191 nonfinancial firms from 1999 to 2003 were collected. The primary source of these data was the 2004 Gulf Investment Guide (GIG) (Zughaibi and Kabbani institution), used to obtain the majority of the data. In addition, the directories of the national stock exchanges for each state were obtained to provide data that were not available in the Gulf Investment. Nevertheless, it was difficult to obtain data on government ownership and business risk (BETA). Such data were obtained from a Saudi financial consulting firm, Zughaibi and Kabbani (www.zkfc.com). Government ownership data for Kuwait were gathered from the daily national newspapers from 1999 to 2003. Government ownership data for Qatar were obtained from an unpublished report supplied by the Doha Stock Exchange. The data on business risk for all firms, listed in the GCC stock exchanges, were also collected from Zughaibi and Kabbani.6. Random Effects Tobit Models

Most existing studies employed a linear regression model, such as a multiple regression, random effects and fixed effects linear model, to investigate dividend payout ratios. In such models, it is assumed that the values of all dependent and explanatory variables are treated as known for the entire sample. Nevertheless, there are some cases, in real-life, where the sample is limited by truncation or censoring. Censoring of a sample only occurs when the explanatory variables are observed for the entire sample, but there is limited information about the response variable for a portion of the sample (Long, 1997). In

other words, the response variable is not observed for the entire range of the sample. For instance, many of the non-financial firms in the GCC stock exchange do not pay dividends (the response variable of the dividend policy model) to their shareholders (Table 2). Therefore, the dividend payout ratios by these firms are not observed, even though the characteristics of the firms (the explanatory variables of the model) are observed for all firms. Truncation limits the sample more severely, than censoring by excluding the observations based on a threshold of the dependent variable. An example of a truncation sample, in our case, would be one that excluded from the sample, those firms that do not pay dividends. It is worth noting that censoring does not change the sample, but truncation does.

Table 2: Listed number of non-financial firms and the summary of their dividend payoutsCountry Total number of

firmsNumber of firms

never paid dividend

Number of firms who did not pay

dividend consistently

Number of firms always paid

dividend

Bahrain 12 0 1 11Saudi Arabia 57 18 20 19Kuwait 37 2 24 11Muscat 75 36 24 15Doha 10 1 2 7Total 191 57 71 63Percentage 100% 30% 37% 33%

The firms in our sample were divided into two categories: (1) firms about whom the information on the explanatory variables, such as firms’ characteristics (e.g., government ownership, and free cash flow), as well as the response variables, such as the dividend payout ratio, is available, (2) firms about whom only the information on the explanatory variables is available. Therefore, our sample is a censored sample. The appropriate model for such a censored sample is the Tobit model (Tobin, 1958).The structural equation of the standard Tobit model is:

(1)

Where: is the latent dependent variable observed for values greater than 0 and censored for values less than, or equal to, 0. xi is the vector of the explanatory variable observed for all cases, β is the vector of coefficients to be estimated, and εi is the error term which is assumed to be independently normally distributed, that is, εi ~ N(0,σ2). The censored variable, observed over the entire range, is defined by the following measurement equation:

(2)Estimation of the structural equation by OLS is conducted with the censored sample by taking y = 0

when , or the truncated sample (that is, the sample with only y > 0), gives inconsistent estimates, that is, it underestimates the intercept and overestimates the slope, or vice versa (Long 1997; Gujarati, 2003; Woolbridge, 2002; Hsiao, 2002). These studies have suggested the use of the Tobit model presented below:Substituting equation (1) in (2) results in:

(3)

Notice that the observed 0’s on the dependent variable can mean either a “true” 0 or censored data (that is, cannot be observed). At least some of the observations in a sample must be censored data,

otherwise yi would always equal and the true model would be a linear model, rather than a Tobit model. The use of OLS estimation in the presence of censoring has been found to result in inconsistent estimates. The suitable estimation method, therefore, is the Maximum Likelihood (ML) estimator, as such estimates are consistent and asymptotically normal (Greene, 2003, Long 1997). This being the case, the observations were divided into two groups: (1) uncensored observations in which ML behaves the

same way as the linear regression model, (2) censored observations where the specific value of is not known, but the probability of being censored is used. As per Long (1997), the log-likelihood function for both censored and uncensored observations is given by:

(4)Where: φ(.) and Φ(.) represent the probability density function (pdf) and cumulative density function (cdf) of the standard normal distribution, all other terms in this model have been defined previously. The standard Tobit model may not be appropriate for the modeling of censored panel data, due to the presence of unobserved heterogeneity. Therefore, the unobserved effects (fixed and random) need to be taken into consideration. This resulted in two types of unobserved effects in Tobit models: (1) fixed, and (2) random. The literature suggests that the estimation of a fixed effects Tobit model is complex. This is for the reason that, at the present time, there is no feasible estimator for a fixed effects Tobit model (STATA, 2000). Therefore, the fixed effects Tobit model is not considered in this study. The random effects Tobit model is presented as follows (Long, 1997):

(5)Where: αi is the unobserved firm-specific effect assumed to be uncorrelated with xit, and independently and identically distributed with a zero mean and constant variance, that is, αi ~ N(0, σ2

it). The ML estimator is used to estimate the models parameters.7. Results and DiscussionsTable 3 presents the descriptive statistics for the variables, related to firms’ characteristics, comprised in the models to examine the dividend policy of non-financial firms listed on the stock exchanges of GCC countries for 1999 to 2003. The mean of the dividend payout ratio of the 191 non-financial firms indicates that the firms distributed an average of 43% of their net profits as dividends. The standard deviation of the dividend payout ratio was 59.8, suggesting that the dividend payout ratio was highly dispersed. It is noticeable that the Q2 of the dividend payout ratio paid by the firms, where the government owned a proportion of the shares, was higher (45%) than that of all firms (7%). This is largely due to the fact that a large proportion of firms did not pay dividends, either consistently, or for some of the years. For instance, it was found that 30% of the firms (i.e., 57 firms) did not pay a dividend during the study period and 37% of the firms (i.e., 71 firms) paid a dividend for some years, but did not pay a dividend for all years. The third quartile, Q3, (i.e., 75th percentile) of the dividend payout ratio is 77.6, implying that 25% of firms paid dividend above 77.6%.Table 3: Descriptive statistics of the variables used in the study for the nonfinancial firms listed on the

GCC countries’ stock exchanges for the period of 1999 to 2003Variables Mean Standard

DeviationQuartiles

Q1 Q2 Q3ALL GS ALL GS ALL GS ALL GS ALL GS

Dividend ratio 42.858 48.526 59.811 59.019 0.00 0.00 7.00 45.00 77.60 79.20

(DIV)Government ownership

(GOV)

10.098 18.358 17.512 20.152 0.00 3.00 1.00 8.00 10.22 30.00

Free Cash Flow (FCF

0.003 0.019 0.261 0.247 –0.10 –0.09 0.02 0.03 0.12 0.14

Market Capitalisation

(MC)$000

629179 981619 3810788 5102016 9984 6669 53269 26975 210600 210956

Growth Rate (GROW)

0.428 0.515 2.722 3.027 –0.02 –0.01 0.05 0.05 0.20 0.19

Firm Leverage (LEV)

204.875 108.420 2714.892 171.528 17.30 15.10 49.90 45.20 129.00 133.00

Business Risk (BETA)

0.394 0.333 0.474 0.475 0.01 0.01 0.28 17.00 0.72 0.57

Firm Profitability

(PROF)

8.649 9.744 12.558 11.068 0.00 0.00 5.50 7.30 13.80 15.00

Key:ALL = Data for all firmsGS = Data from the firms where the government owned a proportion of shares

Multicollinearity between explanatory variables may result in the wrong signs, or implausible magnitudes, in the estimated model coefficients, and the bias of the standard errors of the coefficients. To avoid this problem, the Variance Inflation Factor (VIF) test was used. The results of this test are presented in Table 4. The mean VIF was 1.06, which is much lower than the threshold of 10. The VIF for individual variables was also very low. This indicates that the explanatory variables comprised in the model were not substantially correlated with each other.

Table 4: Variance Inflation Factor (VIF) for the explanatory variablesVariables VIF Tolerance

Firm Size (MC) 1.09 0.9186Government ownership (GOV) 1.09 0.9196Firm profitability (PROF) 1.08 0.9271Free cash flow (FCF) 1.08 0.9302Business Risk (BETA) 1.05 0.9519Growth Rate (GROW) 1.01 0.9949Firm Leverage (LEV) 1.00 0.9976Mean VIF 1.06

To further test whether the explanatory variables were correlated, a pair-wise correlation matrix among the explanatory variables was estimated. The results are illustrated in Table 5, where it can be seen that the correlation coefficients were low (all < 0.300), suggesting that there was no multicollinearity problem among these variables.

Table 5: Correlation coefficients among the explanatory variablesVariables GOV FCF MC GROW LEV BETA PROF

Government ownership

(GOV)

1.0000

Free cash flow (FCF)

0.1006 1.0000

Firm size 0.2516 0.0229 1.0000

(MC)Growth rate

(GROW)–0.0112 –0.0346 –0.0204 1.0000

Firm Leverage

(LEV)

–0.0308 –0.0355 –0.0071 0.0084 1.0000

Business risk (BETA)

0.0785 0.1135 0.1511 –0.0616 –0.0091 1.0000

Firm profitability

(PROF)

0.1089 0.2367 0.0202 0.0046 –0.0297 0.1239 1.0000

Since the dataset is a time-series cross-sectional dataset, the random effects (RE) Tobit model was used, instead of the standard Tobit model. The estimation results for the random effects Tobit model are presented in Table 6. The model parameters were estimated using the maximum likelihood estimation (MLE) method. As can be seen from Table 6, the statistically significant variables at the 95% confidence level are government ownership, firm size, firm leverage and firm profitability. The insignificant variables are free cash flow, growth rate and business risk. Since the variables of free cash flow, growth rate, and business risk were not significant, H3, H4, and H6 could not be supported by the data from the 191 nonfinancial firms considered in this study. The significant variables are discussed in more detail below.

Table 6: Estimation results for the random effects Tobit modelModel A Model B

Explanatory Variables

Coeff. t-stat Coeff. t-stat

Government ownership (GOV)

0.9071 4.96 0.7853 4.00

Free cash Flow (FCF)

–1.1757 –0.09 1.3283 0.11

Ln(Firm size (MC) in US$)

10.5845 4.96 11.3535 3.79

Growth Rate (GROW)

–0.8066 –0.65 –0.7939 –0.62

Firm Leverage (LEV)

–0.0705 –2.7 –0.0568 –2.15

Business risk (BETA)

–9.7570 –1.21 –8.6363 –1.03

Firm profitability (ROE)

1.0668 4.16 0.9813 3.68

Country-specific dummiesSaudi Arabia (Reference)

– –

Bahrain 45.1264 2.92Kuwait 28.6671 2.47Oman –1.4781 –0.10Qatar 15.6187 0.97Constant –109.8740 –4.71 –130.5485 –3.42Descriptive

StatisticsWald statistic 110.7800 131.6200P-value>Wald statistic

0.0000 0.0000

Observations 929 929Left-censored observations

460 460

Uncensored observations

469 469

Log-likelihood function

–3073.3784 –3063.6868

Akaike Information criterion (AIC)

6162.7568 6143.3736

As can be seen, two types of models were estimated: (1) the RE Tobit model, estimated without country-specific dummies (Model A), and (2) the RE Tobit model, estimated with country-specific dummies (Model B). The dummy variable for Saudi Arabia was taken as a reference, when the other four dummy variables (Bahrain, Kuwait, Oman, and Qatar) were estimated. Even though both models provided similar results, in terms of their significant variables, the value of the log likelihood function was higher in Model B, suggesting that it was superior to Model A. Nevertheless, the number of parameters were different in these two models. Therefore, the Akaike Information Criterion (AIC) was used to control for parameters, while comparing the goodness-of-fits for these models. The smaller the value of the AIC, the better the result. AIC for Model A was 6162.76 and for Model B was 6143.37, suggesting that Model B is superior to Model A. It was also noticed that the values of the coefficients in Model B were smaller than those of Model A, for some variables. This is not surprising, as the country specific dummies in Model B will tend to offset some of the effects of the explanatory variables.The RE Tobit model, with country-specific dummies, were also estimated after removing the outliers found in the sample data associated with the GCC stock exchanges. Even though the sets of significant and insignificant variables of this model were the same as the model with the outliers comprised, the values of the parameters were different (Table 7). Moreover, the log likelihood value of the RE Tobit model, after removing the outliers, was higher than that of the model before removing the outliers. Nevertheless, the number of observations between these models was different.The Bayesian Information Criterion (BIC) was used to control for the number of observations, while comparing the goodness-of-fit between these models. The smaller the value of the BIC, the better the goodness-of-fit. The BIC for the RE Tobit, after removing the outliers, was 6048 and for the RE Tobit model, before removing the outliers, was 6202.56. This result suggests that excluding the outliers offers an improvement. Therefore, the rest of the interpretation of the results will be based on the RE Tobit model, with the outliers excluded.

Table 7: Estimation results for random effects Tobit model: The results prior to, and after, removing the outliers

RE Tobit Model(before removing outliers)

RE Tobit model(after removing outliers)

Explanatory Variables

Coeff. t-stat Coeff. t-stat

Government ownership (GOV)

0.7853 4.00 0.7150 4.05

Free cash flow 1.3283 0.11 4.0360 0.36

(FCF)Ln(Firm size (MC) in US$)

11.3535 3.79 9.0409 3.29

Growth rate (GROW)

–0.7939 –0.62 –2.4158 –1.06

Firm leverage (LEV)

–0.0568 –2.15 –0.0495 –2.09

Business risk (BETA)

–8.6363 –1.03 –6.1822 –0.81

Firm profitability (ROE)

0.9813 3.68 1.1923 4.17

Constant –130.5485 –3.42 –102.2439 –2.93Descriptive statisticsWald statistic 131.6200 135.4500P-value>Wald statistic

0.0000 0.0000

Observations 929 920Left-censored observations

460 455

Uncensored observations

469 456

Log-likelihood function

–3063.6868 –2986.5688

Bayesian Information Criterion (BIC)

6202.5600 6048.0000

As discussed in the methodology section, the coefficients of the RE Tobit model represent the underlying propensity to pay a dividend, that is, the impact of a change in an explanatory variable on the unconditional expectation of the unobserved or latent variable, y* . Figure 1 illustrates the comparison between the model predicted dividend ratio value at the expected E(y*|x) and the observed dividend ratio.

Figure 1: Observed and predicted values of the dividend ratioIt can be seen from Figure 1, that the random effects Tobit model, with country dummies, over-predicts the dividend ratio. One measure of accuracy is the model prediction error (MPE). This is given by:

(6)All symbols are as previously defined. The results reveal that the MPE is -1.36%. In addition, a paired sample t-test was conducted to determine whether the means of the observed value and the model predicted value were the same. The results are presented in Table 8. In Table 8, it can be seen that the p-value associated with the test was 0.29, implying that the means were not significantly different.

Observation Mean Standard errorObserved value 920 41.89467 1.810964Predicted value 920 43.95026 0.7232347t-statistic –1.0541p-value 0.292Decision Means are not different

Nevertheless, the main interest is to estimate both the marginal effects (ME) and the elasticities of both censored (y) and uncensored (y > 0) variables, with respect to the continuous independent variables comprised in Model B. The results are illustrated in Table 9 for ME and Table 10 for the elasticities. The ME and elasticities were evaluated at the means of the variables. It is noticeable that the marginal effects of market capitalisation (MC) were relatively low. This is for the reason that the units of the MC were in thousands of US dollars (i.e., US $1000).

Table 9: Marginal effects (ME) on both censored and uncensored variables (ME) with respect to the continuous explanatory variables

Explanatory Variables ME of E(y|x) ME for E(y|y>0,x)Government ownership (GOV)

0.3934 0.2801

Free cash flow (FCF) 2.2206 1.5813Firm size (MC) in 000 US$

8E-06 5.6E–06

Growth rate (GROW) –1.3292 –0.9465Firm Leverage (LEV) –0.0272 –0.0194Business risk (BETA) –3.4015 –2.4222Firm profitability (PROF) 0.6560 0.4671

The sign of the marginal effect of a variable was the same as the sign of corresponding coefficient from the RE Tobit model (Table 7). As can be seen, the marginal effects of the independent variables on the censored (y) and uncensored (y>0) variables were lower than for the marginal effects of the variables on y*. One explanation for this is the relative expected values of latent, censored, and uncensored variables. These values were found to be E(y*|x) = 13.79, E(y|x) = 43.95, and E(y|y>0,x) = 74.37.If one unit in an explanatory variable was changes, then the amount of the dividend ratio paid by the firms that always paid a dividend (i.e., y>0) would be less affected than the amount of the dividend ratio by all firms comprised in the sample. This is also true for the case of the elasticities shown in table 10. For instance, if all else were equal, a 10% increase in government ownership would lead to an increase

of 1% in the dividend payout ratio for all firms comprised in the sample, and only 0.4% for the firms who always paid the dividend.

Table 10: Elasticities of both censored (y) and uncensored (y > 0) variables, with respect to the continuous explanatory variables

Explanatory Variables ElasticitiesCensored Variable

Uncensored Variable

Government ownership (GOv)

0.0937 0.0391

Free cash flow (FCF) 0.0002 0.0001Firm size (MC) in 000 US$ 0.1171 0.0488Growth rate (GROW) –0.0094 –0.0039Firm leverage (LEV) –0.0682 –0.0284Business risk (BETA) –0.0313 –0.0130Firm profitability (PROF) 0.1314 0.0548

The statistically significant variables at the 95% confidence level were government ownership (H1), firm size (H3), firm leverage (H5), and firm profitability (H6). The insignificant variables were free cash flow, growth rate, and business risk. As a result, the hypotheses H2, H4 and H6 could not be supported by the data from the 191 non-financial firms considered in this study. The interpretation of the significant variable is given below.Government ownership appears to be a statistically significant determinant of dividend policy in the companies listed on the stock exchanges of GCC countries. This result supports Hypothesis (H1), which suggests that government ownership and the dividend ratio should have a positive relationship. The slope coefficient of this variable is 1.08, suggesting that a 1 unit increase in government ownership would have an increase of 1.08 units in the dividend ratio (ceteris paribus). Furthermore, the elasticity of the dividend payout ratio, with respect to government ownership, is found to be 0.25, suggesting that a 10% increase in government ownership would increase the dividend ratio by 2.5%. One explanation for the positive association between the dividend payout ratio and government ownership is that firms in which the governments own a percentage of their shares are able to pay higher dividends, for the reason that government ownership itself can attract external funds more easily. Consequently, they have less difficulty raising external funds to finance investments. In contrast, firms with low, or no, government ownership are more likely to experience difficulty raising funds and are, therefore, likely to depend on retained earnings for investment purposes, thus reducing the dividend payout (Gul, 1999a).Another possible reason for this positive relationship is that in GCC countries, where the legal protection for outside shareholders is poor, investors need to be protected. For the reason that the government, which may be seen as acting on behalf of minority shareholders, is a powerful investor, the controlling shareholders may be forced to pay a large dividend to avoid exploiting minority shareholders, and thereby, reducing agency conflict (Glen et al., 1995; Naser, 2004). An alternative hypothesis suggests that government involvement may exacerbate the agency problem. It may also promote a positive association between their ownership and dividend payout. In this case, agency problems may occur between citizens (who are not directly in control) and government representatives, since they might not act in the best interests of citizens. In addition, this may be true between the state-owner and other managers, for the reason that managers often look to benefit themselves in the expense of outside shareholders. Therefore, governments may solve this problem by encouraging the company to pay large dividends, which would reduce free cash flow in the hands of managers, and, at the same time, be in line with the preference of outside shareholders (Al-Malkawi, 2005).

Furthermore, the governments of GCC countries are looking to diversify their economic resources, for the reason that of ongoing deficits in state budgets and the negative impact on the economies of GCC countries of fluctuations in the price of, or decreased demand for, crude oil. One way to diversify their economic resources, and reduce the dependency on oil revenue and the government sector, would be to develop and encourage investment in the private sector. Therefore, governments may force firms to pay large dividends, so that these large dividends can enhance the reputation of the private sector by suggesting that the exploitation of minority shareholders is avoided. This good reputation may then attract small or minority shareholders to invest in such companies.In summary, government ownership was found to have a significant effect in promoting dividend payouts. There could be several reasons for this association: (i) government ownership itself attracts external funds more easily, (ii) a government shareholder, in countries where the legal protection is weak, becomes a powerful investor able to force the firm to disgorge cash, to avoid exploiting minority shareholders, (iii) to reduce the doubled agency conflict, and (iv) to attract investment in the private sector.Firm size was also found to be a statistically significant determinant of dividend policy. This result supports the Hypothesis (H4) that predicts that firm size and dividend ratio should have a positive association. The slope coefficient of this variable was 2.02E-05. It is noticeable that the value of this coefficient is relatively low. This is for the reason that the units of the firm size variable is large, being in US $1000. Nevertheless, this result suggests that the dividend ratio increases with firm size. In addition, the elasticity of the dividend payout ratio, with respect to firm size, is found to be approximately 0.3, suggesting that a 10% increase in firm size, if all else were equal, would lead to an increase of about 3% in the dividend ratio. This result is in line with previous studies, namely, that larger firms are capable of paying larger dividends (Eddy and Seifert, 1988; Jensen et al., 1992; Redding, 1997; Holder et al., 1998; Fama and French, 2000; Manos 2002; Mollah 2002; Travlos et al., 2002; Al-Malkawi 2005). As mentioned previously, larger firms pay a higher cash dividend for several reasons. First, large firms face high agency costs as a result of ownership dispersion, increased complexity, and the inability of shareholders to monitor firm activity closely. Hence, such firms pay a larger dividend to reduce agency costs (Jensen and Meckling, 1976; Lloyd et al., 1985). Second, as a result of the weak control in monitoring management in large firms, a large dividend payout increases the need for external financing, which, in turn, leads to the increased monitoring of large firms by creditors. This may be a quality that is attractive to the shareholders (Sawicki, 2005). Another explanation for this positive association might be related to large firms’ easier access to capital markets, and their ability to raise funds with lower issuance costs for external financing. Consequently, large firms are better able than small firms to distribute higher dividends to shareholders (Holder et al., 1998).The firm profitability ratio appeared to be a very strong and statistically significant determinant of the dividend payout ratio. This result supports Hypothesis H9, which predicted that firm profitability and the dividend ratio should have a positive association. The slope coefficient of this variable was 2.89, suggesting that a 1 unit increase in firm profitability would increase 2.89 units in dividend payout ratio (ceteris paribus). In addition, the elasticity of the dividend payout ratio, with respect to firm profitability, was found to be 0.58, suggesting that, if all else were equal, a 10% increase in firm profitability would lead to an increase of about 5.8% in the dividend payout ratio. This is consistent with the observation that firms normally pay a higher dividend ratio when there is a rise in firm profitability.The observed positive association between dividend payout and current firm profitability is in line with the results of Jensen et al. (1992), Han et al. (1999) and Fama and French (2000). The appearance of profitability as an significant factor influencing the dividend ratio is supported by Adaoğlu (2000), Pandy (2001), Wang et al. (2002), and Al-Malkawi (2005), who indicated that the dividend decision of firms listed on emerging stock exchanges was based on their realised earnings for the same year, which might illustrate that, for these firms, the dividend smoothness/stability was less significant. This finding might

be related to the fact that in GCC countries, as in other developing countries, there is inferior shareholder legal protection; consequently, shareholders will take whatever cash dividend they can get from firm profits (La Porta et al., 2000).The Leverage ratio was found to be strongly statistically significant and negatively associated with the dividend payout ratio. This means that if the leverage ratio of a firm increased, the dividend payout ratio paid by the firm decreased. This is consistent with Hypothesis H6. The marginal effects of this variable on y and y > 0 were found to be -0.03 and -0.02, respectively, suggesting that a unit increase in the leverage ratio would lead to a decrease of 0.03 units in the dividend payout ratio for all firms, and 0.02 units for the firms who always paid dividends if all other factors remained constant. The corresponding elasticities of the dividend payout ratio, with respect to the leverage ratio, were -0.07 and -0.03, respectively, implying that a 10% increase in the leverage ratio would lead to a decrease of about 0.7% in the dividend payout ratio for all firms and 0.3% for the firms who always pay dividends, if all other factors were to remain constant. The reason for this negative association is that highly levered firms carry a large burden of transaction costs from external financing. In this case, firms need to maintain their internal source of funds to meet their duties, instead of distributing the available cash to shareholders as dividends (Crutchley Hansen 1989; Mollah, 2001; Faccio et al., 2001; Aivazian et al., 2004; Naser et al., 2004; Al-Malkawi, 2005). Furthermore, Jensen et al. (1992) and Agrawal and Jayaraman (1994) indicated that for the reason that levered firms had a greater commitment to their creditors, the discretionary funds available to managers would be reduced. This suggests that agency costs will also be reduced. They also conclude that debt can be a substitution for a dividend.Statistically insignificant variablesThe appearance of government ownership, firm size, and firm profitability as the significant explanatory variables, support the idea that the main aim of non-financial firms listed on the GCC countries is to reduce agency conflict and maintain firm reputation. Nevertheless, a number of variables appeared to be statistically insignificant: free cash flow, growth rate, and business risk. What might be notable here is that free cash flow was the only agency theory explanatory variable found to have no influence on dividend policy, and therefore, Hypothesis H4 (the positive association between the amount of dividend payment and free cash flow) cannot be supported. This might be for the reason that the variable, government ownership, actually forced firms with high free cash flow to pay dividends. La Porta et al. (2000) supported this view and proposed an outcome model in which firms in countries with high legal protection paid higher dividends than firms in countries that had poor legal protection.The common transaction cost variables, growth rate, leverage ratio and business risk, also appeared as insignificant variables. This suggests that transaction costs do not have a direct influence on the dividend payout policy. In other words, the firms listed on the GCC countries’ stock exchanges took into account agency conflict and firm reputation, more than transaction costs, when they were making the decision to pay dividends.

