institutions, financial markets, and firm debt maturity

qWe would like to thank Jerry Caprio, Ross Levine, Tim Opler, Fabio Schiantarelli, MaryShirley, and Sheridan Titman for helpful comments and Jim Kuhn for help with the data. We arealso grateful to the referee, Cli!ord J. Smith, Jr., and to the participants of the June 1996 World BankConference on Term Finance, and the 1997 Western Finance Association Meetings in San Diego.The views expressed here are the authors' own and not necessarily those of the World Bank or itsmember countries.

*Corresponding author. Tel.: #1-301-405-2125; fax: #1-301-314-9157.

E-mail address: [email protected] (V. Maksimovic)

Journal of Financial Economics 54 (1999) 295}336

Institutions, "nancial markets, and "rm debtmaturityq

Asli DemirguK c7 -Kunt!, Vojislav Maksimovic",*!The World Bank, 1818 H Street NW, Washington, DC 20433, USA

"Robert H. Smith School of Business, The University of Maryland, College Park, MD 20742, USA

Received 13 January 1998; received in revised form 30 June 1998

Abstract

We examine "rm debt maturity in 30 countries during the period 1980}1991. Incountries with active stock markets, large "rms have more long-term debt. Stock marketactivity is not correlated with debt levels of small "rms. By contrast, in countries witha large banking sector, small "rms have less short-term debt and their debt is of longermaturity. Variation in the size of the banking sector is uncorrelated with the capitalstructures of large "rms. Government subsidies to industry are positively related andin#ation is negatively related to the use of long-term debt. We also "nd evidence ofmaturity matching. ( 1999 Elsevier Science S.A. All rights reserved.

JEL classixcation: G20; G32; K10

Keywords: Capital structure; Debt maturity; Financial institutions; Legal system

0304-405X/99/$ - see front matter ( 1999 Elsevier Science S.A. All rights reserved.PII: S 0 3 0 4 - 4 0 5 X ( 9 9 ) 0 0 0 3 9 - 2

1Diamond (1991,1993) and Rajan (1992) discuss the choice of maturity structures and the choiceof whether to borrow from the market or from an intermediary who has an advantage in monitoring.For recent examples of optimal "nancial structures when investors can observe the "rm's cash #owsbut cannot enforce legal rights to these cash #ows, see Hart and Moore (1995) and Bolton andScharfstein (1993). For reviews of the literature see Ravid (1996) and Harris and Raviv (1990).

1. Introduction

Con#icts of interest between a "rm's insiders and outside investors areimportant determinants of the "rm's ability to obtain capital. Jensen andMeckling (1976), Myers (1977) and Myers and Majluf (1984), among others,suggest that these con#icts can be mitigated by the appropriate choice ofsecurities or contracts between the "rm and its investors. An extensive theoret-ical literature in corporate "nance shows that the optimal choice of securities forthis purpose depends on the information available to investors and their abilityto monitor compliance and enforce their legal rights.1 Since the amount ofinformation available to investors and their ability to protect their investmentboth depend on "nancial and legal institutions, "rms' "nancial structures shoulddi!er systematically across countries. However, little is known about howobserved di!erences in the institutional and legal environments across countriesa!ect the "nancing choices of "rms.

In this paper we examine how di!erences in "nancial and legal institutionsa!ect the use of debt and especially the choice of debt maturity by "rms ina sample of 30 countries in the period 1980}1991. The sample includes bothdeveloped and developing countries as well as countries with both common-lawand civil-law legal systems. We ask four questions. First, are there any system-atic di!erences in the maturity of debt claims issued by "rms in di!erentcountries? Second, can any such di!erences be accounted for by the character-istics of the "rms in each country? Third, can the di!erences in the use of debt beexplained by institutional di!erences, particularly in the development of mar-kets and the enforceability of contracts? Di!erences in the use of debt couldoccur if institutional arrangements in each country facilitate the use of particularsecurities to control opportunistic behavior by a "rm's insiders. Finally, is thereevidence that some "rms, especially small "rms, obtain less long-term debt"nancing in countries with less-developed "nancial systems?

Several authors explore the e!ect of the institutional environment on "rm"nancing choices in speci"c countries. Hoshi et al. (1990) show that membershipin industrial groups linked to banks reduces "nancial constraints on Japanese"rms. Calomiris (1993) examines the e!ect of di!erences between the bankingsystems of the U.S. and Germany on "rm "nancing. He argues that large-scaleuniversal banks in Germany engage in long-term monitoring which is re#ected

296 A. Demirgu( c7 -Kunt, V. Maksimovic / Journal of Financial Economics 54 (1999) 295}336

in a higher propensity of German "rms to issue equity. Rajan and Zingales(1995) explore capital structure decisions of "rms in seven developed countries.They "nd that variables which are commonly used to explain "nancial structurein the U.S. are also correlated with leverage in their sample of foreign "rms.DemirguK c7 -Kunt and Maksimovic (1995) consider "nancing choices in a sampleof ten developing countries and "nd analogous, albeit weaker, results.