8. ConclusionsThe main purpose of this paper was to determine the dividend policies of the nonfinancial firms listed on the GCC countries (i.e., Kuwait, Saudi Arabia, Muscat, Doha, Bahrain) stock exchanges for the period of 1999-2003, and to explain their dividend payment behaviour. Since a significant number of listed firms chose not to distribute cash dividends, in some or all of the years within the study period, the random effects Tobit model was an appropriate model for testing dividend policy. A series of random effects Tobit models were estimated and seven research hypotheses were tested. The statistically significant variables comprised government ownership, firm size, leverage ratio and firm profitability. The results indicated that the firms in which the government owned a proportion of the shares, paid higher dividends compared to the firms owned completely by the private sector. Furthermore, the results illustrated that the firms chose to pay more dividends when firm size and profitability were high.

The model also revealed that the leverage ratio was an additional variable that affected the dividend payout ratio of a firm. One way to extend this study is to investigate the dividend payout ratios of individual states and compare the results with those of GCC countries. Another is to extend this analysis by disaggregating the firms into sectors, such as the service and industry sector. Such an analysis will assist in identifying the sectors whose firms hand out the greatest dividends, and those that offer the least.Source: Global Economy & Finance Journal, Vol. 2 No. 2 September 2009. Pp. 38-63Questions

1. How will you evaluate the growth of non-financial firms listed on the GCC countries stock exchange?

2. Explain how the dividend payout ratio affects the dividend policies of the firms.

Unit 6

APPLIED ISSUES IN CORPORATE FINANCE

INTRODUCTION

With the spread of value-based management, corporate finance has also been gaining significance. Corporate finance is the business function carrying on with financial operations. Financial operations may comprise obtaining debt or equity financing for new business opportunities or acquisitions, and finding investment opportunities for earning interest and increasing the cash flows of the company. Companies may use a financial manager or business analyst to help them decide which opportunities represent the best return on the company's investment. While the majority of corporate finance managers spend copious amounts of time finding effective opportunities for the business, issues may arise that limit the company's ability to accurately assess business opportunities.

Lack of Opportunities

Companies may find out that not sufficient good corporate finance opportunities are accessible in the business environment. The majority of companies have a desired rate of return they want to earn when investing money into new business projects or opportunities. After examining each potential corporate finance opportunity, companies may decide that the present accessible opportunities do not meet the company's guidelines. Companies may choose to lower their investment guidelines to accept an opportunity if a decision completely has to be made. This option may be made by selecting the best option out a list of bad opportunities.

Unable to Finance New Operations

Companies over and over again make use of external financing when beginning new business operations or entering new economic markets. Rather than spending cash to pay for these new operations, companies over and over again use debt or equity financing to pay for these opportunities. The corporate finance function of a company is generally responsible for obtaining the best available bank loan or equity investment for new business operations. Companies may be not capable to secure the right mix of bank loans or equity financing depending on the terms or agreements that must be made to secure this financing. Failing to secure external financing may leave the company devoid of an ability to pay for the new business operation.

Poor Cash Flow from Decisions

An additional significant function of corporate finance and business is shaping the cash flow from existing or new business operations. Wrong valuation methods or other analysis derived from the corporate finance function may effect in lower cash flows than expected. Lower cash flows may necessitate the company to pay for business operations using capital reserves or external financing. Using capital reserves will lower the company's ability to pay for other short-term requirements with its on-hand cash amounts. Using too much external financing to pay for under-performing business operations may produce new cash outflows in the form of bank loan repayments or interest payments to investors.

PENSION FUND MANAGEMENT

Pension fund management necessitates the investment of assets to attain the long-term provision of funding for retirement. The management of pension funds can be distinguished from mainstream asset management in that pension fund management characteristically adopts a much more conservative approach to risk to meet the needs of all pension stakeholders. These stakeholders comprise all existing and future recipients of income from the pension plan as well as those parties contributing to the plan, together with employees and the business, government, or union offering employees the benefit of future retirement provision. As an incentive to decrease their reliance on the state for retirement provision, pension fund contributions usually receive beneficial tax incentives.

Given the huge significance pension funds hold over the long-term financial happiness of their beneficiaries, pension fund trustees play a chief role in the effective governance of a fund, with responsibilities together with ensuring that it sticks rigidly to its mandate and regularly analyzing the performance and abilities of the entity in point of fact making the investment decisions. The challenges facing of those overseeing pension funds are frequently evolving, driven by factors such as national legislation and developments in investment products. For instance, while various pension funds conventionally favored a mix of equities and bonds to help them achieve their long-term investments, debate continues about the role of guaranteed return products and hedge funds to augment fund performance.

Coming in the get up of social changes diminishing the role of the extended family, the ageing of the population in both OECD and Emerging Market (EME) economies is prompting an increased focus on provision of sufficient retirement incomes to the elderly, either by public or private means (World Bank 1994). Pay-as-you-go pension schemes – where wages are taxed to pay pensions directly – have proven workable in the past where the population grows rapidly and the elderly cohort is small. Nevertheless, these systems are facing increasing difficulties as ageing proceeds, for the reason that past benefit promises cannot be maintained without deplorable increases in contribution rates or vast and growing government debt. This situation is putting an increased emphasis on advance funding of pensions, where the transfer element between generations is minimized. Funded systems are themselves not without risks, notably those arising from capital market volatility, or even poor returns due to economic performance in the long term, and there are many others relating to aspects such as system design and institutional background. It is in the context of performance of domestic capital markets that the potential benefits of international investment come to the fore, as a way of minimizing exposure of retirement income of the performance of domestic markets. Tensions may nevertheless arise with domestic regulations that limit such international investment, whose ostensible aim is – mistakenly - often to “avoid risk” or, more plausibly, aid the development of domestic capital markets.

Investment considerations for institutional investorsGeneral portfolio considerations for institutional investorsFor any agent, the most fundamental objective of investment is to attain an optimal trade-off of risk and return by allocation of the portfolio to appropriately diversified combinations of assets (and in some cases liabilities, i.e. leveraging the portfolio by borrowing). As derived by Tobin and Markowitz, with a mean-variance reliant configuration of risk preferences, the precondition for such an optimal trade-off is ability to attain the frontier of efficient portfolios, where there is no likelihood of increasing return exclusive of increasing risk, or of reducing risk without reducing return. The exact tradeoff chosen will

depend on objectives, preferences and constraints on investors. In this context, there are common characteristics of all types of institutional investment. Liabilities are perhaps the the largest part crucial aspect, in the light of which, asset managers may identify the investors' objectives/preferences and constraints.A liability is a cash expenditure made at a specific time to meet the contractual terms of an obligation issued by an institutional investor. Such liabilities differ in assurance and timing, from known outlay and timing to uncertain outlay and uncertain timing. In this context, an institutional investor will seek to earn a satisfactory return on invested funds and to keep a reasonable surplus of assets over liabilities. Risk must be enough to make sure sufficient returns but not so great as to threaten solvency. The nature of liabilities also determines the institutions' liquidity needs. Hence, in terms of objectives, there is a need to assess where on the above-mentioned optimal risk return trade-off the investor wishes to be, in other words his or her risk tolerance in pursuit of return, which will depend on liability considerations. This may in turn influence of holdings of international assets.Uniformly, there are a multiplicity of constraints, all of which may have a marked effect on optimal portfolios, and thus on holdings of international assets. All of these may link to the nature of the liabilities, for example:

• liquidity based restraints relate to the right for investors to withdraw funds as a lump sum or the present needs for regular disbursement;

• the investment horizon connects to the planned liquidation date of the investment, and is often measured by the concept of effective maturity or duration;

• inflation sensitivity relates to the need to hold assets as inflation hedges;• tax considerations may change the nature of the trade-off;• accounting rules can generate different 'optimal' portfolios;• lastly there is the impact of regulations. Besides those linking directly to asset allocation, which

may directly limit international investment, there are sometimes liability restrictions, which may thus influence preferred asset allocations e.g. by enforcing indexation of repayments or minimum solvency levels.

After these conditions are taken into account, investment strategies are developed and implemented. A main decision is to choose the asset categories to be incorporated in the portfolio and whether it should comprise foreign assets. Following this, the investment process is often divided into more than a few mechanism, with asset allocation (or strategic asset allocation) referring to the long term decision on the disposition of the overall portfolio, while tactical asset allocation relates to short term adjustments to this basic choice between asset categories in the light of short term profit opportunities, so-called “market timing”. For the time being security selection relates to the choice of individual assets to be held within each asset class, which may be both strategic and tactical. As noted, the above considerations are based broadly on the mean-variance model, which assumes that the investor chooses an asset allocation based solely on average return and its volatility. Many considerations in respect of liabilities affecting risk preferences give rise to alternative paradigms of asset allocation, which may entail a different approach to investment (Borio et al 1997):Immunization is a special case of the mean-variance approach which entails that the investor tries to stabilize the value of the investment at the end of the holding period, i.e. to hold an exclusively riskless position; this is done characteristically in respect of interest rate risk by appropriately adjusting the duration of the assets held to that of the liabilities. Since liabilities are characteristically in domestic currency, it implies holding of domestic assets. It involves a regular rebalancing of the portfolio - as well as the existence of assets that have a similar duration to liabilities. Matching is a particular case of immunization where the assets precisely imitate the cash flows of the liabilities, including any related option characteristics.

Deficit risk and portfolio insurance approaches put a particular stress on avoiding descending moves, e.g. in the context of minimum solvency levels for pension funds. Therefore, unlike mean-variance they are not symmetric in respect to the weight put on ascending and descending asset price moves. Shortfall risk sees the investor as maximizing the return on the portfolio subject to a ceiling on the likelihood of incurring a loss (e.g. by shifting from equities to bonds as the minimum desired value is approached, or hedging with derivatives). In portfolio insurance the investor is considered to want to avoid any loss but to retain upside profit potential. This may be attained by replicating on a continuous basis the payoff of a call option on the portfolio by trading between the assets and cash (dynamic hedging), or by use of futures and options per se. By these means, the value of a portfolio may be prevented from falling below a given value.An additional issue is if the benchmark for investment is seen in nominal terms, as implicitly understood above, or real terms, reflecting liabilities. Asset management techniques which take into account the nature of liabilities are known as asset liability management techniques (ALM), of which immunization is a special case. They may be defined as an investment technique wherein long term balance between assets and liabilities is maintained by choice of a portfolio of assets with comparable return, risk and duration characteristics to liabilities (although characteristics of individual assets may differ from those of liabilities). This approach may affect inter alia the suitable degree of international diversification of the portfolio.Investment issues for pension fundsIn the context of the above debate, we now go on to assess investment issues for pension funds in more detail. Pension funds collect pool and invest funds contributed by sponsors and beneficiaries to provide for the future pension entitlements of beneficiaries (Davis (1995), Bodie and Davis (2000)). They hence supply means for individuals to collect saving over their working life so as to finance their consumption requirements in retirement. In terms of the framework above, they must shape their assets to the pertinent time horizon and varying degree of liquidity based constraints. Returns to members of pension plans backed by such funds may be purely dependent on the market (defined contribution funds) or may be overlaid by a guarantee of the rate of return by the sponsor (defined benefit funds). The latter have insurance features which are absent in the former (Bodie 1990). These comprise guarantees in respect of replacement ratios (pensions as a proportion of income at retirement) subject to the risk of bankruptcy of the sponsor, as well as possible for risk sharing between older and younger beneficiaries. Defined contribution plans have tended to grow in recent years, as employers have sought to reduce the risk of their obligations, while employees wish funds that are readily transferable between employers.For mutually defined advantage and defined contribution funds, the portfolio distribution and the corresponding return and risk on the assets seek to match or if at all possible exceed the growth of average labor earnings. This will maximize the replacement ratio (pension as a proportion of final earnings) obtainable by purchase of an annuity at retirement financed via an occupational or personal defined contribution fund, and reduce the cost to a company of providing a given pension in a defined benefit plan. This link of liabilities to labor earnings points to a crucial distinction with insurance companies, in that pension funds face the risk of increasing nominal liabilities (for example, due to wage increases), as well as the risk of holding assets, and hence need to trade instability with return. In the context of the framework above, liabilities have an uncertain outcome and timing. In effect, their liabilities are characteristically denominated in real terms and are not fixed in nominal terms. Therefore, they must also focus on real assets which offer some form of inflation protection. Note in this context that domestic equities are a matching asset when liabilities grow at the same pace as real wages, as is typical in an ongoing pension fund aiming for a certain replacement ratio at retirement, for the reason that the labor and capital shares of GDP are roughly constant, and equities constitute capital income. Domestic bonds are not a good equivalent for real-wage based liabilities although they do match annuities for pensions.

An extra factor which will control the portfolio distributions of an individual pension fund is maturity - the ratio of active to retired members. The period of liabilities (that is, the average time to discounted pension payment requirements) is much longer for an immature fund having few pensions in payment than for a mature fund where sizeable repayments are essential. A fund which is closing down (or “winding up”) will have even shorter duration liabilities. Following the ALM approach, Blake (1994) suggests that given the unreliable duration of liabilities it is rational for immature funds having "real" liabilities as defined above to invest mainly in equities (whose cash flows have a long duration), for mature funds to invest in a mix of equities and bonds, and funds which are winding-up chiefly in bonds (whose cash flows have a short duration). Elasticity in the duration of assets, which may need major shifts in portfolios, is therefore necessary over time.There are in addition tax considerations. As revealed by Black (1980), for both defined benefit and defined contribution funds, there is a fiscal incentive to maximize the tax benefit of pension funds by investing in assets with the highest probable spread between pre-tax and post-tax returns. In various countries this tax effect gives an incentive to hold bonds. There is also an incentive to overfund with defined benefit to maximize the tax benefits, as well as to provide a bigger contingency fund, which is generally counteracted by government-imposed limits on funding. The accounting and regulatory framework will influence the approach to investment also. In addition key differences arise in the liabilities and investment approach of defined contribution and defined benefit funds:Defined contribution pension fundsIn a definite contribution pension fund the sponsors are only accountable for making contributions to the plan. There is no guarantee concerning assets at retirement, which depend on growth in the assets of the plan. For that reason the financial risks to which the provider of a defined contribution plan (as opposed to beneficiaries) is exposed are minimal. In a number of cases, exclusively the sponsor and the investment managers it employs choose the portfolio distribution, and hence there is a risk of legal action by beneficiaries against poor investment. But more and more, employees are left also to decide the asset allocation (e.g. in the US 401(k) plans). The outstanding obligation on the sponsor is to maintain contributions.In relation to portfolio objectives, a defined contribution pension plan should in real seek to enhance return for a given risk, so as to achieve as high as possible a replacement ratio at retirement. This entails following closely the standard mean-variance portfolio optimization schema discussed above. As noted by Blake (1997), in order to choose the suitable point on the frontier of efficient portfolios, it is essential to determine the degree of risk tolerance of the scheme member; the high the acceptable risk, the higher the expected value at retirement. The fund will also need to shift to lower risk assets for older workers as they approach retirement, thus lessening duration as discussed above and lessening exposure to market volatility abruptly before retirement which might otherwise risk to sharply reduce pensions. They will imply marked portfolio shifts over time.Until the approach of retirement requires a shift to bonds, the superior returns on equity and foreign assets are likely to ensure an important share of the portfolio is accounted for by equities, depending on the degree of risk aversion. Where employers choose the asset mix, the degree of risk aversion is likely to be related to the fear of proceedings when the market value of a more aggressive asset mix declines, where employees choose the asset allocation it is more direct risk aversion.Defined benefit pension fundsContrasting defined contribution funds, defined benefit funds are subject to a wide variety of solvency risks touching the sponsor. Given there is generally a guarantee of a certain replacement rate at retirement, the fund is subject to risk from earnings growth. Liabilities will also be prejudiced by interest rates at which future payments are discounted, and therefore there are significant interest rate risks. Falling asset returns will affect asset/liability balance. There are also risks of changes in government regulation (such as those of indexation, portability, vesting and preservation) that can vastly and

unexpectedly change liabilities. The instance of the UK, where such changes have been marked, is discussed in Davis (2001).Defined benefit fund liabilities are, owing to the sponsor's guarantee, fundamentally a type of corporate debt (Bodie 1991). Appropriate investment strategies will rely on the nature of the liabilities, together with whether pensions in payment are indexed and the demographic structure of the workforce. Investment strategies will also be prejudiced by the minimum-funding rules forced by the authorities which decide the size of surplus assets, as well as accounting conventions affecting the way shortfalls are presented in annual reports. These imply a focus on shortfall risk as defined above. To assess the suitable investment strategies in the light of liabilities, a number of definitions are required. The wind-up definition of liabilities, the level at which the fund could meet all its current obligations if it were to be closed down entirely, is known as the accumulated benefit obligation (ABO). The projected benefit obligation (PBO) implies that the obligations to be funded include a forward-looking element. It is understood that rights will carry on to accrue, and will be labor earnings indexed up to retirement, as is normal in a final salary plan. The indexed benefit obligation (IBO) also assumes price-indexation of pensions in payment after retirement.

If the sponsor attempts to fund the accumulated benefit obligation, and the obligation is solely nominal, with a minimum-funding requirement in place, it will be suitable to immunize the liabilities with bonds of the same duration to hedge the interest rate risk of these liabilities. Unhedged equities will just imply that such funds incur preventable risk (Bodie (1995)), even though they may be useful to provide extra return on the surplus over and above the minimum funding level. Regulations and practice differ between countries as to which of these is aimed for. With a projected benefit obligation target, an investment policy based on diversification may be most appropriate, in the belief that risk reduction relies on a maximum diversification of the pension fund relative to the firm's operating investments (Ambachtsheer 1988), which could certainly include foreign assets. Furthermore, it is normal for defined benefit schemes which offer a certain link to salary at retirement for the liability to include an element of indexation. Then fund managers and actuaries characteristically suppose that it may be suitable to include an important proportion of real assets such as equities and property in the portfolio as well as bonds. By doing this, they unreservedly diversify between investment risk and liability risk (which are largely risks of inflation). As noted by Blake (1997), minimum funding levels and limits on overfunding offer tolerance limits to the variation of assets around the value of liabilities. If the assets are selected in such a way that their risk, return and duration characteristics match those of liabilities, there is a "liability immunizing portfolio". This protects the portfolio against risks of variation in interest rates, real earnings growth and inflation in the pension liabilities. Such a strategy, which determines the overall asset allocation between broad classes of instrument, may be assisted by an asset-liability modeling exercise (ALM) as detailed above. Severe minimum funding rules will overlay this with shortfall risk conditions also.

LEVERAGED BUYOUT AND GOING PRIVATE CONCERNS

In a leveraged buyout, a company is taken on by a specialized investment firm by means of a relatively small portion of equity and a comparatively outsized portion of outside debt financing. The leveraged buyout investment firms nowadays refer to themselves (and are usually referred to) as private firms. In a typical leveraged buyout transaction, the private equity firms buys mainstream control of an existing or mature firm. This arrangement is distinct from venture capital firms that characteristically invest in young or emerging companies, and characteristically do not obtain majority control. In this section we

spotlight particularly on private equity firms and the leverages buyouts in which they invest, and we will use the terms private equity and leveraged buyouts synonymously.

Leverages buyouts first came out as a significant phenomenon in the 1980s. As leveraged buyout activity enlarged in that decade, Jensen (1989) predicted that the leveraged buyout organizations would finally become the overriding corporate organizational form. He argued that the private equity firm itself combined concentrated ownership stakes in its portfolio companies, high-powered incentives for the private equity firm professionals, and a slant, efficient organization with negligible overhead costs. The private equity firm then applied performance-based managerial compensation, highly leveraged capital structures (often relying on the junk bond financing), and active governance to the companies in which it invested. As per the Jensen, these structures were superior to those of the typical public corporation with dispersed shareholders, low leverage, and weak corporate governance. A few years later, this forecast seemed untimely. The junk bond market crashed; a large number of high-profile leveraged buyouts resulted in default and bankruptcy; and leveraged buyouts of public companies (so called public-to-private transactions) virtually disappeared by the early 1990s.

But the leverages buyout market had not died—it was only in hiding. At the same time as leveraged buyouts of public companies were comparatively scarce throughout the 1990s and early 2000s, private equity firms continued to purchase private companies and divisions. In the mid-2000s, public-to-private transactions re-emerged when the United States (and the rest of the world) experienced a second leveraged buyout boom.

In 2006 and 2007, a record amount of capital was dedicated to private equity, both in nominal terms and as a fraction of the overall stock market. Private equity commitments and transactions rivaled, if not caught up with the activity of the first wave in the late 1980s that reached its peak with the buyout of RJR Nabisco in 1988. Nevertheless, in 2008, with the turmoil in the debt markets, private equity appears to have declined again.

When a listed company is took on and consequently delisted, the transaction is referred to as a public-to-private or going-private transaction. As most such transactions are financed by substantial borrowing, which is used to repurchase most of the exceptional equity, they are called leveraged buyouts (LBOs). An impression of the various types of LBO is given in Table 6.1. Four categories are usually recognized: management buyouts (MBOs), management buyins (MBIs), buyin management buyouts (BIMBOs), and institutional buyouts (IBOs).

Table 6.1: Summary definitions of types of public-to-private transaction

Term DefinitionLBO Leveraged buyout. Acquisition in which a nonstrategic bidder acquires a listed or non-listed

company utilizing funds containing a proportion of debt that is substantially above the industry average. If the acquired company is listed, it is subsequently delisted (in a going-private or public-to-private transaction)

MBO Management buyout. An LBO in which the target firm’s management bids for control of the firm, often supported by a third-party private equity investor

MBI Management buyin. An LBO in which an outside management team (often backed by a third-

party private equity investor) acquires a company and replaces the incumbent management team

BIMBO Buyin management buyout. An LBO in which the bidding team comprises members of the incumbent management team and externally hired managers, often alongside a third-party private equity investor

IBO Institutional buyin. An LBO in which an institutional investor or private equity house acquires a company. Incumbent management can be retained and may be rewarded with equity participation

Reverse LBO

A transaction in which a firm that was previously taken private reobtains public status through a secondary initial public offering (SIPO)

Reasons for the Listed Firms Going Private

Reduction of Stockholder-Related Agency Costs

The fundamental dilemma of principal-agent models is how to get the manager (the agent) to act in the best interests of the stockholders (the principals) when the agent has interests that deviate from those of the principals and an informational advantage.

• The incentive rearrangement supposition states that the gains in stockholder wealth that arise from going private are a result of providing more rewards for managers (through an increased ownership stake) that persuade them to act in line with the interests of investors. In addition, in the case of an institutional buyout, the concentration of ownership leads to enhanced monitoring of management.

• The free cash flow supposition suggests that the expected stock returns follow from debt-induced mechanisms that force managers to pay out free cash flows. Free cash flow is the cash flow in surplus of that necessitate to fund all projects that have positive net present value (NPV) when discounted at the suitable cost of capital. The high leverage does not allow managers to grow the firm further than its optimal size (so-called “empire building”) and at the expense of value creation.

Tax Benefits

The considerable increase in cash flow creates a main tax shield, which increases the pre-transaction (or pre-recapitalization) value. After the buyout, firms pay almost no tax for a period of at least five years. Accordingly, the (new) stockholders gain, but the government loses out.

Management Buyouts• Involves the management team’s purchase of

the bulk of the firm’s shares.• Create a win-win situation for shareholders

who receive a premium for their stock and management who retain control.

• To avoid lawsuits, the price paid must represent a higher premium to the current market price.

• Alternatively, the target may make itself less attractive by divesting assets the bidder wants.

• Cash proceeds of the sale could fund other defenses such as share buybacks.

Reduction of Transaction Costs

The cost of keeping up a stock exchange listing is extremely high. Even though the direct costs (fees paid to the stock exchange) are comparatively small, the indirect costs of being listed are substantial (for example the cost of complying with corporate governance/transparency regulations, which necessitates larger accounting/legal departments, the cost of investor relations managers, and the cost of management time in general, etc.). For a medium-sized listed company these indirect costs are predictable at US$750,000–1,500,000 annually. The going-private transaction eliminates several of the transaction costs.

Wealth Transfers from Bondholders to Stockholders

Gains in stockholder wealth that take place from going private outcome from the expropriation of value belonging to pre-transaction bondholders. There are three mechanisms during which a firm can transfer wealth from bondholders to stockholders: an unanticipated increase in the asset risk (the asset substitution risk); large increases in dividends; or an unexpected issue of debt of higher or equal seniority, or of shorter maturity. In a going-private transaction, the last mechanism in specificity can lead to considerable expropriation of bondholder wealth if protective covenants are not in place.

Defense against Takeover

Frightened of losing their jobs if a antagonistic suitor takes control, the management may decide to take the company private. Thus, an MBO is the ultimate defensive measure against a hostile stockholder or tender offer.

Undervaluation

As a firm is a portfolio of projects, there may be asymmetric information between the management and outsiders concerning the maximum value that can be realized with the assets in place. If management believes that the share price is undervalued in relation to the firm’s true potential, they may privatize the firm through an MBO. Alternatively, if an external party believes that it is

Leveraged Buyouts• Borrowed funds are used to pay for all

or most of the purchase price.• Can be of an entire company or

divisions of a company• The tangible assets of the company are

used as collateral for the loans• Investors in LBOs are referred to as

financial buyers for the reason that they are primarily focused on relatively short- to intermediate-term financial returns

Significant Differences between LBO and MBO• MBO leads to private companies while LBO leaves the

company publicly traded with shareholders receiving “stub” equity in addition to cash payout.

• Under the MBO, the company saves on public reporting costs, but its equity shares remain illiquid securities. LBO preserves equity liquidity but exploits no (or few) savings on reporting.

• Under the MBO, owners are insiders. In LBO, equity investors remain outsiders

• Under MBO, control of the firm changes. In LBO, control may not necessarily change since the stub equity remains in the hands of public shareholders.

• The MBO creates strong conflict of interests, requiring the board to actively represent shareholders in the buyout negotiations. In LBO, ordinary business judgment rules applies.

able to generate more value with the assets of the firm, the firm may be taken over by means of an IBO or MBI.

How High Are the Premiums Paid in LBOs

The premiums (relative to the pre-transaction share price) are in line with those on ordinary takeover transactions: Over last 25 years they have been in the range 35% to 45%. The cumulative average abnormal returns (CAARs) calculated over two months around the event date (the announcement of the going-public transaction) average around 25%, which is similar to those of ordinary takeover transactions. The abnormal returns are equal to the realized returns corrected for the market movement (the return of the market index) and the riskiness of the firm (the beta).

The Phases of the Buyout Process

Figure 6.1 shows the structure of the buyout process, the main research questions for each phase of the process, and their explanations. The first phase (Intent) consists of the identification of good LBO candidates; the second phase (Impact) comprises the actual LBO and an analysis of the expected returns; the third phase (Process) consists of the value creation while the firm is privately listed; and the fourth phase (Duration) concerns the duration of the private phase until the main shareholder exits through an IPO or trade sale. At every phase, eight main hypotheses or triggers can be examined: realignment of incentives, acquisition of control, reduction of free cash flow, wealth transfers from various stakeholders, tax benefits, a reduction of transaction costs, the importance of takeover defense mechanisms, and undervaluation of the target firm.

Figure 6.1: Phases and hypotheses of going private

What makes firms good buyout candidates? The “Intent” phase comprehends the distinctiveness of firms prior to their decision to go private and compares these features to those of firms which stay publicly listed. Out of the eight value drivers (discussed above) to go public in the United States, the decrease in taxation resulting from the tax shield is the major one. Therefore, firms with a high tax bill may regard as going private with a lot of leverage provided that a stable cash flow stream enables the firm to service the debt. In addition, firms with considerable free cash flow (excess cash) that could lead to value-destroying investments have also been exposed to be prime candidates for a public-to-private transaction. In the United States, decisions to go private in the 1980s were frequently motivated by anti-takeover defense strategies.

How does the market react to a buyout? The impact of an LBO offer can be anticipated by analyzing the instant stock price reaction or the premiums paid to pre-transaction stockholders. The CAARs and premiums replicate the expected value creation when the firm becomes privately held. They are larger at the announcement for firms in which pre-transaction managers hold small equity stakes, which entails that the buyout may encourage a realignment of incentives. In addition, the fact that the buyout will reduce large free cash flows triggers positive share price returns. Also, for firms paying a large amount of tax, the buyout announcement leads to positive abnormal returns. As a final point,

bondholder wealth transfers appear to exist but are playing only a very limited role in the wealth gains of pre-buyout stockholders.

Is value created during the private phase? Just the once a company is privatized, what post-buyout processes lead to more wealth creation? The post-transaction performance improvements are in line with those anticipated at the announcement of a going-private transaction. The causes of the performance and efficiency advances are mainly the organizational structure of the LBO (high leverage and strong (managerial) ownership concentration). In the private phase, a firm’s productivity increases due to a focused strategy and the avoidance of excess growth. Post-buyout performance improvements arise from an improved quality of the Research and Development function and intensified venturing activities. This revamped entrepreneurial spirit follows from reduced stockholder-related agency costs. Also typical of firms that go private is an important improvement in the management of working capital.

How long it is before a firm is relisted on the stock exchange? An investor may make a decision to end a company’s private status through an exit via a SIPO (secondary initial public offering, or reverse LBO). Particularly in the United States, some firms seem to use the organizational form of a privatization transaction as a impermanent shock therapy to enable them to restructure professionally, while others view the LBO as a sustainable and superior organizational form. Firms that do a reverse LBO have usually been private for three to six years. In Europe, major stockholders generally do not exit via a SIPO but perform a trade sale. The long life of private ownership and its determinants are studied in the literature on duration.

Safeway and KrogerOne of the key characteristics of going private through an LBO is the high-leverage structure that results. Nevertheless, the discipline of high leverage can also be induced by a leveraged recapitalization without privatization. Denis (1994) investigated the difference between the two approaches by contrasting two grocery store firms—Kroger, which undertook a recapitalization, and Safeway, which took the LBO route. The higher leverage and the pressure to generate cash led to a performance increase at Kroger, but the performance improvement at Safeway was importantly higher. Why should this have been so?

Top managers at Safeway put part of their wealth at stake and hold substantial equity stakes (amounting to a total of about 20%) such that every managerial decision has a important direct impact on their wealth. In addition, management is even more directed towards a focus on value as its bonuses are linked to the market value of assets and managers receive stock options.

There is a major external stockholder (the private equity firm Kohlberg Kravis Roberts, or KKR) that monitors the firm closely and ensures that management does not maximize its private benefits at the expense of other stockholders and the firm itself.

Safeway restructured the board to consist of management and representatives of KKR, who provided expertise on corporate restructuring.

Safeway restructured its operations more drastically than Kroger, closing stores that did not generate sufficient operational cash flows. It also cut back on discretionary expenses, such as advertising and maintenance, to meet its short-term debt obligations, and it cut non-core business.

Safeway removed leverage-induced cash flow streams as fast as possible through asset sales in order to increase capital expenditures.

ConclusionFirms that undergo leveraged (management) buyouts have important advantages over publicly listed firms. First, the high leverage creates value through the tax shield. Second, the management is incentivized to focus on value creation for the reason that it (co-) owns the firm (in the case of a MBO/MBI) or for the reason that strict monitoring of the incumbent management is induced by the major stockholders (in the case of an IBO). The organizational structure reduces the firm’s free cash flow such that money is not squandered by investing in negative-NPV projects. The private status of the firm requires little information disclosure compared to a listed firm, which allows the firm to avoid expenses related to compliance with the regulations on corporate governance/transparency.It should be emphasized that not every firm is a good candidate for LBO. The requirements are: stable cash flows, low and predictable capital investment needs, a liquid balance sheet with collateralizable assets, an established market position, and being in a recession-proof industry.

EXECUTIVE COMPENSATION PLANS

Opposite to media “noise,” the issue of executive compensation is not a recent predicament: economist Adam Smith analyzed it more than 200 years ago. Weak boards, discrete ownership and an ill-informed financial community all add to the problem. But is there a solution? Nowadays, more people than ever before are covered by executive compensation and employee benefits plans and agreements, which have turn into vital tools to attract and retain top talent.

Principles for Executive Compensation

(1) Philosophy: The company should have a clear arching over compensation philosophy. The philosophy should encourage an alignment of interests between management and shareholders. The company’s compensation philosophy should aim to create long-term value while not incentivizing excessive risk taking. The company philosophy should be flexible enough to allow for sensible and fair return in changing market conditions.• Compensation Committee: The compensation committee should be accountable for

developing the philosophy and making sure that it becomes part of the company culture. Roles and accountabilities of members, committee makeup, and information gathering and processing should also be addressed.

• Plan Elements: The company philosophy should address total compensation as well as the relative mix of base, bonus, and long-term incentive elements of the plan. In addition, the company philosophy should address the use of cash, equity, and equity-like compensation.

• Risk management: The company philosophy should address the risks to compensation expenditures as well as risk posed by compensation metrics. Part of this risk analysis should include the issue of unintended drivers or consequences related to incentive compensation.

(2) Design: The design of the company’s compensation policy should be comprehensive and discuss in detail all relevant components. A important portion of plans should be performance based. Discussion should include why certain elements were used as well as why certain elements were not incorporated.• Incentive and Bonus: The policy should contain the intended forms of incentive and bonus

compensation. In addition, the policy should discuss the rationale for use of guaranteed bonuses, the types of metrics used, and the rationale behind adjusted presentation metrics and non-financial metrics.

• Equity Compensation: The policy should address each form of equity and equity-like compensation and the company’s overall objectives in utilizing these tools. Conversation of award structures, as well as the size, timing, valuation, and terms of grants should be included. Additionally, the company’s approach to equity ownership and retention guidelines should be included.

• Dilution & Repricing: Companies should give clear policies on repricing and how compensation plans might lead to dilution of existing shareowners.

• Contractual Arrangements: The policy should enclose the parameters by which the company will make use of employment agreements, severance agreements or other contractual arrangements including post-employment agreements.

(3) Transparency: A well expressed compensation plan increases efficiency and may reduce the instances in which the company and investors are surprised by outcomes related to the compensation program, thereby reducing the negative reaction in the marketplace to compensation related events.• Full disclosures: Companies should provide full disclosure of plans and how they are

intended to be used.• Plain English: Philosophy, plans, and disclosures should be easily understood and presented

in plain English.(4) Accountability: The compensation committee is finally accountable for designing, implementing,

monitoring, and evaluating the executive compensation program.• Roles and responsibilities: Clearly defined responsibilities of the compensation committee

demonstrate rigor in creating and implementing compensation plans. The compensation committee charter should clearly outline these roles and responsibilities.

• Shareholder Approval: Compensation plans, repricing of awards within plans, and an annual advisory vote on compensation should always be submitted for shareholder approval.

(5) Independence: The compensation process should be conducted by independent compensation committee members utilizing only independent advisors and compensation consultants.

Executive Compensation in Present Day Organisational Culture

Executive compensation does not merely indicate base salary; it is the total remuneration an upper-level manager receives in a corporation. This normally comprises benefits such as bonuses, deferred and restricted stock, vesting periods, pensions and perquisites, as well as terms of employment as well as performance metrics, clawback provisions, and golden parachutes.

A Real Opportunity to Improve Incentives

Contemporary situation risks more than bad public relations for public companies: there is a danger of missing out on an opportunity to get better incentives for top managers to increase the long-term value of their enterprises. Taking benefit of this opportunity has main ramifications not only for corporate stakeholders, but for society in common, as it has the potential to create more productive, durable, and valuable companies.

There is a real necessity to bring a unruffled analysis to executive compensation and go beyond the superficial caricatures of unbridled greed. Economics can help by shining a light on the purpose of a corporate form of organization and the costs and benefits of comparative mechanisms for aligning executives’ behavior with shareholders and the broader stakeholder community. Remember that well-constructed executive compensation packages are necessary but not adequate for long-term value creation.

Recognize the Objective

What is the aim of a corporation? Maximizing the long-term total enterprise value of the firm has long been understood as the chief corporate mission. Basically said, this means expanding the organization’s market reach or improving real productive capacity. Demonstrating a culture of long-term value creation implies gaining the loyalty and commitment of all constituencies, comprising employees, suppliers, and the wider community. The main challenge for management is to create the corporate vision, strategy, and tactics to unite and guide all constituents. So much depends on the stock of trust and flow of honest information.

Aligning Interests

A corporation requires to make sure that executive leaders’ incentives are aligned with shareholder interest in long-term value creation. Contrasting owner-managers of small firms, executives at large enterprises often have only a small portion of their own equity at stake. Nevertheless, this disparity leads to a gap in their interest as the “agent” looking after the interest of the “principals” (e.g. shareholders and debtholders). What’s more, they have an information advantage over principals, which can confer rise to a serious conflict of interests and accountability problems. The economics of principals

and agents helps organize thinking about how the structure of the CEO’s compensation will have an effect on managerial behavior in quest of the objectives of the firm’s various stakeholders.

The principal-agent relationship is complicated and costly to uphold effectively over the long run, which means it is critical to take advantage of alternative alignment forces and mechanisms such as government regulation to monitor against fraud and accounting manipulation. Furthermore, the market for corporate control offers outside active investors and strategic companies a way to act as a corrective force against managers who do not maximize the value of their firms and wander away too far from their shareholders’ interests. But recent events suggest that this discipline has restrictions, particularly in the recent era of overvalued equity.

The Role of Corporate Governance

Corporate governance is an instrument for aligning principal-agent interests and incentives, encouraging accountability. How does it accomplish this? Corporate governance seeks to reconcile the relationships among the CEO, board members, stockholders, and the outside financial community of analysts, bondholders, and other creditors by clearly assigning responsibilities, measuring performance, and worthwhile or penalizing managers in line with their impact on the firm’s long-term value creation.

There are some persuasive examples that the sine qua non of a well-run, healthy company is successful coordination of the terms of executive compensation with the corporate governance function: rules, monitoring, and other incentives promoting effective relationships among significant constituencies. Corporate governance, in theory at least, serves as a kind of “check and balance” for a corporation to make sure that executive compensation packages attract and retain the right people, hasten the departure of the wrong people, and provide incentives for high performance.

Dealing With Short-Termism

An important part of the problem in executive compensation can be traced to how compensation packages evolved and became more closely tied to short-term stock prices. For instance, pay plans have tended to move away from using fixed salaries. Compensation plans are now concerted in stock options. And deferred stock and stock purchase options, which tend to vest over short time periods, are common. Furthermore, options are characteristically tied to short-run publicly traded stock prices. The table overleaf shows the recent trend of executive compensation increasing in the form of equity.

Of course, CEOs are supposed to be paid for performance, but this shift in the make-up of compensation packages has led to some unforeseen repercussions: it has skewed managerial decisions to favor the short term and created a marketplace hothouse of overvalued equity.

The introduction of stock options and limited stock grants with such features as short-term vesting periods seemed to be the universal remedy for aligning shareholder and manager interests. Equity compensation soon evolved into the greatest portion of total executive compensation. At the same time, its growth facilitated a climate where information became less reliable, even though superficially more plentiful, making it more and more complex for analysts or investors to make reliable decisions –

for example, financial analysts underestimated Microsoft’s quarterly earnings 41 out of 42 times, according to an SEC investigation and cease and desist order in 2002.

Imagine the CEO of an international engineering and construction company grappling with whether to bite the bullet and overhaul the firm’s engineering software and invest in an even more costly three-year employee training program. If carried out, the costs of the investment could impact earnings and the company’s share price, affecting the CEO’s various stock options and restricted stock share over the next three years. Delaying the investment may advantage the value of the CEO’s personal stock options, but the longer he waits, the greater the decline in long-term company and shareholder value as his engineers carry on to fall behind.

Linking Compensation and Corporate Governance

So, what can this examination tell us about the present argument over executive pay packages? Is the argument symptomatic of a real problem, or an unfortunate matter of timing and appearances? What is clear is that executive compensation and corporate governance are inextricably linked and considerable effort and cost are obligatory to better align incentives and reduce opportunism of all corporate stakeholders – top executives, board chairs and members, stockholders, debt holders – and the financial community.

The seminal idea behind the use of limited stock grants and stock options was to augment the performance of managers, or make “pay for performance” a reality in the corporate world. But the condition was still far removed from the ideal of all responsibility and the fruits of presentation concentrated in the owner-manager. Giving way to the stock options or limited stock shares that vested quickly or over short time-horizons contributed nothing toward managers having “skin in the game.” (Private equity, hedge funds, LBOs and other alike entities have basically solved the “skin in the game” problem in the way that managing partners are compensated.)

It soon became clear that grants of stock options and limited stock were not free of cost to companies and were also failing to have the preferred effect of making top management more vested in companies’ long-term goals. In reply, a number of corporations began to require CEOs to purchase and hold stock with after-tax dollars or variants with similar effects. A number of companies that exchanged cash bonuses, grants of options or restricted stock for their longer-term equivalents include ADC Telecommunications, Arkla, Avon, Baxter, Black & Decker, Clorox, EKCO and General Mills. FTI Consulting will be monitoring the performance of these companies and others that adopt similar compensation strategies over the coming years.

Changing Corporate Culture

Constructing a culture of accountability and responsibility is necessary to address issues around executive compensation. Better transparency and credible, independent checks and balances are the least requirements to regain shareholder and broader stakeholder trust. Companies should also put in place a system that describes potential conflicts of interests and implements countervailing measures. These comprise:

Avoid the intrinsic risk of hiring the same compensation firm for the rank-and-file employees and the CEO or other top managers.

Put into practice a set of measures to monitor accounting and other reporting practices that recommend weaknesses in corporate governance.

Do not provide multi-period compensation packages. Cautiously monitor CEO and financial analyst relationships. Set up an independent compensation committee composed of board members without CEO

participation. Commission financial analysis of compensation packages with another scenarios and expected

outcomes in terms of attraction, separation, and incentives. Limit to only the CEO and certain top executives the grant of options and deferred stock, and

with these, make every effort to set up true estimates of costs to the company and their impact on the CEO and firm value.

As a significant component of corporate governance, executive compensation must continually strive to align the incentives of top managers with shareholders and broader stakeholders. As the value added by management is complex and costly to measure, monitor and verify, this is a complex and challenging responsibility for which much is at stake.

Obviously, ensuring effective measurement and monitoring of the value added by top management is complex and costly but, equally, it is necessary for protecting stakeholders and increasing long-term total enterprise value of the firm.

CORPORATE RESTRUCTURING AND CONTROL

Corporate restructuring is one of the most intricate and basic phenomena that management confronts. Each company has two opposite strategies from which to choose: to expand or to refocus on its core business. While diversifying represents the expansion of corporate activities, refocus characterizes a

concentration on its core business. From this viewpoint, corporate restructuring is reduction in diversification.Corporate restructuring is an episodic exercise, not related to investments in new plant and machinery which involve an important change in one or more of the following

Pattern of ownership and control Composition of liability Asset mix of the firm.

It is an all-inclusive process by which a co. can consolidate its business operations and make stronger its position for achieving the desired objectives:

(a) Synergetic(b) Competitive(c) Successful

It comprises important re-orientation, re-organization or realignment of assets and liabilities of the organization through cognizant management action to improve future cash flow stream and to make more profitable and well-organized.Need for Corporate Restructuring

Corporate restructuring is the procedure of redesigning one or more aspects of a company. The process of reorganizing a company may be implemented due to a number of diverse factors, such as positioning the company to be more competitive, survive a presently unfavorable economic climate, or poise the corporation to move in a totally new direction. Here are some instances of why corporate restructuring may take place and what it can mean for the company.Restructuring a corporate entity is often an essential when the company has grown to the point that the original structure can no longer competently manage the output and general interests of the company. For instance, a corporate restructuring may need for spinning off some departments into subsidiaries as a means of creating a more effective management model as well as taking benefit of tax breaks that would allow the corporation to divert more revenue to the production process. In this situation, the restructuring is viewed as a positive sign of growth of the company and is often welcome by those who wish to see the corporation get a larger market share. Corporate restructuring may also take place as a result of the acquisition of the company by new owners. The acquisition may be in the form of a leveraged buyout, a hostile takeover, or a merger of some type that keeps the company intact as a subsidiary of the controlling corporation. When the restructuring is due to a hostile takeover, corporate raiders often implement a dismantling of the company, selling off properties and other assets in order to make a profit from the buyout. What remains after this restructuring may be a smaller entity that can continue to function, albeit not at the level possible before the takeover took placeIn common, the idea of corporate restructuring is to allow the company to continue functioning in some manner. Even when corporate raiders break up the company and leave behind a shell of the original structure, there is still frequently a hope, what remains can function well adequate for a new buyer to buy the diminished corporation and return it to profitability.Purpose of Corporate Restructuring -

To enhance the share holder value, The company should continuously evaluate its:1. Portfolio of businesses,2. Capital mix,3. Ownership &4. Asset arrangements to find opportunities to increase the share holder’s value.

To focus on asset utilization and profitable investment opportunities.