Recently, Barclay and Smith (1995) and Stohs and Mauer (1996) have exam-ined term "nancing in the U.S. Barclay and Smith (1995) "nd strong support forthe hypothesis that "rms with larger information asymmetries issue moreshort-term debt. Stohs and Mauer (1996) "nd that long-term debt is issued bylarger, less-risky "rms in relatively low-growth industries. There have been fewercross-sectional studies of the e!ect of "nancial and legal institutions on "rm"nancing. DemirguK c7 -Kunt and Maksimovic (1998) explore the relation between"rm growth and access to external "nance for a sample of both developed anddeveloping countries. They show that the proportion in each country of "rmsthat grow at rates that exceed those that can be "nanced internally is correlatedwith the perceived e!ectiveness of the country's legal system and several indi-cators of "nancial market and institutional development. Rajan and Zingales(1998) examine the e!ect of the development of "nancial institutions on industrygrowth in a sample of countries, and DemirguK c7 -Kunt and Maksimovic (1996)explore complementarities in stock market and banking-sector development onthe "nancing decision of "rms in a cross-country sample of "rms, but neitherpaper addresses the question of debt maturity or the quality of enforcement ofcontracts by the legal systems in each country. In a comparative study of legalsystems, La Porta et al. (1998), hereafter LLSV, argue that the tradition onwhich a country's system of law is based as well as several speci"c protectionsare important in determining whether investors can enforce their claims on the"rm's assets. Their paper classi"es the legal systems of a sample of countriesaccording to their legal tradition and whether or not they grant investors thosespeci"c protections. In the tests below, we use their classi"cation of legal systemsto supplement an index measuring the e!ectiveness of each country's legalsystem.

The rest of the paper is organized as follows. In Section 2 we take a prelimi-nary look at the di!erences in term "nancing between countries. In Section 3 wediscuss possible explanations for these di!erences. Section 4 introduces the dataand presents summary statistics. Section 5 reports cross-sectional empirical testsof "nancial maturity across countries. Section 6 concludes.

2. Cross-country comparison of term 5nancing

Financial theory suggests that a major factor in a "rm's choice of capitalstructure is the existence of agency costs. The extent to which these costs can be

A. Demirgu( c7 -Kunt, V. Maksimovic / Journal of Financial Economics 54 (1999) 295}336 297

controlled by appropriate "nancial contracts depends on both the character-istics of "rms and the institutional environment in which the contracting takesplace. Thus, since countries have very di!erent institutional systems and "rmcharacteristics, the observed "nancial structures in a cross-country sampleshould vary systematically across both countries and "rms.

We can obtain an initial assessment of the extent of these di!erences bycomparing the long-term and short-term indebtedness of "rms for a sample ofcountries at di!erent levels of economic development. Our sample consists of"rms in 19 developed economies and 11 developing economies for the period1980}1991. The developed countries in our sample are Austria, Australia,Belgium, Canada, Finland, France, Germany, Hong Kong, Italy, Japan, theNetherlands, New Zealand, Norway, Sweden, Singapore, Spain, Switzerland,the United Kingdom, and the U.S. The developing countries are Brazil, India,Jordan, Korea, Malaysia, Mexico, Pakistan, South Africa, Thailand, Turkey,and Zimbabwe. The selection of countries and the variables discussed in thispaper are described in detail in Section 3.

For the developed economies we obtain "rm-level data from Global Vantage.We include all the countries in the database for which there are more than 40"rms available. We eliminate observations that do not lie within a band centeredon the median observation and having a width 12 times the interval between the5th percentile and 95th percentile observations. (Fewer than 1% of observationsare eliminated.) The "rm-level data for developing countries are from theInternational Finance Corporation's (IFC) database. They consist of "nancialstatement data for the (approximately) 100 largest publicly traded corporationsin these economies. The IFC data are described in detail, together with primarysources, in Singh et al. (1992); the data for Brazil are from the publications of theVargas Foundation of Brazil. For both databases, the number of "rms availablein each country, the years available, and the calculation of each variable we useare described in the appendix.

Fig. 1 displays the average ratios of long-term liabilities (measured as totalliabilities less current liabilities) to total assets for "rms in our sample for each ofthe 30 countries. The developing countries in our sample are denoted by thedarker outline. Norway has the highest ratio of long-term debt to assets,whereas Zimbabwe has the lowest, at about one-"fth of Norway's. There isa marked clustering of developing countries at the bottom of the range, indicat-ing that "rms in these countries do not employ as much long-term debt"nancing. This pattern is con"rmed in Fig. 2, which displays the ratio oflong-term to total liabilities in our sample of countries. Firms in developingcountries use less long-term debt as a proportion of total debt.