Corporate RestructuringActivities

ExpansionMergers & acquisition

Tender offersJoint venture

SELL-OFFsSpin-offSplit-off

Equity carve-out

Corporate controlPremium buy-back

Standstill AgreementsAnti-take overProxy contests

To reorganize or divest less profitable or loss making businesses/products. The company can also enhance value through capital Restructuring, it can innovate securities

that help to reduce cost of capital.Characteristics of Corporate Restructuring

1. To improve the company’s Balance sheet, (by selling unprofitable division from its core business).

2. To accomplish staff reduction ( by selling/closing of unprofitable portion)3. Changes in corporate mgt4. Sale of underutilized assets, such as patents/brands.5. Outsourcing of operations such as payroll and technical support to a more efficient 3rd party.6. Moving of operations such as manufacturing to lower-cost locations.7. Reorganization of functions such as sales, marketing, & distribution8. Renegotiation of labor contracts to reduce overhead9. Refinancing of corporate debt to reduce interest payments.10. A major public relations campaign to reposition the co., with consumers.

Category of Corporate RestructuringCorporate Restructuring entails a range of activities including financial restructuring and organization restructuring.

Financial Restructuring

Financial restructuring is the restructuring of the financial assets and liabilities of a corporation to to create the most advantageous financial atmosphere for the company. The process of financial restructuring is often connected with corporate restructuring, in that restructuring the general function and composition of the company is likely to effect the financial health of the corporation. When finished, this reordering of corporate assets and liabilities can help the company to remain competitive, even in a miserable economy.

Almost all businesses goes through a phase of financial restructuring at one time or another. In a few cases, the process of restructuring takes place as a means of apportioning resources for a new marketing campaign or the launch of a new product line. When this happens, the restructure is often seemed as a sign that the company is financially established and has set goals for future growth and expansion.

Figure 6.2: Corporate Restructuring Activities

CORPORATE RESTRUCTURING- Leveraged Buyout, Hostile Takeover & Merger

Corporate restructuring may take place due to the acquisition of the company by new owners. The acquisition may be in the shape of a leveraged buyout, a hostile takeover, or a merger of some type that keeps the company intact as a subsidiary of the controlling corporation.Hostile TakeoverA hostile takeover is a kind of corporate takeover which is carried out against the wishes of the board of the target company. This unique type of acquisition does not occur nearly as frequently as friendly takeovers, in which the two companies work together for the reason that the takeover is perceived as beneficial. Hostile takeovers can be traumatic for the target company, and they can also be risky for the other side, as the acquiring company may not be able to obtain certain relevant information about the target company.Companies are purchased and sold on a daily basis. There are two kinds of sale agreements. In the first, a merger, two companies come together, blending their assets, staff, facilities, and so forth. After a merger, the original companies cease to exist, and a new company comes up in its place. In a takeover, a company is purchased by another company. The purchasing company owns all of the target company's assets as well as company patents, trademarks, and so forth. The original company may be completely swallowed up, or may operate semi-independently under the umbrella of the acquiring company.Characteristically, a company which wishes to acquire another company approaches the target company's board with an offer. The board members mull over the offer, and then choose to accept or reject it. The offer will be accepted if the board believes that it will promote the long term welfare of the company, and it will be rejected if the board dislikes the terms or it feels that a takeover would not be beneficial. When a company engages in the takeover after rejection by a board, it is a hostile takeover. If a company bypasses the board completely, it is also termed a hostile takeover.Publicly traded companies are at risk of hostile takeover for the reason that opposing companies can purchase large amounts of their stock to gain a controlling share. In this instance, the company does not have to respect the feelings of the board for the reason that it already essentially owns and controls the firm. A hostile takeover may also comprise tactics like trying to sweeten the deal for individual board members to get them to agree.

An acquiring firm takes a risk by attempting a hostile takeover. For the reason that the target firm is not cooperating, the acquiring firm may without knowing take on debts or serious problems, since it does not have access to all of the information about the company. A lot of firms also have trouble getting financing for hostile takeovers, since some banks are reluctant to lend in these situations.

Merger and Acquisition

A merger takes place when two companies unite to form a single company. A merger is very alike to an acquisition or takeover, except that in the case of a merger existing stockholders of both companies comprised retain a shared interest in the new corporation. By distinction, in an acquisition one company purchases a bulk of a second company's stock, creating an uneven balance of ownership in the new combined company.The acquisition of a business may be structured in a number of ways, comprising, an asset sale, a stock sale, or a merger. The structure of the acquisition will be determined by a variety of accounting, business, legal, and tax considerations. In spite of the structure of the transaction, acquisition agreements have the subsequent four common and very significant features which are examined in this discussion: (a) representations and warranties; (b) pre-closing covenants; (c) conditions precedent to closing; and (d) indemnification.

Representations and Warranties

The vendor and the purchaser will make delegacies and warranties to the other in the acquisition agreement. The vendor’s representations and warranties characteristically make up the major part of the acquisition agreement. Delegacies and warranties serve three significant purposes. First, they are informational. The seller’s representations and warranties, united with the buyer’s due diligence, facilitate the buyer to learn as much as possible about the seller’s business prior to signing the definitive acquisition agreement. Second, they are protective. The seller’s representations and warranties provide a mechanism for the buyer to walk away from, or perhaps to renegotiate the terms of, the acquisition, if the buyer discovers facts that are contrary to the representations and warranties between the signing and the closing. Third, they are supportive. The seller’s representations and warranties make available the framework for the seller’s restitution obligations to the buyer after the closing.

Prior to signing the acquisition agreement, the buyer will want to learn as much as possible about the seller’s business. Consequently, the buyer will require the seller to make extensive representations and warranties about its business. Several of these representations and warranties will be specific to the seller’s industry. Nevertheless, the most common representations and warranties comprise:

(a) corporate organization, authority, and capitalization; (b) assets; (c) liabilities; (d) financial statements; (e) taxes; (f) contracts, leases, and other commitments;

(g) employment matters; (h) compliance with laws and litigation; (i) product liability; and (j) environmental protection.

From the seller’s viewpoint, if the buyer is paying the purchase price in cash at the closing, the most significant representations and warranties the seller can elicit from the buyer are those governing the buyer’s corporate authorization and financial condition (i.e., the buyer’s ability to pay the purchase price). If the buyer is paying the purchase price over time or by issuing stock, the seller will need more extensive representations and warranties from the buyer.

Pre-Closing Covenants

The second major characteristic of merger and acquisition agreements is the addition of various pre-closing covenants, or promises to do something, or not do something, during the period between the signing of the acquisition agreement and the closing. Usually, covenants are absolute; nevertheless, some may be subject to a "reasonable efforts" qualification. Other than covenants relating to corporate approvals and governmental filings and approvals, compliance with a particular covenant may be forewent by the party that benefits from the covenant.

There are two types of pre-closing covenants: negative covenants and affirmative covenants. Negative covenants restrict the seller from taking certain actions prior to the closing exclusive of the buyer’s prior consent. Negative covenants protect the buyer from the seller taking actions prior to the closing that change the business that the buyer expects to buy at the closing. Characteristic negative covenants comprise:

(a) not changing accounting methods or practices; (b) not entering into transactions or incurring liabilities outside the ordinary course of business or in

excess of certain amounts; (c) not paying dividends or making other distributions to stockholders; (d) not amending or terminating contracts; (e) not making capital expenditures; (f) not transferring assets; (g) not releasing claims or waiving rights; and (h) not doing anything that would make the seller’s representations and warranties untrue.

Affirmative covenants obligate the seller or the buyer to take certain actions prior to the closing. Typical affirmative covenants include:

(a) allowing the buyer full access to the seller’s books, records, and other properties; (b) obtaining the necessary board and stockholder approvals; (c) obtaining the necessary third party consents; and (d) making the required governmental filings and obtaining the required governmental approvals.

Conditions to Closing

Merger and acquisition agreements usually also enclose numerous conditions to closing, which are certain obligations that must be fulfilled in order to legally necessitate the other party to close the transaction. Other than conditions to closing relating to corporate approvals and governmental filings and approvals, acquiescence with a particular condition to closing may be waived by the party that benefits from the condition. The interplay between pre-closing covenants and conditions to closing is significant. If a particular matter is addressed exclusively as a covenant, the buyer’s only remedy for the breach of the covenant will be monetary damages. Nevertheless, if a particular matter also is addressed as a condition to closing, the buyer can walk away from the transaction if the condition is not satisfied. In addition, and perhaps equally as significant, the buyer may be able to use the threat of not closing as leverage to renegotiate the terms of the transaction.

All merger and acquisition agreements present that, as a condition to closing, the representations and warranties of the parties must be accurate and truthful at the closing, and that the pre-closing covenants have been performed or fulfilled prior to the closing. This is in general confirmed by each party delivering a written certificate to that effect to the other party.

Other typical conditions to closing include:

(a) receipt of the necessary third party consents; (b) receipt of the necessary governmental approvals; (c) receipt of legal opinions and other closing documents; (d) receipt of certain financial statements or the achievement of certain financial milestones; (e) receipt of employment or non-competition agreements from key employees; and (f) satisfactory completion of the buyer’s due diligence of the seller’s business.

Indemnification

The last main characteristic of typical merger and acquisition agreements is indemnification. Nevertheless, indemnification provisions are strange in agreements for the acquisition of a public company. Indemnification requirements protect the parties from definite matters that occur after the closing and allocate the risks and responsibilities for these occurrences between the buyer and the seller. Indemnification provisions characteristically address breaches of covenants or representations and warranties that are discovered after the closing. In addition, indemnification provisions address items that are disclosed in the seller’s representations and warranties and for which the seller retains accountability after the closing. An instance is pending litigation, the outcome and amount of damages of which cannot be predicted and reflected in the purchase price. Consequently, the buyer may need the seller to stay responsible for the litigation after the closing. The buyer may also ask for separate indemnification for environmental and tax liabilities beyond the seller’s representations and warranties.

Usually, indemnification provisions are a great deal negotiated, and the seller will seek to limit its post-closing indemnification obligations in numerous ways. First, the seller will try to limit the time period

after the closing for which it has indemnification obligations. In theory, this time period should be based upon the reasonable period of time within which the buyer, through reasonable diligence, should have discovered the breach and misrepresentation or, if relevant, the time period within which a third party would make its claim. In practice, the parties usually agree on a period of one to three years after the closing. Exceptions may be made for environmental and tax liabilities, for which the time period may be the applicable decree of limitations.

Second, the seller will try to impose a cap on the total amount of its indemnification liability. A lot of sellers try to cap their liability at an amount less than the total purchase price. A lot of buyers will agree to a cap equal to the total purchase price. If the seller’s business is "clean," the risk to the buyer in in accord to an indemnification cap may be small.

Third, the seller will try to negotiate a "basket" or a "deductible" on its indemnification obligations in order to get rid of small indemnification claims. A "basket" or "deductible" provides that the seller does not have liability to the buyer until the amount of the buyer’s losses exceed a definite amount. In the case of a "basket," when the buyer’s losses exceed the agreed upon "basket" amount, the seller is legally responsible for the total amount of the losses. In the case of a "deductible," when the buyer’s losses go beyond the agreed upon "deductible" amount, the seller is liable only for the surplus amount of the losses above the "deductible."

In order to make sure that there are funds available to gratify the seller’s indemnification obligations, the buyer may necessitate that a portion of the purchase price be held in escrow by a third party for a period of time after the closing. On the other hand, the buyer may hold back a portion of the purchase price and give the seller a promissory note for that portion but retain the right to offset the promissory note to satisfy its indemnification claims.

The whole merger process is generally kept secret from the general public, and often from the majority of the employees at the concerned companies. In view of the fact that the majority of merger attempts do not succeed, and most are kept secret, it is difficult to presume how many potential mergers occur in a given year. It is likely that the number is very high, nevertheless, given the amount of successful mergers and the desirability of mergers for many companies.

A merger may be required for various reasons, some of which are advantageous to the shareholders, some of which are not. One use of the merger, for instance, is to combine a very profitable company with a losing company in order to use the losses as a tax write-off to offset the profits, while expanding the corporation as a whole.Growing one's market share is another major use of the merger, chiefly amongst large corporations. By merging with major competitors, a company can come to control the market they compete in, giving them a freer hand with regard to pricing and buyer incentives. This form of merger may cause problems when two overtopping companies merge, as it may trigger litigation concerning monopoly laws.Another kind of popular merger brings in concert two companies that make dissimilar, but complementary, products. This may also comprise purchasing a company which controls an asset your company utilizes anywhere in its supply chain. Major manufacturers buying out a warehousing chain in order to save on warehousing costs, as well as making a profit in a straight line from the purchased

business, is a good instance of this. PayPal's merger with eBay is another good instance, as it allowed eBay to avoid fees they had been paying, while tying two complementary products together.A merger is generally handled by an investment banker, who aids in transferring ownership of the company all the way through the strategic issuance and sale of stock. Some have supposed that this relationship causes some troubles, as it provides an incentive for investment banks to push existing clients towards a merger even in cases where it may not be beneficial for the stockholders.Mergers and acquisitions are means by which corporations unite with each other. Mergers take place when two or more corporations become one. To protect shareholders, state law provides procedures for the merger. A vote of the board of directors and then a vote of the shareholders of both corporations is more often than not required. Subsequent to a merger, the two corporations come to an end to exist as separate entities. In the classic merger, the assets and liabilities of one corporation are automatically transferred to the other. Shareholders of the vanishing company become shareholders in the surviving company or receive compensation for their shares.Mergers may come as the consequence of a negotiation between two corporations interested in combining, or when one or more corporations "target" another for acquisition. Combinations that happen with the approval and encouragement of the target company's management are known "friendly" mergers; combinations that occur despite opposition from the target company are known as "hostile" mergers or takeovers. In either case, these consolidations can bring together corporations of approximately the same size and market power, or corporations of vastly diverse sizes and market power.The term "acquisition" is normally used when one company takes control of another. This can take place through a merger or a number of other techniques, such as purchasing the greater part of a company's stock or all of its assets. In a purchase of assets, the transaction is one that must be negotiated with the management of the target company. Compared to a merger, an acquisition is treated in a different way for tax purposes, and the acquiring company does not essentially assume the liabilities of the target company.A "tender offer" is a most famous way to purchase a majority of shares in another company. The acquiring company makes a public offer to buy shares from the target company's shareholders, therefore by passing the target company's management. In order to induce the shareholders to sell, or "tender,” their shares, the acquiring company characteristically offers a purchase price higher than market value, often considerably higher. Certain situations are often placed on a tender offer, such as necessitating the number of shares tendered be adequate for the acquiring company to gain control of the target. If the tender offer is successful and a sufficient percentage of shares are acquired, control of the target company through the normal methods of shareholder democracy can be taken and after that the target company's management replaced. The acquiring company can also use their control of the target company to bring about a merger of the two companies.Habitually, a successful tender offer is accompanied by a "cash-out merger." The target company (now controlled by the acquiring company) is merged into the acquiring company, and the left behind shareholders of the target company have their shares transformed into a right to receive a certain amount of cash.Another ordinary merger variation is the "triangular" merger, in which an auxiliary of the surviving company is formed and then merged with the target. This protects the surviving company from the liabilities of the target by keeping them within the subsidiary rather than the parent. A "reverse triangular merger" has the acquiring company make a subsidiary, which is then merged into the target company. This form preserves the target company as a continuing legal entity, though its control has passed into the hands of the acquirer.Usually, mergers and other sorts of acquisitions are performed in the hopes of realizing an economic gain. For such a transaction to be justified, the two firms concerned must be worth more together than

they were apart. Some of the potential benefits of mergers and acquisitions comprise attaining economies of scale, combining complementary resources, garnering tax advantages, and eliminating inefficiencies. Other reasons for bearing in mind growth through acquisitions comprise getting proprietary rights to products or services, increasing market power by purchasing competitors, shoring up weaknesses in key business areas, penetrating new geographic regions, or providing managers with new opportunities for career growth and advancement. Since mergers and acquisitions are so compound, nevertheless, it can be very complicated to evaluate the transaction, define the related costs and benefits, and handle the resulting tax and legal issues.When a small business owner opts to merge with or sell out to another company, it is at times called "harvesting" the small business. In this condition, the transaction is anticipated to release the value locked up in the small business for the benefit of its owners and investors. The thrust for a small business owner to pursue a sale or merger may comprise estate planning, a requirement to diversify his or her investments, a failure to finance growth independently, or a simple need for change. Furthermore, some small businesses find that the best way to grow and compete against larger firms is to merge with or acquire other small businesses.Principally, the decision to merge with or acquire an additional firm is a capital budgeting decision much like any other. But mergers vary from ordinary investment decisions in at least five ways. First, the value of a merger may depend on such things as strategic fits that are hard to measure. Second, the accounting, tax, and legal aspects of a merger can be complex. Third, mergers often involve issues of corporate control and are a means of replacing existing management. Fourth, mergers clearly affect the value of the firm, but they also affect the relative value of the stocks and bonds. Lastly, mergers are often "unfriendly."Bird’s Eye View of the Benefits Accruing from Mergers and AcquisitionsThe main advantages from mergers and acquisitions can be listed as increased value generation, increase in cost efficiency and increase in market share. Mergers and acquisitions often lead to an increased value generation for the company. It is anticipated that the shareholder value of a firm after mergers or acquisitions would be bigger than the sum of the shareholder values of the parent companies. An increase in cost efficiency is exaggerated through the procedure of mergers and acquisitions. This is for the reason that mergers and acquisitions lead to economies of scale. This in turn promotes cost efficiency. As the parent firms amalgamate to form a larger new firm the scale of operations of the new firm increases. As output production rises there are chances that the cost per unit of production will come down. DemergerDemergers are circumstances in which divisions or subsidiaries of parent companies are split off into their own independent corporations. The process for a demerger can differ somewhat, depending on the reasons behind the implementation of the split. Usually, the parent company maintains some degree of financial interest in the newly formed corporation, although that interest may not be enough to uphold control of the functionality of the new corporate entity.

It results in the transfer by a company of one or more of its undertakings to one more company. The company whose undertaking is transferred is known as the demerged company and the company (or the companies) to which the undertaking is transferred is referred to as the resulting company.

A demerger may take the form of - A spinoff or a split-up.

Methods of Corporate Restructuring

Joint ventures Sell off and spin off Divestitures Equity carve out Leveraged buy outs (LBO) Management buy outs Master limited partnerships Employee stock ownership plans (ESOP)

Joint Venture -

Joint ventures are new-fangled enterprises owned by two or more participants. They are characteristically created for special purposes for a limited duration. It is a mixture of subsets of assets contributed by two (or more) business entities for a explicit business purpose and a limited duration. Each of the venture partners continues to exist as a separate firm, and the joint venture represents a new business enterprise. It is a agreement to work together for a period of time each participant expects to gain from the activity but also must make a contribution.Reasons for Forming a Joint Venture

Construct on company's strengths Spreading costs and risks Improving way in to financial resources Economies of scale and benfits of size Right of entry to new technologies and customers Right of entry to innovative managerial practices

Rational For Joint Ventures

To supplement inadequate financial or technical ability to enter a particular line or business. To share technology and generic management skills in organization, planning and control. To diversify risk To attain distribution channels or raw materials supply To attain economies of scale To make bigger activities with smaller investment than if done separately To take advantage of favorable tax treatment or political incentives (particularly in foreign

ventures).Tax aspects of joint venture.

If a corporation lends a patent technology to a Joint Venture, the tax effects may be less than on royalties earned though a licensing arrangements.

Example –One partner leads the technology, while another leads depreciable facilities. The depreciation offsets the revenues falling to the technology. The J.V. may be taxed at a lower rate than any of its partner and the partners pay a later capital gain tax on the returns realized by the J.V. if and when it is sold. If the J.V. is organized as a corporation, only its assets are at risk. The partners are responsible only to the extent of their investment, this is chiefly significant in hazardous industries where the risk of workers, production, or environmental liabilities is high.

Spin Off

Spinoffs are a way to eliminate the underperforming or non-core business divisions that can pull down profits.

Process of spin off

1. The company determines to spin off a business division. 2. The parent company files the essential paperwork with the Securities and Exchange Board of the

Country. 3. The spinoff becomes a company of its own and must also file paperwork with the Securities and

Exchange Board of the country. 4. Shares in the new company are distributed to parent company shareholders. 5. The spinoff company goes public.

Take in considerations that the spinoff shares are disseminated to the parent company shareholders. There are two reasons why this creates value:

1. Parent company shareholders infrequently want anything to do with the new spinoff. After all, it's an underperforming division that was cut off to get better the bottom line. As a consequence, several new shareholders sell right away after the new company goes public.

2. Big institutions are often prohibited to hold shares in spinoffs due to the smaller market capitalization, increased risk, or poor financials of the new company. Consequently, a lot of large institutions automatically sell their shares directly after the new company goes public.

Trouble-free supply and demand logic says us that such great number of shares on the market will obviously decrease the price, even if it is not basically justified. It is this temporary mispricing that gives the enterprising investor an opportunity for profit.There is no money transaction in spin-off. The transaction is treated as stock dividend and tax free exchange.

Split-offIs a transaction in which some, but not all, parent company shareholders take delivery of shares in a subsidiary, in return for dispensing with their parent company’s share.In other words some parent company shareholders take delivery of the subsidiary’s shares in return for which they must give up their parent company shares.FeaturesA segment of existing shareholders receives stock in a subsidiary in exchange for parent company stock.Split-upIs a transaction in which a company spins off all of its subsidiaries to its shareholders & ceases to exist.

The entire firm is broken up in a series of spin-offs. The parent no longer exists and Only the new offspring survive.

In a split-up, a company is split up into two or more independent companies. As a continuation, the parent company vanishes as a corporate entity and in its place two or more separate companies come forth.Squeeze-out: the elimination of minority shareholders by controlling shareholders.

Sell Off

Selling a part or all of the firm by any one of means: sale, liquidation, spin-off and so on. Or General term for divestiture of part/all of a firm by any one of a no. of means: sale, liquidation, spin-off and so on.Partial Sell-Off

A partial sell-off/slump sale, comprises the sale of a business unit or plant of one firm to an additional firm.

It is the mirror image of a purchase of a business unit or plant. From the seller’s perspective, it is a form of contraction; from the buyer’s point of view it is a

form of expansion.

Motives for sell off

Raising capital Curtailment of losses Strategic realignment Efficiency gain.

Strategic RationaleDivesting a subsidiary can achieve a variety of strategic objectives, such as:

Anti-trust – Break up a business in response to anti-trust concerns. Corporate defense – Divest "crown jewel" assets to make a hostile takeover of Parent Company

less attractive. Eliminating dissynergies – Reduce bureaucracy and give Spin Company management complete

autonomy. Institutional sponsorship – Promote equity research coverage and ownership by sophisticated

institutional investors, either of which tend to validate SpinCo as a standalone business. Motivating management – Improve performance by better aligning management incentives with

Spin Co’s performance (using Spin Co’s, rather than Parent Company, stock-based awards), creating direct accountability to public shareholders, and increasing transparency into management performance.

Public currency – Create a public currency for acquisitions and stock-based compensation programs.

Undiversification – Divest non-core businesses and sharpen strategic focus when direct sale to a strategic or financial buyer is either not compelling or not possible

Unlocking hidden value – Establish a public market valuation for undervalued assets and create a pure-play entity that is transparent and easier to value

Divestitures Divesture is a transaction through which a firm sells a portion of its assets or a division to a different company. It comprises selling some of the assets or division for cash or securities to a third party which is an outsider. Divestiture is a form of narrowing for the selling company and means of expansion for the purchasing company. It represents the sale of a division of a company (assets, a product line, a subsidiary) to a third party for cash and or securities.

Mergers, assets purchase and takeovers lead to expansion in some way or the other. They are established on the principle of synergy which says 2 + 2 = 5! , divestiture on the contrary is based on the principle of “anergy” which says 5 – 3 = 3!.

Among the different techniques of divestiture, the most significant ones are partial sell-off, demerger (spin-off & split off) and equity carve out. Some authors define divestiture rather scarcely as partial sell off and some scholars define divestiture more generally to comprise partial sell offs, demergers and so on.

Motives for Divestitures

Antitrust Change of focus or corporate strategy Defend against takeover Good price. Need cash Sale to pay off leveraged finance Unit unprofitable can mistake

Equity Carve-Out

A transaction in which a parent firm extends some of a subsidiaries common stock to the general public, to bring in a cash infusion to the parent with no loss of control.Put differently equity carve outs are those in which some of a subsidiaries shares are offered for a sale to the general public, bringing an infusion of cash to the parent firm with no loss of control.Equity carve out is also a means of reducing their exposure to a riskier line of business and to boost shareholders value. Features of Equity Carve Out

A new control group is immediately created. A new legal entity is created. It is the sale of a minority or majority voting control in a subsidiary by its parents to outsider

investors. These are also referred to as “split-off IPO’s” The equity holders in the new entity need not be the same as the equity holders in the original

seller.Difference between Spin-off and Equity carve outs:

1. In a spin off, distribution is made pro rata to shareholders of the parent company as a dividend, a form of non cash payment to shareholders

1. In equity carve out; stock of subsidiary is sold to the public for cash which is received by parent company

2. In a spin off, parent firm no longer has control over subsidiary assets.In equity carve out, parent sells only a minority interest in subsidiary and retains control.Master Limited PartnershipMaster Limited Partnership’s are a kind of limited partnership in which the shares are openly traded. The limited partnership interests are divided into units which are traded as shares of common stock. Shares of ownership are called as as units. MLPs usually operate in the natural resource (petroleum and natural gas extraction and transportation), financial services, and real estate industries.

The benefit of a Master Limited Partnership is it combines the tax benefits of a limited partnership (the partnership does not pay taxes from the profit - the money is only taxed when unit holders obtain distributions) with the liquidity of a openly traded company.There are two types of partners in this type of partnership:

1. The limited partner is the person or group that provides the capital to the MLP and receives periodic income distributions from the Master Limited Partnership's cash flow

2. The general partner is the party responsible for managing the Master Limited Partnership's affairs and receives compensation that is linked to the performance of the venture.

Employees Stock Option Plan (ESOP)

An Employee Stock Option is a kind of defined contribution benefit plan that buys and holds stock. ESOP is a qualified, defined contribution, employee benefit plan designed to invest first and foremost in the stock of the sponsoring employer.Employee Stock Option’s are “qualified” in the sense that the ESOP’s sponsoring company, the selling shareholder and participants receive a variety of tax benefits.With an ESOP, employees never buy or hold the stock directly.Features

Employee Stock Ownership Plan (ESOP) is an employee benefit plan. The scheme provides employees the ownership of stocks in the company. It is one of the profit sharing plans. Employers have the benefit to use the ESOP’s as a tool to fetch loans from a financial institute. It also provides for tax benefits to the employers.

The benefits for the company: increased cash flow, tax savings, and increased productivity from highly motivated workers.The benefit for the employees: is the ability to share in the company's success.How it works?

Organizations strategically plan the ESOPs and make arrangements for the purpose. They make annual contributions in a special trust set up for ESOPs. An employee is eligible for the ESOP’s only after he/she has completed 1000 hours within a year

of service. After completing 10 years of service in an organization or reaching the age of 55, an employee

should be given the opportunity to diversify his/her share up to 25% of the total value of ESOP’s.

RECENT DEVELOPMENTS IN MERGERS AND ACQUISITIONS (M&A) ACTIVITIES

United States of America

Various major merger movements have taken place in the United States and each was more or less dominated by specific kind of merger. All of the merger movements occur when the economy experienced constant high rate of growth and conceded well particular development in business environments.

1895-1904 movements: the combination movement at the turn of the century constitutes chiefly of horizontal mergers, which resulted in high concentration in many industries; as well as heavy manufacturing industries. The period was one of rapid economic expansion. The movements peaked in 1899 and almost ended in 1903, when a severe economic recession set in mergers completed in the period 1887 through 1904 were estimated to involve 15 %of the total numbers of plants and employees comprising manufacturers in 1900 (Markhar, 1955).

The mergers of 1895-1904 followed by major changes in economic infrastructure and production technologies. The period was accompanied by the completion of the transcontinental railroad system, the advent of a national economic market and thus, making the way for regional firms becoming national firms (Markhan, 1955 Salter and Weinhold, 1980). While the preceding argument by Markhan does mergers some economists cast doubt on the possibility that large scale production was a motive to combination. Lynch points out three problems in economy of scale national.

(A) although scale economics can be more simply attained in combination of small firms than of large firms; the merger activity was concentrated in the large-firm category. Nevertheless, Markhan notes that nearly all the tabulation of early mergers were based on large mergers and did not comprise merger comprising a capitalization of less then $1 million (Markhan, 1955).

(B) Lynch’s second point is that combinations resulted in multi plant operations but scale economics are obtained when production is integrated by investments in large replacement facilities. Clearly, horizontal combination between geographically estranged plants will not have any production economics of scale in the absence of their physical integration. Nevertheless, economics of scale can exist not only in production but also in administration and marketing.

(C) Finally, lynch observe that merger activity in the early period occurred in a wide range of industries and technological advancements motivating horizontal mergers cannot surface in a number of industries with in a short time span. This appears to miss the point that economies of scale available not shape technological advancements in individual industries but rather became attainable from the reduction in transportations costs, whose impact could have been pervasive.

Europe

Since the middle 1950s a wave of takeovers, historically unprecedented in its scope, and its effects had swept through British industry. In a study of UK manufacturing industries by Ajit Singh (1971), it was found that 2,126 firms engaged in manufacturing, which were quoted on the U.K. stock exchanges in 1954; more than 400 had been acquired by 1960. Out of the next 100 large firms in 1954, 10 were taken over during the next 6 years. The number of unquoted manufacturing companies and other smaller concerns acquired in the same period runs into thousands. This take over movement has been far larger than those which occurred at the turn of the century and in the early 1920s.

The reasons for the enormous volume of acquisitions in the 1980s were manifold. The stock market in the UK, in harmony with markets in other countries, experienced a strong bull phase while culminated as the October crash of 1987, there was a mire relaxed, laissez-faire governmental attitude to mergers

and acquisitions embodied in the new vision of Thatcherism. The 1980s also witnessed divestments on a large scale.

The simultaneous increase in acquisitions and divestments suggest a considerable amount of corporate restructuring in the UK economy in recent years. Such restructuring has been made possible by new organizational innovations like, leveraged buyouts, management buyout (Sudersan 1995).

Even though Taggarts’ figure does not cover the period from 1987 to the present, junk financing was certainly curtailed by market conditions starting in late 1989. Even if junk bonds did not play a major role in the 1980s mergers deals, the dramatic increase in corporate leveraging during the 1980s served to deteriorate the quality of debt worldwide especially when coupled with high interest rates supported by governments contending with inflation caused by increase lightly, itself due in part to more lending and easier consumer credit. This has led to the highest levels of foreclosure and bankruptcies since the great depression in the 1930s.

THEORIES OF M&A ACTIVITIES

There is a very old and ongoing debate concerning firms’ motives for engaging in mergers and acquisitions. Neoclassical theories depict collective merger activity as firms’ value-enhancing response to industry-wide and/or economy-wide shocks. Behavioral and agency theories, on the other hand, view M&A’s as the resulting from investors’ and/or managers’ cognitive biases, or the inherent conflicts of interests between managers and investors. Some significant empirical aspects of the aggregate merger activity are widely accepted: mergers occur in waves; within each wave, they tend to cluster by industry; and, within industries, higher merger activity is associated with larger positive or negative shocks. Nevertheless, why these patterns emerge remains an open question.

The majority of existing theories of merger timing come out from completion in product markets. In this section we explore the impacts of product market competition and industry structure on firms’ incentives to connect in horizontal mergers and on the consequential dynamics of mergers. Ceteris paribus, a horizontal merger gains the combined value of the merging firms, due to the post-merger collusion in output markets. Nevertheless, the reduced level of competition following a merger also attracts new firms to enter the industry. Possible entry, in turn, minimizes the value of the officeholders and their incentives to merge. When an industry is in expansion, the value of the entry option is high in spite of of the industry structure, and the incumbents cannot deter entry by not merging. When an industry is in decline, entry is unbeneficial in spite of of the incumbents’ decision whether to merge. Therefore, in the extreme states of demand, the incumbents’ merger decision has limited effect on the entrant’s decision, and the effect of higher incumbents’ profits due to the post-merger collusion dominates the merger decision. In intermediate states, the merger decision has a extra marked effect on the likelihood of potential entry, and the incumbents may be better off not merging in order to delay entry. The reason above suggests that as a result of the interaction between incumbents’ decision to merge and new firms’ decision to enter an industry, horizontal mergers should occur with greater frequency during periods of extreme growth or decline in demand in comparatively intense industries.

There exists a considerable body of empirical proof reliable with the two main driving forces in our model: mergers increase incumbent firms’ market power and they may make possible entry by outsiders. For instance, Akhavein, Berger, and Humphrey (1997) and Prager and Hannan (1998) document that mergers between banks lead to greater than before market power and lower deposit interest rates. Borenstein (1990), Kim and Singhal (1993), and Singhal (1996) report alike evidence for the airline industry. A clinical study of acquisitions in the microfilm industry by Barton and Sherman (1984) provides evidence of important post-merger price increases. In addition, the effect of mergers on the incentives of new firms to enter an industry is not only a clear result of an oligopolistic competition model, but also a matter of explicit consideration by Antitrust authorities for the period of merger application reviews. Berger, Bonime, Goldberg, and White (2004) and Seeling and Critchfield (2003) find that mergers induce entry in the banking industry. As a result, it seems sensible that strategic considerations may effect firms’ incentives to merge and the resulting dynamics of horizontal mergers.

One proposition of the existing models of merger timing is that firms’ incentives to merge in periods of economic recession are diverse from those in periods of expansion. Lambrecht (2004) examines mergers motivated by operating synergies. In his model, mergers are likely to occur in expansions. In Lambrecht and Myers (2007), takeovers give out a mechanism to force disinvestment in declining industries. Their arguments lead to takeover transactions occurring mostly in industries that have experienced negative economic shocks. Similarly, in Mason and Weeds (2006) mergers in expansions are motivated by production synergies, while those in recessions allow consolidation and disinvestment. We show that in oligopolistic industries, strategic considerations increase the likelihood of observing horizontal mergers both in periods of rising and declining demand and that strategic incentives to attempt or delay horizontal mergers are more significant in more concentrated industries.

To highlight the effect of strategic considerations on the relation between the state of industry demand and takeover activity, in the base model we deliberately focus on the strategic aspects and abstract from other significant motives for merging. Strategic considerations, nevertheless, are but one of numerous factors that may affect firms’ decision to merge. Specifically, we do distinguish that assuming no production synergies, no merger costs, no operating leverage, and duopolistic competition may create overly stylized results. For that reason, we also discuss the intuition behind a number of extensions of the base model that relax these assumptions. These extensions show that strategic considerations carry greater weight in firms’ merger decisions when the costs of merging, operating synergies, and operating leverage are comparatively low, and when the degree of industry attentiveness is relatively high.Our theory is strongly related to recent models incorporating product market competition into the analysis of merger dynamics (e.g., Lambrecht (2004), Hackbarth and Miao (2007), and Yan (2006)). Our contribution to this literature is screening that the threat of potential entry is a crucial determinant of the dynamics of mergers, while entry is not allowed in existing models. Our model is also related to studies that examine entry into an industry within a dynamic setting (e.g., Dixit (1989), Baldursson (1998), Grenadier (2002), Fries, Miller and Perraudin (1997), Lambrecht (2001), and Zhdanov (2007)). We put in to the dynamic entry literature by incorporating the option of a merger initiated by incumbent firms. In addition, our analysis is related to studies that examine the link between incumbents’ incentives to merge and outsiders’ incentives to enter the industry (e.g., Cabral (2003), Marino and Zábojník (2006), Toxvaerd (2008), and Werden and Froeb (1998)). We contribute to this literature by examining the timing of mergers in the presence of potential entry in rising and declining industries.

The difficulty we examine is by its very nature dynamic, for the reason that a key driver of the relation between demand shocks and mergers in the model is the impact that mergers have on the timing of future entry. We rely on a continuous-time real options framework to typify the dynamics of both mergers and entry, for the reason that this approach results in time-independent equilibria and, thus, is more logically well-mannered than a finite-period dynamic game. In addition, employing a framework similar to that of other recent models of merger timing (e.g., Lambrecht (2004), Lambrecht and Myers (2007), Hackbarth and Morellec (2007), Morellec and Zhdanov (2005, 2008), Leland (2007), and Hackbarth and Miao (2007) among others) allows us to straighten out the effects of product market interaction on the timing of mergers from other, non-product-market-related effects.We donate to the empirical merger literature by showing, using parametric and semi-parametric tests, that a significant reason for the U-shaped relation between economic shocks and merger intensity, recognized in Mitchell and Mulherin (1996), Andrade and Stafford (2004), and Harford (2005), is the effect of demand shocks on firms’ incentives to merge horizontally. In addition, the evidence shows that the U-shaped relation between horizontal merger concentration and the state of industry demand is there in relatively concentrated industries, whereas it is absent in relatively competitive ones, in which strategic considerations are likely to play a smaller role. This confirmation is reliable with our model, which predicts that, ceteris paribus, horizontal mergers within oligopolistic industries are more probable to take place in times of high and low demand relative to times of intermediate demand, and that such pattern disappears within relatively competitive industries. The modelSetupAssumption 1:There are two incumbents in the industry. Each incumbent is endowed with capital, K. Moreover, entry by one firm is allowed, with an equal amount of capital, K. The firms’ production functions are of the Cobb-Douglas specification with two factors and constant returns to scale:

1 1

2 2i iq K L (1)

Where qi is the firm i’s output, and Li is the amount of labor it employs.

The cost of one unit of labor per unit of time is denoted pl. The amount of capital is fixed, hence labor is the only variable input. At any given instant, each firm can costlessly adjust its labor input to produce any output quantity. Since firms are not able to alter the level of capital, firm i’s instantaneous variable cost of producing qidt units is

2

( ) ii i l

qC q p dt

K

(2)

Note that this specification assumes away technological (production) synergies.

Assumption 2

The firms are subject to heterogenous-products Bertrand competition.

The heterogeneous-products Bertrand competition allows us to accommodate different degrees of substitutability among the rivals’ products and analyze comparative statistics with respect to the extent

of competition. The logic and results of the model, nevertheless, are robust to choosing different types of product market competition.

Assumption 3

The demand-side of the industry is characterized by a representative consumer with quadratic instantaneous utility function

2

1 1

1( ) 2

2

n n

i i i ji i j i

U q x q q q q

(3)

Where α, β, and γ are the parameters of the utility function, qi is (annualized) consumption of good i, n is the number of active firms in the industry and, thus, the number of available products, and x is the stochastic shock to the representative consumer’s utility. This specification of the utility function is typical of partial equilibrium models commonly used in the industrial organization literature. We further assume that x follows a geometric Brownian motion

t t t tdx x dt x dW

Where, Wt is a standard Wiener process on a filtered probability space (Ω, F, P).

We impose the standard conditions: α > 0 and β > γ > 0. γ > 0 implies that the goods produced are substitutes, which is reasonable for products of firms competing in the same industry. α> 0 and β > γ imply that the utility function is concave in each of its arguments. The specific functional form of the relation between the representative agent’s utility and the state of the stochastic shock, x, is made for analytical convenience. It is common in the industrial organization literature to assume that shocks to demand correspond to changes in the intercept of the demand function. Consistent with this norm, as

shown below, the term x in the linear term of the utility function translates into a linear relation

between x and the intercept of the demand function. This in turn, translates into a linear between x and firms’ instantaneous profits.

Equating the marginal utility that the representative consumer derives from consuming product i to its price and solving the resulting system of n equations in n unknowns (quantities) defines the demand function for product i as a function of i’s own price and the other products’ prices:

( )i i jj i

D p xa bp c p

(4)

Where

,( –1)

( – 2)

[ ( –1) ]( – )

[ ( –1) ( – )]

an

nb

n

cn

(5)

As mentioned above, the only benefit of merging in the base-case model is that the incumbents can coordinate their pricing strategies. The entrant benefits from the merger, due to the lower level of competition. This leads to the following intuitive results.

Lemma 1

(i) The combined instantaneous profit of the two incumbents is higher if they merge than if they stay separate, ceteris paribus.

(ii) The entrant’s instantaneous profit is higher if the two incumbents merge than it they stay separate ceteris paribus.

(iii) The incumbents’ combined instantaneous profit is higher in the case of no merger and no entry than in the case of merger and entry.

When the two incumbents merge, they charge higher prices for the reason that they internalize the cannibalization effect of raising one product’s price on the quantity sold of the other product. This benefits the entrant and increases its instantaneous profit. Combining the first and third parts of Lemma 1 results in the following set of relations for the combined instantaneous profits of the incumbents in the four possible scenarios (merger/no merger combined with entry/no entry):

(noentry,merger) (noentry, no merger) (entry, merger) (entry, no merger)inc inc inc inc (6)

This result is significant. The first and third inequalities show that given the presence/absence of the entrant in the industry, the incumbents are always better off merging (part (i) of Lemma 1). The second inequality (part (iii) of Lemma 1) is at the heart of our analysis and implies that the incumbents may be better off not merging, if by staying separate they can deter potential entry of the new firm in the industry.

Absent the threat of new entry, the optimal strategy of the incumbents is to initiate a merger attempt independent of the state of the industry demand, for the reason that it is costless to do so. The threat of potential entry makes the problem more realistic and interesting, by introducing an opportunity cost the incumbents must bear when attempting a merger. On the other hand, the entrant’s profit and thus, its decision to enter depends on whether the incumbents have merged.

Assumption 4

Incumbent firms are endowed with an option to initiate one attempt to merge with each other (combine operations). Attempting and implementing merger among incumbents entails no out-of-pocket costs.

Once initiated, the merger attempt is successful with probability p < 1.

The assumption that a merger attempt does not necessarily result in a successful merger is consistent with the empirical evidence. First, a merger attempt can be unsuccessful due to difficulties in the negotiation process. Anti-takeover provisions, such as staggered boards, limits to shareholder bylaw amendments, poison pills, golden parachutes, supermajority requirements, and state anti-takeover legislation reduce the likelihood of successful merger attempts. Boone and Mulherin (2007) report that only 27% of potential bidders that sign a confidentiality agreement and only 78% of bidders that submit a private written offer succeed in acquiring their target. Second, even when the firms involved are willing parties to the negotiation, antitrust authorities may oppose the transaction on anticompetitive grounds or provide conditional approval, requiring firms to divest the operations that pose the highest anti-competitive threat. This, of course, is an aspect that is particularly relevant in the context of our model. In Eckbo’s (1983) sample of 191 horizontal mergers that occurred between 1963 and 1978, for instance, 65 were challenged by either the Justice Department or the Federal Trade Commission.

Finally, even if the merger deal is successfully completed, there remains uncertainty about the ability of the merging firms to successfully integrate their operations. Mitchell and Lehn (1990) and Lehn and Zhao (2006) show that there are badly conceived mergers that are subsequently thwarted through divestitures, bust-up takeovers, and management turnover. There is also anecdotal evidence that suggests that post-merger successful integration turnover. There is also anecdotal evidence that suggests that post-merger successful integration is far from being certain.

The assumption that the two incumbents are endowed with only one option to initiate a merger attempt is made for analytical tractability. From the modeling perspective, allowing multiple merger attempts is equivalent to raising the probability of merger success, p. In the base model, for the reason that we assume merging entails no out of pocket costs, allowing a finite number of merger attempts would result in a series of attempts that would stop either after a successful merger or after all attempts have been exhausted. Uncertainty about the success of the merger attempt is crucial in our model. This is for the reason that absent uncertainty, entry is independent of the merger attempt and vice versa, which makes mergers optimal in all states of world.

Assumption 5

Upon successful consummation of the merger, the shareholders of each incumbent receive a 50% stake in the merged entity. We abstract from the analysis of how merger gains are allocated. All that is required for a deal to take place is that the combined value of the merging firms increases as a result of a merger. For the reason that the two incumbents are identical in all respects, we simply assume that the gains are split evenly between the merging firms to ensure that the merger is desirable for both sets of shareholders. Moreover, in our model the merger payment method (cash or stock) is irrelevant, as long as capital markets are efficient and securities are correctly priced. Therefore, we do not analyze misvaluation-driven merges, as in Rhodes-Kropf and Viswanathan (2004) or in Shleifer and Vishny (2003).

Assumption 6Entry requires the outsider to incur a fixed irreversible cost, I, to obtain capital K. This assumption precludes immediate entry for low realizations of the demand shock. Consistent with economic intuition, the 1997 Horizontal Mergers Guidelines highlight that costless entry precludes mergers motivated by the pursuit of market power. Given that strategic (market—power-related) considerations are the focus of our analysis, we assume entry is indeed costly.Assumption 7We normalize the amount of installed capital of each firm, K, the cost of labor, pl, and the coefficient on the quadratic term of the utility function, β, to one. This assumption is made for analytical convenience only. Normalizing β to one is innocuous. In addition, it is straightforward to show that the general version of the model with K and pl that are different from one produces the same conclusions as the more restrictive model we are examining here.