The di!erences in "nancing patterns across countries re#ect di!erences ininstitutions and contracting environments across countries. However, "rms withdi!erent characteristics have di!erent access to "nancial markets and institu-tions even within the same economy. In particular, smaller "rms are likely to


Fig. 1. Long-term debt/total asset ratios. The "gure presents the average long-term debt to totalasset ratios for "rms in each country for the 1980}1991 period. Developing countries are denoted bythe darker outline. The countries in the "gure are ordered by their utilization of long-term debt"nancing.

have higher monitoring costs than larger "rms, relative to the amount of theloan. We expect these di!erences to be re#ected in di!erent "nancing patterns.Fig. 3 depicts the ratios of short-term, long-term, and total indebtedness to totalassets and the ratio of long-term to total debt by "rm size. We divide the "rms ineach country in the sample into quartiles by value of total assets, and report theaverage debt ratios of each quartile calculated across countries. Inspection of the"gure reveals that there are marked and consistent di!erences across quartiles inthe use of long-term debt. Large "rms report higher ratios of long-term debt tototal assets and long-term debt to total liabilities. By contrast, there do notappear to be di!erences in the ratios of short-term debt to total assets across"rm-size quartiles.

The "gures indicate that there are di!erences in "nancing patterns for coun-tries at di!erent levels of development and for large and small "rms. The mostpronounced di!erences are in the use of long-term debt contracts. In principle,greater reliance on long-term debt in more-developed countries could be at-tributable to di!erences in the type of assets owned by "rms in developed anddeveloping countries. Thus, if "rms in developed countries own more "xedassets, which have longer maturity, then the di!erences in capital structures can


Fig. 2. Long-term debt/total debt ratios. The "gure presents the average long-term debt to totaldebt ratios for "rms in each country for the 1980}1991 period. Developing countries are denoted bythe darker outline. The countries in the "gure are ordered by their debt maturity.

be explained by simple maturity matching. We explore this possibility byplotting the average ratio of net "xed assets to "xed assets for the sample of "rmsin Fig. 4. Inspection of Fig. 4 shows that "rms in several developing countrieshave higher ratios of net "xed assets to total assets than "rms in many developedcountries. Thus, simple maturity matching cannot explain the variation inlong-term "nancing.

A more formal measure of the covariation of the level of development andlong-term corporate indebtedness can be obtained by regressing the ratio oflong-term debt to total assets on gross domestic product (GDP) per capita andcountry dummies. For large "rms these variables &explain' 44% of the covari-ation in long-term "nancing over the sample period. The size of the coe$cientindicates that di!erences in per capita GDP in our sample are associated withvery di!erent levels of long-term debt. Thus, a relatively small $1000 increase inper capita GDP (the di!erence between, say, Pakistan and Thailand) translatesinto an increment of 0.09 in the value of the ratio of long-term debt to totalassets. Increases of $10,000 in per capita GDP (the di!erence between, say,Pakistan and Singapore) translates into an increase of 0.09 in long-term lever-age, whereas di!erences between some of the richest and poorest countries in thesample (the di!erence of approximately $20,000 between, say, Pakistan andNorway) is associated with an impressive increase of 0.18 in the value of the ratioof long-term debt to total assets. The results of corresponding regressions of the


Fig. 3. Debt ratios: small vs. large "rms. The "gure presents the average long-term debt to asset ratio(LTD/TA), short-term debt to assets (STD/TA), total debt to total assets (TD/TA), and long-termdebt to total debt (LTD/TD) across 30 countries by "rm size. The "rms in each country are dividedinto quartiles by value of total assets, and the average debt ratios of each quartile, calculated acrosscountries, is reported. Countries in the sample are: Australia, Austria, Belgium, Brazil, Canada,Germany, Finland, France, Hong Kong, India, Italy, Jordan, Japan, Korea, Malaysia, Mexico,Netherlands, Norway, New Zealand, Pakistan, Singapore, Spain, Switzerland, Sweden, Thailand,Turkey, United Kingdom, United States, South Africa, and Zimbabwe.

ratio of long-term debt to total assets for small "rms on per capita GDP arequalitatively similar. Together with country dummies, per capita GDP &explains'41% of the variation.

In the remainder of the paper we take a closer look at the di!erences in the"nancing of "rms across countries and test whether it can be explained by "rmcharacteristics and the characteristics of contracting environments and "nancialinstitutions.

3. Markets, institutions, and debt maturity

In order for a "rm to obtain outside "nancing, and particularly loans, the "rmmust credibly commit to investors to respect contracts that control opportunis-tic behavior. The types of contracts that permit commitment in any particularcase depend both on "rm characteristics (Smith and Warner, 1979) and on theinstitutions in the economy that facilitate monitoring and enforcement of"nancial contracts.