AnalysisWe now proceed to the formal analysis of the model. In what follows, we incorporate the derivations of the firms’ instantaneous profits under different industry structures, found in the proofs of Lemma 1, and introduce the following simplifying notation for the firms’ instantaneous profits under different scenarios:

2 2,

2 2

1 (2 – )(noentry,no merger)

2 (4 – )ne nminc incx

(7)

2, 1

(noentry,merger)2 4(2 )

ne nminc incx

(8)

2 2 2 2,

2 2 3 2

1 (4 3 – 3 ) (2 – 2 )(entry,merger)

2 4(1 ) (8 4 – 9 2 )e nminc incx

(9)

2 2,

2

1 (1 )(1 – )(entry,no merger)

2 2(2 )e nminc incx

(10)2 2

2

1 (1 )(1 – )(no merger)

2(2 3 )nment entx

(11)2 2 2

2 3 2

1 2 (2 )(1 – )(merger)

(1 )(8 4 – 9 2 )ment entx

(12)We start by establishing the optimal thresholds xe,m and xe,nm, corresponding to the cases in which the incumbents have already attempted a merger, either successfully or not.Lemma 2If the incumbents have already exercised their option to attempt a merger and have not succeeded, then the optimal entry threshold is given by

1,

1

( – )

–1e nm nment

I rx

(13)

Where β1 is the positive root of the quadratic equation

21( –1) – 0,

2r

1 2 2 2

1 1 2– –

2 2

r

(14)

If the incumbents have successfully merged, then the optimal entry threshold is

1,

1

( )

1e m ment

I rx

(15) As argued above, in order to find xl and xu we need to examine the incumbents’ and entrant’s optimization programs simultaneously. The following two results provide a set of equations that determine the solution to the entrant’s and incumbents’ optimization problems.Lemma 3If xl < x < xu then the entrant’s value is given by

1 2( )nmentV x Ax Bx (16)

Where A and B are constants to be determined below, and β2 is the negative root of the quadratic

equation

2 ( –1) 0,2

r

2

2 2 2 2

1 1 2

2 2

r

(17)The following conditions must hold at xl and xu:

1 21

(1 )m nmu u ent u ent uAx Bx p x p x I

r

(18)

1 21 11 2

1| (1 ) ,m nm

u u ent entAx B x p pr

(19)and

1 1

1 2 , ,

, ,,

|(1 )

m nment e m ent e nml l

l le m e nm

x xx xAx Bx p I p I

x r x r

(20)

Equations (18) and (20) are the value-matching conditions, which stipulate that the values of the entrant at the two merger thresholds are exactly equal to their respective expected post-merger-attempt values. (These values are the weighted averages of the values conditional on a successful and unsuccessful merger attempts.) Equation (19) is the smooth-pasting condition that ensures the optimality of the outsider’s entry decision.Lemma 3 provides us with three equations in four unknowns (two constants, A and B, and the optimal merging thresholds, xu and xl). Thus, we need additional conditions in order to solve for the optimal merging thresholds. The remaining conditions come from the optimization program of the incumbents. These conditions are derived as follows.Lemma 4If xl < x < xu then the value of each incumbent is given by

1 2

,

( ) ,ne nminc

inc

xV x Cx Dx

r

(21)Where C and D are constants to be determined together with A, B, xu, and xl. The following conditions must hold at the upper and lower merging thresholds, xl and xu:

1 2

1

,

, , , ,,

,

1(1 ) ,

ne nminc u

u u

e m ne nm e nm ne nmuinc u inc u inc inc e nm

e nm

xCx Dx

r

xp x p x x

r x

(22)

1

1 2

1

,, , ,

,,

, , ,,

,

1{

(1 ) ( ) },

ne nme m e nm ne minc i l

l l inc inc e m inc le m

ne nm ne nm ne nmlinc inc e nm inc l

e nm

x xCx Dx p n s x

r r x

xp x x

x

(23)

1

1 2

1

1

1

1,1 1 , , ,1

1 2 ,,

1, , ,1

,,

1{

( )

(1 ) ( ) }( )

ne nme m ne m ne minc l

l l inc inc e m ince m

e nm ne nm ne nmlinc inc e nm inc

e nm

xCx Dx p x

r r x

xp x

x

(24)

In (21) the term

,ne nminc x

r

refers to the present value of each incumbent’s perpetual entitlement to

instantaneous profits if the incumbents never merge and the outsider never enters. The remaining

terms, 1 2Cx Dx , account for the change in each incumbent’s value due to the incumbent’s option to

merge among themselves and for the threat of new entry.Equations (22) and (23) are the value-matching conditions for each incumbent’s optimization problem, while (24) is the smooth-pasting condition that must obtain at the lower merging threshold. The first term on the right-hand side in (22) is the post-merger value of an incumbent if the merger attempt is successful, and the second term is the value of the incumbent in case of an unsuccessful merger attempt. Note that the expression on the right hand side of (22) (unlike that of (18)) accounts for the fact that the merger attempt would actually precede entry, so the new entrant is able to postpone entry if the merger attempt is unsuccessful. On the contrary, (18) does not have the same term on the right hand side for the reason that the upper merger threshold is determined as the one that makes entry optimal even if the potential entrant cannot anticipate the result of the merger attempt.

CONGLOMERATE MERGERSAccording to the definition, a conglomerate merger is a kind of merger whereby the two companies that merge with each other are implicated in different types of businesses. The significance of the conglomerate mergers lies in the fact that they help the merging companies to get better.

Types of Conglomerate MergersThere are two major types of conglomerate mergers – the pure conglomerate merger and the mixed conglomerate merger. The pure conglomerate merger is one where the merging companies are doing businesses that are completely unconnected to each other.The mixed conglomerate mergers are ones where the companies that are merging with each other are doing so with the key reason of gaining access to a wider market and client base or for intensifying the range of products and services that are being supplied by them.

There are also a number of other subsections of conglomerate mergers like the financial conglomerates, the concentric companies, and the managerial conglomerates.

Reasons of Conglomerate MergersThere are more than a few reasons as to why a company may go for a conglomerate merger. Among the more general reasons are adding to the share of the market that is owned by the company and putting its feet into cross selling. The companies also look to add to their overall synergy and productivity by adopting the method of conglomerate mergers.

Benefits of Conglomerate MergersThere are numerous benefits of the conglomerate mergers. One of the major advantages is that conglomerate mergers help out the companies to diversify. As a result of conglomerate mergers the merging companies can also bring down the levels of their exposure to risks.

Implications of Conglomerate MergersThere are numerous implications of conglomerate mergers. It has often been seen that companies are going for conglomerate mergers in order to increase their sizes. Nevertheless, this also, at times, has unfavorable effects on the functioning of the new company. It has usually been observed that these companies are not able to perform like they used to before the merger occurred. This was obvious in the 1960s when the conglomerate mergers were the common trend. The term conglomerate mergers also entails that the two companies that are merging do not even have the same customer base as they are in completely dissimilar businesses. It has usually been seen that a lot of companies that go for conglomerate mergers are able to manage a broad diversity of activities in a particular market. For instance, these companies can carry out research activities and applied engineering processes. They are also able to add to their production as well as make stronger the marketing area that makes sure of better profitability. It has been seen from case studies that conglomerate mergers do not have an effect on the structures of the industries. Nevertheless, there might be important impact if the acquiring company happens to be a leading company of its market that is not concentrated and has a large quantity of entry barriers.

CORPORATE GOVERNANCE

Corporate governance came out as an issue of international concern and argumentative discussion in the early 1980s, through 1990s and this has continued into the twenty-first century. Corporate governance is not new; it has existed while the incorporation of business began (Vinten, 2003). The recognition of the centrality of major enterprises in allocating resources in the economy underlies contemporary debates about corporate governance. Nevertheless, ideas about the concept of corporate governance date back to 1776 (Tricker, 2000; Denise, 2001) when Smith, in The Wealth of Nations, raised the issue of a lack of incentives on the part of directors to look after other people’s money with as much care as the owners themselves would:

The directors of companies, being managers of other people’s money rather than their own, it cannot be expected that they should watch over it with the same anxious vigilance with which the partners in the private copartnery watch over their own (Smith, 1776 cited by Tricker, 2000).

Smith did not use the word corporate governance; the term only emerged in the 1980s (Tricker, 2000). Nevertheless, this remark shows that he had a sound

“Good corporate governance practices instill in companies the essential vision Processes and structures to make decisions that ensure longer-term sustainability. More than ever we need companies that can be profitable as well as achieving environmental social and economic value of society.”Rachel Kyte — Vice President Business Advisory Services IFC

understanding of the issue of corporate governance: when the owners of a corporation are dissimilar from those who manage it, incentive problems tend to occur. The nature of the debate on corporate governance is prejudiced by the way in which corporations are viewed. Clarke and Clegg (1998) challenge that the early conceptualisation of corporations tended to treat corporations as the property of equity capital providers (shareholders) for the pursuance of their economic interests. Nevertheless, a necessary characteristic of a corporation is its ability to have a split existence apart from those who own it. When a corporation has attained its own different and unique existence, the issue of control arises.Mintzberg (1984) points out that, traditionally, control of a corporation was exercised by its owners either in a straight line or through control of management. Nevertheless, when ownership and management are separated, as when ownership becomes fragmented, control of the corporation presents an important challenge. The issue of the separation of management from ownership, which results in the transfer of control of corporations from owners to professional managers (Scott, 1997), received greater emphasis following the Berle and Means article: The Modern Corporation and Private Property (Berle and Means, 1932).They observe that:

…in the modern corporation, these two attributes of ownership (control and economic rights) no longer attach to the same individual or group. The stockholder has surrendered control over his wealth. He has become a supplier of capital, and a risk taker pure and simple, while ultimate responsibility and authority of ownership is attached to the stock ownership, the other attribute is attached to corporate control. Must we not, therefore recognise that we are no longer dealing with property in the old sense? Does the traditional logic of property still apply? For the reason that an owner who also exercises control over his wealth is protected in the full receipt of the advantages derived from it, must it necessarily follow that an owner who has surrendered control of his wealth should likewise be protected to the full?

Berle and Means suggest that the notion of ownership of property when employed to corporations particularly large ones, is not simple. It is in this context that Minztberg (1984) poses the question: who should organize the corporation, and for the pursuit of what goals? Minztberg contends that as ownership of corporations became discrete, owner-control weakened and corporations came under the contained control of their managers. Stiles and Taylor (2002) point out that although corporate governance as a subject has fascinated extensive interest internationally, the nature of the debate about it is still basically shaped by the Berle and Means analysis.Causes of corporate governance challengesThe issues that have encouraged interests in the phenomenon of corporate governance, spot to particular causes of corporate governance crises. These comprise feeble legal and regulatory systems, contradictory accounting and auditing standards, and poor banking practices. Thin and poorly regulated capital markets, unproductive oversight by corporate boards of directors, and little regard for the rights of minority shareholders are also problems with respect to corporate governance (World Bank, 2000).The problem of pathetic legal and regulatory systems is usually viewed as a problem of developing countries. Developed economies tend to have developed and complicated regulatory systems, while less industrialized ones tend to display less efficient systems of law and regulations (Lin, 2000). The thinness and lack of effective stock exchange regulation may also be seen as chiefly part of the problem in developing countries. This is connected with the low level of market development in such economies (Lin, 2000; World Bank, 2000). When legal and regulatory systems are weak, the enforcement of contracts becomes difficult. For instance due to weak legal and regulatory systems, particularly the enforcement of laws in Russia and the Czech Republic, controlling shareholders were

able to draw off profits leading to a loss of investments by minority shareholders (World Bank, 2000).The application of varying accounting and auditing standards is another challenge to corporate governance (Clarke and Clegg, 1998). This problem emanates from the use of a variety of financial accounting standards by corporations whose operations span to diverse countries in the preparation and presentation of financial statements (Bradley et al., 1999). For instance, US Corporations employ an American system of GAAPs developed by FASB, whereas UK-based corporations apply a dissimilar set of accounting standards (SSAPs) developed by the Accounting Standards Board (ASB) and the Financial Reporting Council (FRC). This use of unlike accounting standards makes the evaluation of performance across companies operating internationally complicated (Bradley et al., 1999). This challenge has led to the need to correspond standards through the use of accounting standards promoted by the International Accounting Standards Board (IASB). This initiative is reflected in the current attempts to harmonize accounting standards in Eastern, Central and Southern African regions through the Eastern, Central and Southern Africa Federation of Accountants -ECSAFA (Gathinji, 2002).The poor banking practices reported by the World Bank are particularly related to the Asian crisis where banks provided credit to companies under the influence of the political elite. Either one family, or a corporation under a family’s control, generally own Asian firms. Such families have close connections with the government, and politicians, and dominate the national economy to a large extent (Hanazaki and Liu, 2003). Using these connections, corporations have been able to borrow funds from banks without the proper disclosure of the information required to enable full evaluation of company performance and establish creditworthiness (World Bank, 2000).The cases of Maxwell, BCCI, Nomura and a number of other large corporations show how the lack of effective oversight by directors can lead to corporate governance crises. This lack of effective oversight by boards of directors has resulted in boards' failures to prevent a large number of fraud cases and the subsequent collapse of corporations (Tricker, 2000; Stiles and Taylor, 2002). Mace (1971) argues that boards of directors are 'Christmas ornaments' and do not effectively control senior managers. Demb and Neubauer (1992), posit that this is paradoxical since chief executive officers exercise the power of corporations which should be the preserve of directors.Perspectives on Corporate GovernanceCorporate governance can be addressed from two broad perspectives: the liberalist and the communitarian perspectives (Bradley et al., 1999; Clarke and Clegg, 1998). The liberalist perspective views corporations as only accountable to shareholders. In this perspective, a corporation's legitimate goal is to serve the interests of those who own it i.e. the shareholders. The legitimate claims of other stakeholders are satisfied by meeting the contractual terms between them and the corporation. Within the communitarian perspective, corporations are required to be accountable to other stakeholders than the shareholders alone. In this respect, shareholders become one stakeholder group among a number of stakeholder groups. The results of these differing views are different goals for corporations, for which managers are held accountable (Bradley et al., 1999). The case of SGL, a German manufacturer of graphite and carbon which wanted to attract capital from American investors, is provided to illustrate the implications of these perspectives.

In 1992, SLG AG lost more that $71 million as part of a restructuring programme to turn the company around, top management altered the firm’s governance structure by: adopting US transparent accounting practices, establishing corporate goal of enhancing

“A well-governed company takes a longer-term view that integrates environmental and social responsibilities in analyzing risks discovering opportunities and allocating capital in the best interests of shareowners. There can be no better way to restore public confidence in both businesses and markets and build a prosperous future.”Georg Kell—Executive Director UN Global Compact

shareholder value which is a norm in the US corporate governance, listing its stock on the New York Stock exchange, and even changing the company official language to English. Top management wanted to make the company look “Less German”. By 1995, SGL had seen its revenues increase to more that $159 million dollars. Top management attributes much of the improvement to the changes in corporate governance. Thus, corporate governance provides firm-specific competitive advantages in the global capital market (Rubach and Sebora, 1998).

The case of SGL highlights the debate that has been going on for over a hundred years regarding the primacy of individuals versus the primacy of community (Bradley et al., 1999). This debate tends to be settled in each society in favour of one or the other: in Anglo-Saxon societies, the individual is assumed to have primacy and therefore his/her freedom to pursue their own goals, aimed at the maximization of their own utility, is argued for. This freedom is recognised in economic spheres and is actualised through the notion of private property. Respect for private ownership is central in the pursuit of economic activities, and in enjoying the benefits of such activities. A corporation is viewed in the same way as private property, and considered to belong to those who have put in their money as capital for the purpose of pursuing their economic interests (Scott, 1997). The goal is thus the maximisation of the wealth of its shareholders. In the above case, shareholders in the US required SGL to commit itself to this goal.The SGL case implicitly points to an alternative perspective regarding the notion of freedom one in which the primacy of individuals is recognised but within the bounds of the larger society. In this context, the argument is advanced for individuals to pursue freedoms subject to constraints imposed by society, and the economic activities pursued should serve some social goal (Fisher and Lovell, 2003). In the context of Germany, where SGL is based, corporations are expected to have goals that reflect the interests of a number of stakeholders in addition to shareholders (Weimer, 1995). This perspective is usually viewed as being broad (Clarke and Clegg, 1998).The SGL case shows that the globalisation of capital in a world that is dominated by American corporations is forcing other countries (or corporations in other counties) to re-evaluate their governance systems in an attempt to meet the demands of multinational corporations. Monks and Minnow (2002) point out that when Mercedes Benz wanted to raise capital on the New York Stock exchange, it had to adapt to the US accounting practices to facilitate investors’ evaluation of the company. The American system of corporate governance requires corporations to apply accounting practices that align financial reporting with the interests of parties outside the corporation with respect to the provision of information these parties require. The accounts are required to present a “true and fair” view of the business in the American system, as opposed to the “true and correct” requirement of the company law in Germany (Kendal and Sheridan, 1991). These authors point out that such a system helps parties outside the firm, particularly shareholders, to evaluate the corporation’s performance to assess the return on investment. In the German communitarian perspective, corporations are required to comply with detailed civil laws governing accounting.The two perspectives are of great importance in the discussions on corporate governance. Corporate governance practices are expressions of these perspectives, and are embedded in the problem of social conflict (Roe, 2003). Corporate governance evolves to address incentive problems brought about by the

Why is corporate governance significant?

Corporate governance refers to the way that Boards oversee the running of a company by its managers, and how Board members are held accountable to shareowners and the company. This has implications for company behavior not only to shareowners but also to employees, customers, those financing the company, and other stakeholders, including the communities in which the business operates.

Research shows that responsible management of environmental, social and governance issues creates a business ethos and environment that builds both a company’s integrity within society and the trust of its shareowners.

“Good corporate governance is the glue that holds together responsible business practices, which ensures positive workplace management, marketplace responsibility, environmental stewardship, community engagement, and sustained financial performance. This is even more true now as we work worldwide to restore confidence and promote economic growth.”Thierry Buchs— Head Private Sector Development Division of Switzerland’s State Secretariat for Economic Affairs (SECO)

separation of ownership from management (Tricker, 1994). The two perspectives shape the manner in which corporate governance develops. The Anglo-Saxon perspective suggests a conflict perspective, and thus corporate governance evolves to protect the interests of shareholders (Albert, 1993 cited by Macquand, 1993). In other countries, for example Japan and Germany, corporate governance is viewed as enhancing performance for the long-term survival of the corporations and for the benefit of multiple stakeholders (Sebora and Rubach, 1998; Weimer, 1995; Maassen, 1999; Letza et al., 2002). The distinct perspectives indicate that understanding the prevailing perspective in a particular context is of paramount importance in understanding corporate governance in that context.Corporate Governance in the Context of TanzaniaThe central planning system for economic coordination in Tanzania, and the ownership of corporations by the state, implies that the important experiences of corporate governance will be related to stateowned corporations. Between 1967 and 1992, state-owned corporations were the most common type of large corporations found in Tanzania. In these corporations, corruption, (embezzlement and nepotism) managerial incompetence, political interference and government subsidisation of failing corporations were the predominant characteristics of corporate governance. Corporations were shielded from the discipline of the market (Bagachwa, 1992; Kihiyo, 2002). Control and accountability became the prime casualty within these corporations. Bagachwa et al. (1992) point out that the lack of accountability and effective control of these corporations left the managers with unfettered powers. They attribute these problems to the ambiguous property rights in the state-owned corporations.The problems reported by Bagachwa et al. apply to a large extent to other countries in sub-Saharan Africa. Etukudo (1999) reports the paucity of corporate governance in sub-Saharan Africa arising from the ambiguous relationship between the state, as the owner of the corporations, the boards of directors and senior management. The paucity of corporate governance in state-owned corporations in Tanzania has resulted in dismal performance and the failure of these corporations (Wangwe, 1992). One result of the poor performance of these corporations has been their inability to provide the necessary “push” for the attainment of social and economic development as envisaged by the post-independence government (URT, 1999).The system of central planning, including the state ownership of corporations, is being reformed through a series of market-promoting schemes. This process formally started in 1986 following an agreement between the Government of Tanzania and multinational financial institutions – the IMF and the World Bank - in 1986 (Mukangara, 1993; World Bank, 2002). The reforms included adoption of competition friendly policies and the transfer of ownership of state assets/corporations to private shareholders. There had been earlier minor reforms towards a market orientation, e.g. the National Economic Survival Programme (NESP) of 1981-1982. Nevertheless, these were largely unsuccessful, and this justified the need for the more comprehensive reforms that began in 1986 (Bagachwa et al., 1992). Following reforms, a number of corporations have been privatised. Some of the privatised corporations have shown important improvements in their performance (URT, 2003). Indeed, privatisation has been viewed as a solution to the problem of governance (Wangwe, 1992).Developments in the African region and worldwide have also created the need to develop an understanding of corporate governance practices in Tanzania. As a member of the African Union (AU), Tanzania is required to join the comity of other African countries in improving corporate governance practices. The Africa Union has developed a development vision in which member countries are individually required to implement initiatives to improve corporate governance practices within them. This vision, called New Partnership and Development (NEPAD), recognises corporate governance as one of the key issues that need to be addressed to achieve social and economic development on the African continent. The vision underscores that issues of poverty alleviation are best addressed through wealth creation, in which corporate governance plays a key role since it improves efficiency in the allocation of

resources. Thus, NEPAD draws a direct link between corporate governance and wealth creation (Gathinji, 2002).Away from the African region, Tanzania is also a member of the British Commonwealth. The member countries of the Commonwealth have agreed to undertake measures to perk up corporate governance practices (CACG, 1999). CACG points out that corporate governance is significant in improving the competitiveness of member states in attracting capital and in enhancing the performance of corporations.The Overview of the ContextTanzania is a developing country situated in the Eastern African region, positioned around 5° South of the Equator and 45° East, and covers 945,087 sq km. It is bordered to the North and North East by Uganda and Kenya and to the North West and West by Rwanda, Burundi and the Democratic Republic of Congo. To the South and Southwest lie Mozambique and Zambia respectively, and the Indian Ocean lies to the East.Tanzania is made up of Tanganyika and Zanzibar. Tanganyika regained its independence in 1961 from the British, and became a republic in 1962. Zanzibar regained its independence from the Arab Sultanate in 1964. The two joined to form the current union – called the United Republic of Tanzania (URT) - in April 1964. Tanzania was a single party state from 1965 to 1991. In 1992 a multiparty political system was re-introduced; and twelve political parties had obtained their registration by 1993 (Muya, 1998). Chama Cha Mapinduzi (CCM) has since been the dominant political party. Elections for both the presidency and parliament are held every five years. Tanzania is viewed as having been one of the most politically stable countries in Africa since independence. The transition towards a multiparty political system has also been peaceful. This stable environment is viewed as providing a base for rapid economic growth (URT, 1999)Dar es Salaam constitutes the operating de facto capital of Tanzania. The designated de júre capital is Dodoma. The decision to move the capital to Dodoma was made in 1973. Thirty years on, the government continues to insist on moving to Dodoma. Nevertheless, only four out of the current thirteen ministries have moved to, and currently operate from, Dodoma. The rhetoric of moving the capital to Dodoma continues, although with much less vigour than in earlier times. For example, in July 2003, the President reminded ministers to prepare themselves for the move to Dodoma, insisting that the decision to move to Dodoma was irreversible. This illustrates one of the gaps between ideology and practice in the country.Tanzania, in terms of economic potential, is endowed with a rich natural resource base and easy access for international trade. 46% of its land is suitable for agriculture (with only 6.7% of it being cultivated), it also has a large hydropower potential, a wide range of mineral deposits including gold, diamonds, tin, iron, uranium, phosphate, gemstones, and nickel, and also natural gas. Other resources include exotic varieties of wildlife and a number of tourist attractions (World Bank, 2002). Despite this potential and rich resource endowment, Tanzania remains one of the least developed countries in the world; poverty remains pervasive and deep. As illustrations, about half of the Tanzanian citizens are poor, 32 percent illiterate and the infant mortality rate stands at 99 per 1000 live births (World Bank, 2002). This suggests that the resources are not being sufficiently utilised to bring about social and economic development. The appalling level of poverty forms the basis for the current strategy for poverty eradication (Chachage, 2003). The economy experienced a severe crisis in the 1980s. Nevertheless, recently, changes have been taking place with positive growth being registered. In comparative terms, the economy of Tanzania has been showing positive growth since mid-1990s. Table 6.2 provides the output growth rates for Tanzania, Africa, developing and developed countries over recent years.

Table 6.2: Comparative Output Growth Rates: 1998-2003Country/Region 1998 1999 2000 2001 2002 2003Tanzania 4.0% 4.7% 4.9% 5.9% 6.2% 5.6%

Africa 3.3% 2.3% 3.0% 3.6% 3.4% 4.1%Developing Countries 3.0% 3.8% 5.8% 3.9% 4.6% 6.1%Developed Countries 2.0% 3.4% 5.8% 0.8% 1.8% 2.1%The World 2.0% 3.5% 4.8% 2.3% 3.0% 3.9%

Table 6.2 shows that the economy has been growing over the last six years; with annual growth rates of more than 4%. This growth is greater than that for the whole of Africa, and also that of both developing and developed countries as well as the world's overall performance. The strong economic growth rates point to increasing efficiency in the utilisation of available in the economy.Tanzania remains largely an agrarian economy. The rural population, about 76% of the total population, relies on agriculture for its livelihood, and forms an significant source of foodstuffs for the urban population (World Bank, 2002). Agriculture has been contributing about 48% to the GDP over the last ten years. This sector is linked to a large number of other sectors of the economy particularly industry and commerce. It is for this reason that the belief is held that the development of agriculture will result in rapid economic development in Tanzania (TANU, 1967; Nyerere, 1968, World Bank, 2002).Other significant sectors of the economy include trade, hotels and restaurants; financial business services; and manufacturing. Trade, hotels and restaurants (combined) form the second most significant sector of the economy, and has been contributing about 16% of GDP over the last ten years. This sector is gradually growing reflecting recent developments linked to the adoption of market reforms that began in 1986. The reforms have led to increased trade and growth in the tourism industry. For example, in 1998, tourism contributed 7.6% to the GDP up from a paltry 1.5% in the early 1990s. This sector has recently been growing at an annual rate of 22% (World Bank, 2002).The financial services sector is the third largest contributor to the economy of Tanzania. This sector has had a stable contribution to the GDP of about 10% over the last ten years. The manufacturing sector ranks fourth in terms of contribution to GDP, with a consistent contribution of about 8% between 1992 and 2002. The public sector, transport and communication, construction, mining and quarrying, electricity and water are other significant sectors of the economy. With respect to the phenomenon of corporate governance, the analysis of relevant context is considered in this research to begin with the colonisation of the country since this marked the beginning of corporate development leading to current practices.From a historical perspective, the abolition of the slave trade in the later part of the nineteenth century (around the 1870s) saw an increased interest by the Western powers seeking to establish spheres of influence on the African continent and in other parts of the world. In the context of Tanzania, the first colonial experience began with the German colonisation of Tanganyika following the outcome of the Berlin Conference of 1884-1885. German rule lasted from 1884 to 1919. With respect to corporate governance, laws constitute the main way in which colonial experience influences prevailing corporate governance practices in colonised countries (Djankov et al., 2003).Nevertheless, the influence of German law on corporate governance is nowadays nonexistent in Tanzania. This is for the reason that World War Two led to Tanzania becoming a protectorate of the British following the endorsement of the League of Nations (now called the United Nations). This was the point at which British colonial influence penetrated Tanzania, the legacies of which are still present. The most important legacy of this linked colonial past, with respect to corporate governance, is the British legal system which has provided the framework for corporate governance in Tanzania to the present time.In addition to the legal system, systems of economic coordination are significant determinants of corporate governance practices (Gilpin, 2001). They reflect assumptions about social life in different societies, and how economic activities are coordinated. Corporate governance practices are embedded in systems of economic coordination. The cultural, legal and institutional aspects of different societies influence corporate governance (Tricker, 1994; Sebora and Rubach, 1998). An understanding of the

systems of economic coordination is helpful for placing corporate governance practices in their appropriate context.Implications for Corporate GovernanceCorporate governance under central-coordinationThe two systems of economic coordination within which business has been operating imply that corporate governance can be discussed under the two forms of economic coordination. The centrally-controlled system of coordination had implications for corporate governance practices with respect to state-owned corporations. This was the most significant form of corporation during the Ujamaa and Kujitegemea era. Figure 6.3 indicates the system of governance that evolved with respect to state-owned corporations under the centrally-controlled system. The figure shows three different levels of governance: the people, the ruling party, and the government and parliament (Mwapachu, 1983). The way in which state-owned corporations were to be governed was stated in the Public Corporations Act and by individual Acts that established specific corporations. In this regard, the Public Corporations Act of 1969, amended in 1976, applied until 1992, when the current Public Corporations Act was enacted.

Figure 6.3: Corporate governance of state-owned corporationsSource: Based on Mwapachu, 1983; Public Corporations Act 1969 and 1992

(i) People and party controlTANU aimed at placing the ownership of enterprises in the hands of the Tanzanian people through their government and that they would exercise ultimate control of corporations both directly and indirectly. Direct governance involved the use of social pressure brought to bear on the managers of the corporations. Social pressure is recognised as one of the mechanisms by which corporations can be controlled (Demb and Neubaeuer, 1992). For example, complaints about the conduct of managers reported in the media constituted one of the deterrents to the abuse of the power entrusted to managers (Mwapachu, 1983). People could also send reports to the Permanent Commission of Inquiry which was established to address the problem of abuse of power and corruption practiced by civil servants including the managers of state-owned corporations.Indirect control by the people was exercised through the ruling party, TANU and later CCM. Control through the party was more prevalent during the early days following the Arusha Declaration and was mainly exercised through party directives to the government (Mwapachu, 1983). The notion of party supremacy is applied to express the extent of control that the ruling party exercised (Kiondo, 1993).

Party control was also exercised through the appointment of party representatives9 who were placed in corporate offices. These officers reported directly to the national party organs on the performance of the managers of the corporations.They could send confidential reports to the national party organs such as the National Executive Committee (NEC). This way of reporting gave representatives important influence on the management of state-owned corporations; managers feared the disciplinary action that could be taken against them if they were reported to the NEC.Parliamentary controlParliamentary control was (and still is) exercised at two stages. At the stage of debating and approving the bill, and during the life of a corporation. At the establishment stage, the government, through a sector ministry under which the corporation would be placed, is required to satisfy the parliament of the relevance of the corporation to be established. This was to ensure that the corporation being proposed fitted the national interests in terms of development plans, and that public funds were directed into significant areas. There were problems, nevertheless, in relation to those state-owned corporations which were established by presidential order10. In this event, the parliament has no way of vetting the usefulness of a corporation.Parliamentary deliberations about the performance of corporations were (and still are) useful ways of exercising control over state-owned corporations. In 1978, the Parliament established a Parastatal Organisations’ Accounts Committee which served as an significant control mechanism. This committee verified whether funds utilised were legally appropriated and whether funds have been expended on approved services and projects. The Committee also looked into budget overruns, and the reasons for them, and checked for wasteful utilisation of public funds. In discharging its responsibilities, the committee employed the services of the controller and auditor general (CAG) and those of the Tanzania Audit Corporation (Mwapachu, 1983).(ii) Governmental controlPresidential controlPresidential control has been both formal and informal. Formal control applied to specific corporations that, by the Act establishing them, had to report directly to the President. These include the Capital Development Authority (CDA) and the Rufiji Basin Authority (RUBADA). Presidential control was exerted over all other corporations through the appointment of the chief executives/chairmen of the holding corporations. According to the 1992 Act, the President appoints the chairmen of those corporations in which the government is the only shareholder.According to the 1969 and 1992 public corporation acts, the appointments have to be based on recommendations by the sector ministries. Further presidential control, over state corporations, was exercised informally through consultation and planned visits by the President to these corporations. The President also exercised control through his powers to require corporations solely owned by the government to keep records as the President directed. The borrowing powers of such corporations are also vested in the president.Mwapachu (1983) reports that during the customary annual consultation sessions between the president and the heads of public corporations, particularly holding corporations, the managers of the corporations were required to present their latest briefs on their performance and problems. These occasions also served as forums for managers to present their problems and receive support in areas such as foreign exchange. The occasional visits provided a means for interaction between the president and directors and senior management of public corporations. During these visits the President is

9 Party representatives were individuals appointed by the party (i.e., CCM) to promote the party’s policies at places of work.10 In Tanzania, corporations could be established either by Act of Parliament or by presidential order.

reported to have made orders on steps that needed to be taken. The President also had a fully-fledged advisory economic unit which is regarded as having been useful in gathering information about the performance of corporations and advised the President accordingly. This provided the President with further opportunities and means for exercising control over corporations.Ministerial controlMinisterial11 control concerns sectoral control over state-owned corporations. The 1969 Presidential Circular No 2 stipulates that each state corporation should be responsible to one sectoral ministry (Mwapachu, 1983). Sector ministers exercise control over corporations through: ministerial directives, ministerial appointment of directors of parastatals, access to information on demand, control of borrowings, budget approval, control over investments and audit scrutiny. Ministers are empowered by law to provide directives of both general and specific character to state corporations. The ministers are responsible for ensuring that the corporations operate in the national interests. Mwapachu (1983) contends that the lack of a clear definition of the national interests left the issue open to interpretation by the responsible minister. This control could be applied in ways that could frustrate management and lead to the poor performance of the corporations. The 1992 Act still requires the minister of the parent ministry to give corporations directives. Section 6 of the Act states:

“Where the government is the sole shareholder, the minister responsible for the parent ministry may in writing under his hand give the board of directors of the Public Corporation direction of a general or specific character as to the performance of their functions”

The 1992 Act also gives the minister of the parent ministry the power to appoint directors for corporations in which the government is the sole shareholder. With this Act, every proposed appointment has to be forwarded to the treasury registrar. The Act specifies that where the government is not the sole shareholder, instruments under which such corporations are established govern the appointment of directors. When the directors are appointees of the minister, the boards provide the link between the minister and the management of the corporation.The responsibilities of boards of directors of holding corporations include the development of mechanisms for sending information to the ministers of the parent ministries and reporting about the property of the corporation as well as its operations. Directors were also required to submit, to the minister, accounts (including annual reports) and any other reports the minister demanded. A copy of the audited annual accounts had also to be made available to the sector ministry.As part of ministerial control over investments, most of the parastatals could not (and still cannot) take out loans or overdrafts without prior approval of the sector minister. In some cases, both the sector minister and the Minister for Finance must approve any borrowings. For example, Tanzania Railways Corporation (TRC) cannot undertake borrowing without the prior approval of the sector ministry as well as of the Minister for Finance. The 1981 laws covering miscellaneous amendments introduced this double control for all state-owned enterprises.In relation to budget approval, budgets prepared and approved by boards of directors must be in the form and detail specified by the parent ministry; final approval of the budget rests with the minister. Investments of funds by public enterprises are also subject to approval by the parent ministries. Nevertheless, there were variations with respect to investments. For example, in the case of Tanzania Posts and Telecommunications Corporation12 there was no such requirement, whereas such a requirement exists for the Board of Internal Trade, National Lotteries, and the National Milling Corporation (Mwapachu, 1983).11 In Tanzania, the President appoints all the ministers with the prime minister only requiring parliamentary approval. Ministerial control can therefore be considered as indirect presidential control.12 This company has been split into three separate companies for Postal Services, Telecommunications and the Postal Bank.

With respect to holding-subsidiary relationships, holding corporations act on behalf of the parent ministry. The performance of the holding corporations reflect the performance of their subsidiaries. Prior to 1992, holding corporations, operating under parent ministries, were treated as custodians of government equity in the operating companies (subsidiaries and associates13). They were charged with the responsibility of overseeing the individual companies. These holding corporations included the National Development Corporation (NDC), the National Textile Corporation (NATCO), the National Agricultural and Food Corporation (NAFCO), the State Mining Corporation (STAMICO) and the Livestock Development Authority (LDA) (Mwapachu, 1983). In relation to their overseeing function, the holding companies were required to carry out a number of activities: planning, promoting, organizing and integrating subsidiaries into a specific sector of the economy, and acting as entrepreneurs on behalf of government. Thus, holding corporations were required to identify new areas of economic investment and to participate in carrying out such investments. They were also required to coordinate the activities of subsidiary companies in terms of defining the economic and financial objectives, guide and direct their performance for optimal capacity utilization, and review and control performance using budgetary controls. Holding corporations were also required to appoint directors and the top management of subsidiaries, and provide research, marketing and development services, recruitment and training services at the management levels. Auditing, financial management services and the determination of salary structures, in collaboration with relevant Government authorities particularly SCOPO, were other duties of holding corporations with respect to subsidiaries.Central controlCentral control refers to the exercise of control by the treasury registrar. This is part of the formal control exercised by the government. Formal mechanisms are ones that have been stipulated in the law that established state-owned corporations, i.e. the 1969 Act and relevant regulations. The post of treasury registrar was first established by the colonial government for the purpose of controlling and allocating the resources of that government. The post independence government adopted the same model14. This office, in the Ministry of Finance, is there to acquire, hold and manage investments of the government.Under the control of the treasury registrar various mechanisms were instituted. These include the establishment of committees designed to enable the treasury registrar to exercise control. For example, the treasury registrar established a special committee on parastatal management agreements for the purpose of evaluating proposals for management agreements with state-owned corporations. It also established a finance and credit plan council, under the direction of the Ministry of Finance. This was to ensure that financial resources, accruing to the government and its institutions, were put to priority use. It also ensured the involvement of financial institutions in financial management and credit allocation. Central controls also included a control function for banks and financial institutions by the Bank of Tanzania (BOT) through the Government Loans, Guarantee and Grants Acts of 1975.The Tanzania Audit Corporation (TAC), the National Price Commission (NPC) and the Standing Committee on Parastatal Organisations (SCOPO) were also involved in the control of state-owned corporations. The control role of the TAC related to powers and obligations to provide audit and auxiliary services including advisory and accounting services. NPC exerted control on enterprises through its powers to fix and control prices of products and goods both produced within the country and imported.The argument behind the establishment of the NPC was to ensure that corporations realised profits by increasing efficiency rather than through exploitative pricing. In that way, NPC would protect consumers

13 A subsidiary company is one a holding corporation has a controlling stake (above 50% equity) while an associate is one in which it has shares but a controlling stake.14 Treasury Registrar Act of 1959, 1980.

from unreasonable pricing as well as limit the competition that existed between private and public enterprises. While this objective was plausible it was not achieved. Wangwe (1992) points out that price controls did not offer incentives for efficiency for the reason that inefficient managers could price their goods/services at the stipulated prices where they could have been lower if efficiency was effected.In the early days of the Arusha Declaration, the remunerations of employees of state corporations were set by SCOPO. SCOPO was established immediately after the nationalisation policy of the Arusha Declaration to: formulate training programmes for employees of state-owned corporations, review salary and benefits of parastatals to bring them into line with those of other public servants with comparable skills, review and approve the conditions of employees seconded from public service to state-owned corporations, approve transfers between parastatals and the Ministry of Manpower Development and to review localisation arrangements with state corporations to bring them in line with the government policy requiring all posts be filled by competent citizens. According to the 1992 Act, all the activities that SCOPO once performed have become the domain of boards of directors of these corporations. Figure 6.2 depicts the governance system of state-owned corporations.Reflection on the Corporate Governance SystemThe corporate governance of state-owned corporations reflects the role of the state in coordinating economic activities. The former corporate governance system of stateowned corporations was a complex system which faced a number of problems with respect to control and accountability. At the level of the people and the party, it can be argued that, when ownership is so fragmented, free rider problems15 occur which lead to an absence of monitoring (Hart, 1995). It cannot, be expected that the people would have the motivation and adequate information about the performance of corporations to exert sufficient pressure on management.At the presidential/ministerial level, there are also problems of information and incentives. For example, with regard to holding corporations, the President appointed the CEO/chairmen while the sector minister appointed other directors. The directors could not exercise control over the decisions of the CEO/chairman. The minister under whom the corporation was placed could not discipline the CEO if they wanted to, except by making recommendations to the appointing authority. Kihiyo (2002) raises the problem of information not reaching the President in time for action. Nevertheless, it can be argued that information might not reach him at all and, even if it did, it could easily be distorted in line with the interests of the people in the chain of command in the matter being reported. The system was therefore informationally deficient, permitting problems to go unchecked.In addition to the problems relating to the structure of the corporate governance system, state-owned corporations operated without clear objectives. Social and economic objectives were imposed, instructions to management from government ministries conflicted those from other government agents, political interference was rampant and the role of the state as shareholder and regulator of business was ambiguous (Wangwe, 1992). The result was a lack of specific criteria for assessing performance. The system was also bureaucratic in that decisions had to follow a number of steps before action could be taken. This slow pace of decision-making translated into a lack of the flexibility needed to take advantage of opportunities.The ineffectiveness of the corporate governance of state-owned corporations contributed to their poor performance. Gregory and Simms (in Monks and Minow, 2002) suggest that the effectiveness of corporate governance will be reflected in a firm’s performance16. A study conducted by the Ministry of Finance in 1974 of 24 parastatals showed that, by the end of 1973, they had accumulated TAS 178 million (equivalent to US$ 35.6 million at that time) in losses which accounted for 91% of the total capital allocated to these parastatals. By the late 1980s, the public corporations were regarded as such a burden on the state that they had to be divested. These governance problems resulted from a system that made managers accountable to those who appointed them rather than to the people; these “management problems” point to the lack of control and accountability (Chachage, 2003).

15 Free rider problems refer to owners of a corporation who individually hold only a small number of shares and who thus individually lack an incentive to monitor management and take corrective action.16 There are no clear measures to establish the link between performance and corporate governance. Nevertheless, it is believed that corporate governance has an important influence on the performance of corporations.

Corporate Governance under the Market-based SystemAt the level of the corporation, the ongoing reforms have implications for the form of corporate governance that will evolve. High levels of poverty characterise the social context in which economic reforms are taking place in Tanzania (Mtatifikolo, 1992; World Bank, 2002). This limits the extent to which the majority of local people can effectively participate in the process of reforms, with particular reference to the privatisation of the state-owned corporations. For the reason that of the shortage of local capital, privatisation through strategic investors17 has become an significant method for privatising state-owned corporations.Privatisation through strategic investors has implications for corporate governance: it provides an opportunity for a small number of individuals and corporations to acquire significant holdings in the privatised corporations and hence result in a concentration of ownership in the corporations. This issue was of concern during the early years of privatisation. Ngemera (1993) cautioned that if privatization was not carefully handled, it could end up by creating an economy which was either foreign dominated, or locally dominated by a small group of people. Notwithstanding these warnings, the majority of the privatised corporations have single controlling shareholders18. Although poverty is generally considered to have reduced the ability of indigenous people to acquire the privatised corporations, leading to a greater reliance on strategic investors, other reasons for the reliance on strategic investors have been suggested. Chachage (2003) observes that:

“…what we are seeing is a situation whereby stupendous wealth is accumulated by a tiny fraction of the population through exploitation and pillage of human, mineral, and natural resources during mounting poverty, destitution and structural adjustment”

The above comment implies that a number of individuals have exploited the reform process for private gain. The process of privatisation has been criticised by members of the public, including members of parliament, for being indifferent to the interests of local people (Simba, 2003)19. Table 6.2, provides a list of some of the companies that have been privatised, and the extent of ownership concentration in such companies. The majority of the corporations in Table 6.4 have been referred to as major corporate taxpayers by the Tanzania Revenue Authority (TRA)20 which points to their importance in the economy particularly with respect to their contribution to the government revenues (URT, 2003).

Table 6.4: Structure of ownership in successful privatized firms in TanzaniaCompany name Ownership in % Company name Ownership in %

Private shareholders

Govt Other shareholders*

Private shareholders

Govt Other shareholders

TCC* 75 2.5 19.5 Tanzania Cables 51 49 -TBL* 52.8 4 43.2 Kilombero Sugar Estate 75 25 -ABB Tanalec Ltd 70 30 - Tanzania Portland 55 45 -Blankets and Textile 100 0 - Mbeya Cement 100 0 -MKONO (formerly called HANDICO)

100 0 - TANESCO Wood Plant 100 0 -

Morgoro Canvas 100 0 - NABICO 100 0 -DAHCO 51 0 49 Tanzania Pharmaceutical 60 40 -Mtibwa Sugar Estate 95 5 - Sungra Textiles 100 0 -Maponde Tea Factory

100 0 - NBC 55 30 15

Source: The Economic Survey, 2001 and annual reports+ Listed on the DSE and with the largest shareholder being foreign-based.* These include individuals, other corporations and institutional investors.

17 A strategic investor is one who acquires an important number of shares in a corporation in order to become either a controlling or a important shareholder entitling him to important control rights.18 The PSRC annual reports for 2002/2003 indicate that the majority of privatized corporations have controlling shareholders.19 The rise of the indigenization debate in Tanzania is reflective of the indifference by the government to local interests. The whole privatization process in Tanzania has been criticized for a number of reasons.20 TRA has established a special unit that deals exclusively with these major corporate taxpayers.

The legal Framework for Corporate GovernanceThe legal framework for privately-owned corporations in Tanzania is linked to the colonial past. Djankov et al. (2003) point out that when European powers conquered and colonised other nations, they brought with them a large number of their political, legal and regulatory institutions, and most significantly their laws. England transplanted its laws to the United States, Canada, Australia, and other members of the Commonwealth including Tanzania. In this respect, the company ordinance of 1932, whose origin is in mid-1800s Great Britain, is relevant in Tanzania. For this reason, the corporate governance framework in Tanzania has some semblance to that of the UK.Essential elements of the legal framework include the incorporation and governance of corporations. The model that applies in Tanzania was developed in England in the 17th century and was transplanted to Tanzania through colonial channels.In England, an 1844 Act laid down the mode of a company in which people who subscribed their names to a memorandum of association became its shareholders (Tricker, 2000). The concept of limited liability for members was introduced in 1854. In general, the underlying principle of incorporation has not changed. This principle is reflected in the Company Ordinance (Cap. 212) as well as in the Company Act of 2002 which repeals the Company Ordinance when it comes into operation. This law reflects the English individualism (at the onset of enlightenment) of the 13th century characterised by the market, urbanizing society and eventually industrialization (Tricker, 2000). Ownership in a company provided shareholders with property rights which permitted them to participate or be represented in the company decision-making organs such the board of directors and annual general meetings. Company law, Cap. 212 section 26, states:

(1) The subscribers to the memorandum of a company shall be deemed to become members of the company and on its registration shall be entered in its register of members

(2) Every other person who agrees to become a member of a company and whose name is entered in its register of members, shall be a member of the company.

The Company Act, 200221 has not departed from Cap. 212 importantly. Section 24 (1) and (2) of the Act defines members of the company as those who have subscribed to the memorandum of the company; sec. 133 (1) (a-f) describes the meetings of members where disclosure is executed by directors through financial reporting and the directors’ report as means of accountability. Founders of a company prepare MEMARTS which detail the way in which the company is governed internally, and the way in which it relates to the outside world. The MEMARTS have to be drawn up within the framework of company law and to suit the shareholders of a company. Other relationships between the company and various other constituencies e.g. employees and managers are regulated by employment contracts and other relevant contracts. This approach to corporate governance has been referred to as the contractarian approach (Bradley et al., 1999).Figure 6.3 depicts the model of corporate governance that applies in Tanzania. It involves the relationship between shareholders as providers of capital and directors and management as agents of shareholders. Table A of Cap. 212 requires companies to be managed by directors appointed by members. The ordinance specifies a number of issues with regard to the power of directors and members (shareholders) in the governance of the companies. In terms of Figure 6.3, the model that applies in Tanzania, as provided in Figure 6.4, falls both within the classical and challenge perspectives of economic coordination. Both these perspectives respect the freedom of the individual, including their property.

Figure 6.4: Relationship between shareholders, directors and top management

21 The new law has been passed by Parliament and accented to by the President of Tanzania. Nevertheless, its operationalization awaits regulations from the Minister for Trade and Industry.

Source: Clarke and Clegg, 1998The Company Act 2002 requires directors to act in good faith and in the best interests of the company. Cap. 212 does not specify this role. Section 182 of the Act states:

“Subject to this section, a director of a company, when exercising powers or performing duties, must act honestly and in good faith and in what the director believes to be in the best interests of the company” [Company Act, 2002, sec 182 (1)]

The new Company Act also expands the scope of the accountability of directors to include their accountability to employees and assigns equal importance to the interests of both parties. Section 183 of the Act states:

(1) “The matters to which the directors of the company are to have regard in the performance of their functions include, in addition to the interests of members, the interests of employees”.

(2) “The duty imposed by this section on the directors is owed by them to the company (and the company alone) and is enforceable in the same way as any other fiduciary duty owed to the company by its directors”

Consistent with vision 2025, the establishment of the Business Registration and Licensing Agency (BRELA) under the Government Executive Agency Act No. 30 of 1997 is an attempt to promote the private sector and entrepreneurship in the country in line with the new philosophy of a market-based system. BRELA is a semi-autonomous agency, under the Ministry of Industries and Trade, charged with the responsibility of facilitating and regulating business activities in the country. The responsibilities of the agency include: registration of both local and foreign companies; registration of business names; registration of trade and service marks; granting of patents; overseeing copyrights and neighbouring rights administration in Tanzania; issuing business and industrial licenses.The four forms of companies that are recognised under Tanzanian company law are private companies, public companies, foreign companies and state-owned corporations. Private companies are defined as those that restrict the right to transfer shares, limit the number of its members to fifty and prohibit inviting the public to subscribe for any shares or debentures in the company (Cap. 212 sec 27 [1(a-c)]. The shares of private companies are not tradable on the stock exchange. Public companies are those without a maximum limit on the number of shareholders but required to have a minimum of seven shareholders to be allowed to operate. These companies can be listed on the stock exchange, as their shares are freely transferable. In this respect, they are required to issue a prospectus, which is an invitation to the public to subscribe to shares on issue. It is these types of corporations that are listed on the Dar es Salaam stock exchange.The third category of companies recognised under BRELA is foreign companies, these are companies operating in Tanzania as branches of companies incorporated outside Tanzania. They are not viewed as Tanzanian companies. The final category of company is a form of private company in which the government holds more than 50% of the voting shares. The Corporations Act of 1992 applies to these corporations. This form of corporation was the most significant between 1967 and 1992 but is currently diminishing as privatisation continues.

CONCLUSIONS

Tanzania is a developing economy with agriculture its mainstay. Other sectors of the economy are gradually beginning to make a important contribution to the economy. Two systems of economic coordination have existed in this country over the past forty-three years of independence: a centrally-coordinated system and then a market-based system. The developments in 1967 led to the adoption of policies that sought to encourage the evolution of an egalitarian society based on state-ownership of the major means of production including corporations.Developments in the late-1980s have changed the course of events. The economic reforms that were introduced in 1986 have the objectives of promoting and encouraging competition, and reducing the role of the state in economic activities giving individualism more room to thrive. This type of economic coordination is advocated in classical/neoclassical liberalism. The implications of such change for corporate governance is the evolution of large shareholders. The majority of privatised corporations have controlling shareholders. The company ordinances (Cap. 212), currently provide the framework in which corporations are governed. This framework promotes a liberal, shareholder oriented system of corporate governance. This is for the reason that the framework recognizes shareholders only in the key decision-making organs of the corporation: annual general meetings and boards of directors.Mergers and takeovers have introduced a new dynamism into the U.S. economy, producing gains from increasing incentives for efficiency improvements. Management is challenged to demonstrate continuing value increasing contributions to shareholders. But many examples of mistakes and excesses can also be cited.

The merger movement at the turn of the century was followed by economic development and growth in the U.S. M&As in the 1980s have also been associated with a long period of economic expansion. Obviously M&As have not been the only significant economic factor operating. But neither have they been a major negative or destructive force to date.

REVIEW QUESTIONS1. Discuss the current issues in corporate finance of Tanzania.2. What is pension fund management? How it is beneficial both to employees and enterprises?3. State the differences between defined contribution pension funds and defined benefit pension

funds.4. What is leveraged buyout?5. State the reasons for which the listed firms go private.6. How the executive compensation plans are beneficial?7. How will you justify the corporate restructuring and control?8. What are the benefits of mergers and acquisitions?9. State the difference between split-off and spin-off with suitable examples.10. What do you mean by conglomerate mergers?

CASE STUDYINTEGRITY ENVIRONMENT AND INVESTMENT PROMOTION

“The Case of Tanzania”INTRODUCTION

An environment of integrity is essential for private sector development, promotion of foreign investments and economic growth. Its requirements include building an anticorruption environment, provision of high quality government services, predictability and consistency of government procedures and regulations, and overall transparency of the system. Increasingly, many African countries are now realising that capable and responsible governments are a prerequisite for development.All over, initiatives are underway to reduce bureaucratic delays and uncertainties in servicing the business sector and in ensuring integrity, rule of law, transparency and regulatory consistency to all.Maintaining high level of Integrity and transparency correlates with creating a favourable environment for private sector growth and Investment. If African countries are to meet the Millennium Development Goals (MDGs) both; governments and businesses need to share responsibility in any drive to create a eliminate corruption free in integrity environment that will automatically improve investment environment levels and quality.The reverse is true. Corruption for instance, can and does sabotage national development. Corruption leads to a loss of government legitimacy and of public trust and support. It inhibits the functioning of the market and distorts the allocation and use of resources, hence hampering trade and deterring investment. Corruption is often a collusive arrangement between some officials in government and individuals in business who have little or no concern for the social, economic, environmental, political or human consequences of their actions. Such decisions are taken not for the public benefit but merely to serve personal interests.Many countries in this era of democracy and transparency have intensified efforts at deepening reforms in service delivery institutions such as the judiciary, parliament, local government, public service management, etc. Progress has been achieved in many African countries though the journey ahead is long and difficult. Lack of an integrity environment impedes FDI flows too. Many African countries are pursuing market oriented economic policies, including divestiture of public enterprises, and creating an environment more conducing for business. This includes improving efficiency in production and service delivery. Yet the business environment overall is not yet fully conducive for FDI inflows.As per Transparent International report 2004 as well as the findings by the Economic Commission for Africa (ECA) study on governance, most of African countries performed badly in efforts to control corruption and creating integrity environment. There is a clear need to create and nurture an integrity environment which will prevent corruption and create a better environment for private sector development and investment promotion.RECENT INTERNATIONAL DEVELOPMENTSThe International community recognising the detrimental effects of corruption on development, has endorsed the UN Convention against Corruption. The Convention enhances cooperation and mutual legal assistance, particularly regarding return of stolen assets. As more and more countries ratify the Convention, corrupt practices in a foreign country are no longer beyond the reach of domestic jurisdictions. For Africa the adoption of African Union Convention on Prevention and Combating Corruption and Related Offences was a most important development. Nevertheless, as reported by Transparency International, the major weakness of the Convention is that it permits any signatory to opt out of some or all provisions!Under the NEPAD/African Union Programme, 24 African countries representing about 75% of the continent’s population have agreed to take part in peer reviews of their governance performance – African Peer Review Mechanism (APRM). Countries that endorsed APRM must conform to agreed values in four areas including, democracy and political governance, economic governance, corporate governance and socio-economic development. Tanzania is amongst African countries that endorsed the APRM. Nevertheless the success of creating an integrity environment and fight against corruption depends on commitment and political will to implement change.

INTEGRITY AND GOOD GOVERNANCE - THE CURRENT SITUATION (STATUS AND CONSTRAINTS) IN TANZANIA“We have persisted in our resolute struggle against corruption, including through rolling out plans to combat corruption; the establishment of anti-corruption bureaus at the district level; and enhanced accountability for resources transferred from the Central Government to the district level. Tanzania’s efforts in fighting corruption are starting to win international recognition.”(His Excellency President Benjamin W. Mkapa addressing the 4th International Investors’ Round Table (IRT) meeting at the Golden Tulip Hotel, 23rd November 2004)BackgroundTanzania like many other Sub-Saharan Africa countries achieved its independence with a severely underdeveloped economy and extremely limited infrastructure. Nevertheless, Tanzania has made concerted efforts to improve its economy, raise living standard of its people and create a conducive environment for private sector development & investment. Since early 1980s, governments of developing countries have been supporting and implementing strategies of encouraging competitive free markets, privatisation of state owned enterprises (parastatals), move from closed (no trade) to open (trading) economies and opening up the domestic economy through free trade and attracting foreign direct investment. This was done as a way of recognising the lead role that private sector can play in economic development. The private sector expressed concern that the system of governance in the region is still tinged with corrupt practices. According to the Transparency International’s annual Corruption Perception Index (CPI) for 2003, out of the 133 countries that were surveyed, the East African countries of Kenya, Uganda and Tanzania ranked relatively high in levels of corruption. The rankings were 122 for Kenya (with a CPI score of 1.9 out of 10), 113 for Uganda (with a CPI score of 2.2 and 92 for Tanzania (with a CPI score of 2.5).In all the three countries, efforts are being taken to curb corruption as systems are made to become more transparent with better placed to apply the rule of law in government operations. Nevertheless, a recent PWC report has suggested that the war on corruption in East Africa is being lost for the reason that of lack of political will in the high echelons, inadequate funding and equipment for anti-corruption institutions, an inappropriate legal framework and lukewarm enforcement as most bureaucrats who are charged with the responsibility of fighting corruption are themselves corrupt.Integrity and the Fight against Corruption

“Tanzania stands at the threshold of a new era: The new era demands a transformed public service. This will be a service that is truly transparent and accountable to the public. The service will have zero tolerance for corrupt behaviour. The service will guide the Nation as it crosses irreversibly into a poverty-free zone in the 21st Century.”

— H.E. President Benjamin W. Mkapa at the launch of PSRP on June 20th 2000.

In most African countries including Tanzania corruption was at a relatively low level during’ colonial rule. Nevertheless, after independence and the move to single party systems, which concentrated power into small cliques corruption, began to rear its ugly head. As time passed, the integrity environment became dramatically eroded.In order to foster integrity environment and create a better environment for investment and private sector development, the Tanzania government has taken several steps. First, in 1991, the government set up the Prevention of Corruption Bureau (PCB). In 1995, the government formed the Public Leadership Code of Ethics to curb impropriety at higher levels of public service and later it established the Commission for Human Rights and Good Governance (2001). In reality the government was preparing the legal framework to curb corruption and bring about integrity in public service. Later, the government established the Permanent Commission of Inquiry (Ombudsman) in 1996 to check abuse of power by government officials and its agencies.

When, President Benjamin William Mkapa came into power in 1995, he declared war on corruption so as to enhance the integrity environment. As a first step he established the Presidential Commission Against Corruption in January 1996. Ten months later, the Commission produced one of the most respected and commended analyses of corruption by any Africa state (commonly referred to as the Warioba Report). The report identified areas/environments where corruption occurs and also revealed regulations and procedures that facilitate corruption. It also cited examples of dubious decisions/contracts in several departments that were perceived corrupt.The Warioba report concluded that there was much evidence of Corruption. The report classified corruption into two categories. The first type relates to those who receive bribes to cater for their daily living needs (Petty Corruption) while the other group involves high level leaders and public officials, who are motivated by excessive greed for wealth accumulation and money (Grand Corruption). The Warioba Report had the further benefit of opening up public discussion on corruption.Accordingly, the government took several measures to address the problem. Such measures include: -

Appointment of a good governance Minister, who is responsible for among other things in monitoring overall strategy & implementation of anti-corruption measures,

Adoption of Natural Anti-Corruption Strategy for Tanzania. The strategy focuses on the need for transparency and accountability in government,

Appointment of the Prevention of Corruption Bureau (PCB). This is a unit that investigates and prosecutes corruption with the approval from the Director of Public Prosecutions (DPP),

Establishment of the Commission for Ethics to deal with administering and enquiring into senior public appointee’s Declaration of Assets and making recommendations to the president.

Along with above measures, which were implemented simultaneously with other government reform measures, Tanzania took holistic strategic approach to improve its governance system as reflected in “The National Framework on Good Governance, (NFGG) (Dec, 1999). This framework gave rise to the Accountability, Transparency and Integrity (ATIP) programme.This programme aims at supporting NFGG through: -

Strengthening the legal and judiciary system, Enhancing public financial accountability, Strengthening oversight and watchdog institutions.

The Public Sector Reform Programme of Tanzania, which was officially launched by His Excellency, the President of the United Republic of Tanzania in June 2000 aims to transform the public service into a result-oriented public service. Among other things, it aims to create a public service of the high calibre and integrity that is both responsive to and supportive of national efforts to deliver service to be competitive, to ensure good governance and to facilitate poverty reduction. The PSRP has been designed in pursuit of the vision, mission, core values and guiding principles that have been promulgated in the new public services management and employment policy. Major structural and institutional changes had been effectively implemented by the end of the PSRP phase of Government reform. These changes may be highlighted under the following seven headings:

Contracting and streamlining of Government structures; Reduction in employment numbers and wage bill control; Installation of an integrated Human Resources and Payroll Management system; Improved pay structure and enhanced salary levels; Restructuring and decentralisation for improved service delivery; Capacity building; and Improved policy and legislative environment for sustaining reforms.

The president speaks frequently on his determination to fight corruption. Nowadays there is the greater awareness of the evil effects of corruption, which are being openly discussed. Previously, some of the

greatest obstacles to curbing corruption were weak political will, weak institutions, and inadequate adherence to the rule of law, entrenched patronage, weak Private sector and weak civil society.Integrity Environment and Investment Promotion in TanzaniaIn recognition of the necessity to build an integrity environment hand in hand with economic reforms, Tanzania took several measures such as:

Tax reforms aimed at streamlining the tax structure, broadening the tax base, and establishing a sound institutional framework (establishment of TRA and introduction of Value Added Tax (VAT), in 1997.

Parastatal sector reforms through privatisation of state owned companies and institutional reforms. The government adopted privatisation programme whereby about more than 80% of 400 public entities that existed prior to 1993, had been divested by the end of 2003.

In 1999 commercial courts were established to speed up hearing of commercial disputes thus building investors’ confidence.

Public Service Reforms; which introduced client service charters in all its ministries and agencies.President’s Office assists departments and agencies to prepare the charters. These charters are derived from or are part of mission statements and focus on what the institution sets out to do for its customers. Charters increase accountability by setting performance standards. There are consequences when standards are consistently not achieved. Embedded in the charters are annual incentives such as awards to best performing workers and institutions. In addition, courses on public service customer care and anticorruption are being run by the department of Good Governance in the President’s Office to enhance service excellence, and adherence to a code of moral or ethical values including incorruptibility.

Financial sector reforms have been achieved by way of privatisation of state owned banks, allowing foreign banks to operate alongside local banks, establishment of the Capital Market and Securities Authority (CMSA) and setting up of the Dar es Salaam Stock Exchange (DSE).

Establishment of the Tanzania Investment Centre (TIC) with expanded mandate as “one stop centre” to promote and facilitate all investments and to advise the Government on investment matters. To minimize bureaucracy, the investment code also set a maximum period of 14 working days within which relevant government agencies were to have processed applications sent to them by TIC and that “where the Centre does not receive a written objection from the relevant authority within the specified time, the necessary license or approval shall be deemed to have been granted

Formation of the National Investment Steering Committee headed by Honourable Prime Minister. The Committee is entrusted with the task of investment policy formulation and solving intersectoral problems of investors on a fast track basis. Again this initiative was put in place to provide further momentum to the investment process in Tanzania.

Other measures taken to improve economic governance in summary were, Implementation of an integrated Financial Management System; Adopting an inclusive Public Expenditure Review/Medium Term Expenditure Framework

process; Establishment of a commercial court – as stated earlier. Revision of Public Finance Management and Public Procurement Act; Legal sector reform programme (LSRP) Local Government Reform Programme;

In Tanzania, there is in place a National Anti-Corruption Strategy and Action Plan (NACSAP), which basically brings together a coalition of stakeholders (public/private) to combat corruption.The NACSAP is envisaged to improve the quality of public service delivery through the following ways: -

Effective and transparent system of procedure and regulations. Enhance and organizational capacity to deliver high quality standard of service.

Public awareness on procedures, standards of services, codes of conduct and their rights in general.

Inputs for the NACSAP emanate from the key three stakeholders namely the Government, civil society and the private sector. This relationship is critically significant in the fight against corruption, No anticorruption strategy can succeed without the joint efforts of government, civil society and business. Thus, nurturing this relationship is vital in the fight against corruption.Recent Reforms and FDI InflowsTanzania has made major progress over recent years towards putting into place a policy environment for investment promotion. Increase investment flows are a sign of an improving integrity environment. Government priorities have included stable macroeconomic environment, privatization, promoting good governance and elimination of corruption, building a democratic nation, poverty eradication and development of a strong civil society.Policy reforms like trade liberalization, financial sector reforms, local government reforms, legal sector reforms, tax reforms, decentralization, civil services reforms and enforcing accountability and measures against corruption, have enabled Tanzania to improve efficiency and to increase economic growth and FDI inflows. The government has been fully supportive of public/private dialogue institutions. In this respect the Tanzania National Business Council (TNBC) was established and has provided a forum for public/private sector dialogue to improve business environment in the country. Its membership is 50% public and 50% private with President as Chair while the Vice Chairman is Chairperson of the Tanzania Private Sector Foundation. The Council addresses investment issues affecting domestic and foreign investors through respective Round Table meetings at which government’s service delivery is called to account, with cabinet ministers attending. The Government has in addition, embarked on the implementation of a programme of Business Environment Strengthening for Tanzania (BEST). The purpose of the programme, whose implementation started in December 2003, is to reduce the burden of businesses by eliminating as many procedural and administrative barriers as possible and to improve the quality of services provided by the government in the private sector including commercial disputes resolution. The five-year BEST plan shows that the targets will be achieved through five inter-linked components (Box).

Box: Summary of the Interlinked Components of the Best Programme

Components Main OutputsAchieving Better Regulation Unnecessary regulations removed

Sustainable process established for ensuring business friendly laws, regulations and administrative procedures

Improved efficiency and transparency of government institutions dealing with business

Improving Commercial Dispute Resolution

Improved accessibility to the court system for formal and informal business

Speed and quality of service provided by court system for business improvement

Strengthening the Tanzania Investment Centre

Increased number and value of local and foreign investment in

Tanzania Enhanced promotion of Tanzania as

investment destinationChanging the Culture of Government

Improved customer service ethos for services provided to the private sector by the public and judicial service

Empowering Private Sector Advocacy

Improved capacity of private sector stakeholders to identify regulatory problems and solutions and advocate for an improved business environment

Source: URT (Best programme Volume 1, 2003)

The progressive increase in FDI stock and flows revealed in figure 1, correspond with the investment policy evolution and promotional efforts that the Government undertook since early 1990s as discussed in the previous section. It is evident here that investors are sensitive to domestic policy actions. After the establishment of IPC in 1990, for example, the volume of FDI at first rose sharply and then stagnated after 1995 in response to administrative weaknesses in the IPC.

Figure 1: shows considerable rise in annual FDI flows between 1990-2004 and a continuous increase in annual FDI stocks from 1998-2004.

Key: e=BOT estimates; p=BOT projectionsSource: IFC (1997) for 1990-1991 and Bank of Tanzania for 1992-2004.Tanzania has generally acquired a sound reputation internationally. For instance, in April 2004, Tanzania was rated by UNCTAD survey of as the 2nd most attractive destination in Africa, just behind South Africa. In 19th November 2004 Tanzania Investment Centre was chosen as the Africa’s best Investment Promotion Agency of the year (2004) by Africa Investor Media Group, (http\\www.Africa-investors.com).Obstacles to Integrity Environment and Good Governance in Investment Promotion

As per the Warioba Report and other surveys on corruption and improving integrity such as the study “Striving for Good Governance in Africa” done by UN - Economic Commission for Africa (ECA), it has been revealed that although constitutional government is getting stronger and more democratic, countries perform badly in efforts to control corruption and building an integrity environment. For instance corruption is rampant in the tax systems, police and the judiciary. The ECA study released in October 2004 showed that in many African countries police and the military violate the rights of civilians, electoral commissions need more independence, and costs and red tape greatly hinder business and investment activities in Africa.The report identifies 10 areas in need of urgent action including strengthening parliaments, protecting the autonomy of the judiciary, improving the performance of the public sector, supporting the development of professional media encouraging private investment and decentralizing the delivery of services.On the other hand, the UNCTAD Study on Good Governance in Investment Promotion and Facilitation (2002), major obstacles to Good Governance in Investment Promotion are: -

Limited infrastructure, and institutional capacity Capital resources/financial constraints, Skills and education, Transportation, Legal system and crime, Bureaucracy and corruption, Institutionalised negative mindset.

As per the TIC/BoT/NBS investment survey (2003) on investors’ perceptions on Good Governance, results show that seven out of eleven factors had negative effect and only four were rated favourably (Figure 2). Factors considered included bureaucracy, corporate corruption, custom procedures, domestic political stability, legal system, internal security, investment facilitation, public sector corruption, regional political stability, speed of government decision-making and tax collection efficiency.

Figure 2: Investors’ Perception on Political and Governance Factors

Regional and domestic political stability, investment facilitation, internal security, are four out of eleven factors that were positively rated (Figure 2). On the other hand, factors that were negatively rated include; bureaucracy, corruption, effectiveness of the legal system, customs procedure, tax collection efficiency and speed of Government decision-making process. All these factors though rated negative, have been improving between “start-up” and 2003 periods. Reasons behind the improvements include, among others, the sustained efforts by the Government on the fight against corruption as well as reduction of bureaucracy by transforming the speed of Government decision-making process.

THE NEED FOR CORPORATE GOVERNANCEAs the private sector is increasingly becoming a key partner in development in many African countries, the importance of creating better environments for private sector development is critical. Similarly, in order to build integrity and attract both local and foreign investors, the private sector has to accept, and implement Corporate Governance. As a recent World Bank report stated, “Corporate Governance is concerned with holding the balance between economic and social goods and between individual and communal goals, the aim being aligning as nearly as possible the interest of individuals, corporations and society.” In many African countries nowadays, corporate governance has become a critical element of business management and economic growth. Elsewhere, massive economic crises and corporate failures have been closely associated with the lack of good corporate governance. Lack of sound corporate governance has fuelled corruption and cronyism while suppressing sound and sustainable economic decisions. Results from some surveys carried out, demonstrated that there is a positive correlation

between good corporate governance and increasing capital flows. It has been suggested that good corporate governance reduces cost of capital as it provides a positive global premium that tends to attract investment. Pillars of corporate governance include transparency, accountability, probity and respect for the rights of all stakeholders.The OECD Principles on corporate governance (2004) constitutes a balanced benchmark for corporate governance. The Principles are intended to assist OECD and non-OECD governments in their efforts to evaluate and improve the legal, institutional and regulatory framework for corporate governance in their countries, and to provide guidance and suggestions for stock exchanges, investors, corporations, and other parties that have a role in the process of developing good corporate governance. They also provide the basis for an extensive programme of cooperation between OECD and non-OECD countries and underpin the corporate governance component of World Bank/IMF Reports on the Observance of Standards and Codes (ROSC). Nevertheless, as the world is becoming smaller and smaller and with increased globalisation, companies/corporations worldwide cannot escape from the global movement that shapes standard principles of corporate governance. Tanzania has so far developed her own national code of corporate governance.

THE WAY FORWARDThough in many African countries, leadership has have done much to advance their national economies over the past decade nevertheless much remains to be done, as overall economies are small and underdeveloped. There is no conducive environment for investment promotion and economic growth. Good governance and enhanced Integrity environment as pillars to better economic and investment climate are yet to be achieved. Many countries are perceived to be corrupt as the political will to counter corruption is perceived as lacking. It is true in many countries that there is “much talk” but little real action against corruption.African Leaders have to institute and promote good governance and integrity environment as a foundation for creating a favourable investment climate. Government operating an integrity environment creates willingness in people and the business community to trust and cooperate. Courage is of utmost importance in maintaining integrity. Without strong, clear leadership and support throughout the government, its various departments and the law enforcers, it is very difficult for any real advance in the fight against the evils of corruption to succeed.Government and corporate leaders must: -

Live and openly role-model integrity in all their dealings, Make morally and ethically correct decisions regardless of cost or difficulty, Hold on to the vision and of clean government and clean business, Be courageous in sharing their sentiments and recommendations, Have clear values to guide them so as to take bold decisions, Have the courage of owning their mistakes and learning from them.

All African Countries must rectify the African Union Convention on Prevention and Combating Corruption and Related Offences to remove avoidance clauses. The must also agree to take part in peer reviews of their governance performance. Measures like strengthening the legal and judiciary system, enhancing public financial accountability and strengthening oversight and watchdog institutions are critical in promoting integrity environment.Another significant factor in promoting good governance is public awareness on procedures, standards of services, codes of conduct and their rights in general. Information regarding services provided by Ministries, Departments and Agencies should be spelt out clearly in the client charters and disseminated freely. Success in creating an integrity environment for investment promotion and economic growth depends on commitment and political will to implement necessary reforms as recommended.

Finally, WITHOUT GENUINE COMMITMENT – ALL IS LOST!!Source: http://www.oecd.org/dataoecd/11/37/34571058.pdfQuestions:

1. Do you agree with the words “WITHOUT GENUINE COMMITMENT–ALL IS LOST!!” in case of Tanzanian Corporate Governance? Write your views.

2. What other factors except those discussed in the above case study can be implemented to promote good governance in the Tanzania?