When the legal system is ine$cient or costly to use, short-term debt is morelikely to be employed than long-term debt. As Diamond (1991,1993) and Rajan


Fig. 4. Net "xed assets/total assets ratios. The "gure presents the average net "xed asset to totalasset ratios for "rms in each country for 1980}1991. Developing countries are denoted by the darkeroutline. The countries in the "gure are ordered by the value of the ratio.

2We are grateful to the referee for bringing this aspect of the debt management policy of the NewZealand government to our attention.

(1992) argue, short-term "nancing makes it more di$cult for borrowers todefraud creditors. Shorter maturities limit the period during which an opportun-istic "rm can exploit its creditors without being in default. The creditors canreview the "rm's decisions more frequently and, if necessary, vary the termsof the "nancing before su$cient losses have accumulated to make default bythe borrower optimal. Thus, we would expect an inverse relation between theine$ciency of a country's legal system and the use of long-term debt.2 To theextent that there are "xed litigation costs in enforcing contracts, long-term debtis likely to be used most heavily by large "rms. The "xed costs also make the useof long-term debt, particularly by small "rms, less responsive to small year-to-year changes in the economic environment. Of course, this presupposes theexistence of a trade-o! between the use of long-term and short-term debt.

Governments can facilitate the issuance of long-term debt by maintaininga predictable currency value. High and/or variable rates of in#ation make itcostly for investors and "rms to contract. This contracting problem caused by


in#ation is compounded when the legal resolution of disputes is subject to delay.In principle, debt contracts can be indexed. For example, in Brazil all contractsspecify a government price index used to adjust the nominal payments forin#ation. However, the indexes are subject to adjustments made for politicalreasons. Furthermore, the judicial system does not index judgments, which aresubject to appeal and other delays. Perhaps not coincidentally, Table 1 revealsthat Brazilian "rms have little long-term debt.

The government can promote the use of long-term "nancing directly bygranting implicit loan guarantees when it adopts a policy of subsidizing loss-generating "rms or sectors. Governments can also encourage the development oflonger-maturity public debt markets by choosing to issue public debt with longmaturities because a market in long-term government debt gives investors in-formation about the risk-free term structure.

Financial intermediaries and stock markets directly in#uence the "nancialstructures of "rms. A prime function of "nancial intermediaries, such as banks, isthat of monitoring borrowers. As Diamond (1984) argues, intermediaries haveeconomies of scale in obtaining information. Intermediaries also have greaterincentives to use the collected information to discipline borrowers than do smallinvestors subject to free-rider problems. Thus, we would expect that a developedbanking sector would facilitate access to external "nance, particularly amongsmaller "rms. The implications for the debt maturity of "rms are less clear.A developed banking sector leads to an increase in the availability of short-term"nancing, since this form of "nancing enables intermediaries to use their com-parative advantage in monitoring. However, banks' economies of scale and theirability to monitor covenants also permit them to o!er long-term loans thatwould not be available in a market without intermediaries. Which of thesetendencies predominates is an empirical question.

Developed stock markets provide opportunities for diversi"cation by entre-preneurs. Thus, in countries with developed stock markets, there is an incentivefor "rms to switch from long-term debt to equity. However, stock markets alsoa!ect transmission of information that is useful to creditors. As Grossman (1976)and Grossman and Stiglitz (1980) demonstrate, prices quoted in "nancialmarkets at least partially reveal information that more-informed investorspossess. This revelation of information makes lending to a publicly quoted "rmless risky. As a result, the existence of active stock markets increases the abilityof "rms to obtain long-term credit. On the other hand, the additional liquiditythat stock markets provide makes it easier for informed shareholders to escapethe consequences of failed gambles, and therefore encourages risk-taking behav-ior costly to shareholders. Which of these e!ects predominates is anotherempirical question. Initial evidence by DemirguK c7 -Kunt and Maksimovic (1996)suggests that the informational e!ect is stronger and that in countries withdeveloping "nancial markets debt}equity ratios increase with an increase instock market size and activity.


The amount of long-term and short-term debt that is optimal even when"nancial markets are perfect depends in general on the opportunities that the"rm's insiders have for diverting resources and on the assets that the "rm cano!er as collateral. Thus, theory predicts that "rms whose principal asset is thepresent value of growth opportunities do not optimally borrow against thatasset (Myers, 1977). By contrast, "rms with a large quantity of "xed assetsalready in place do not distort their incentive value when they borrow. The "xedassets also facilitate borrowing by serving as collateral.

Barclay and Smith (1995) "nd that these predictions of the theory aresupported in the U.S. We expect that in an international context, the observed"nancial structure choices depend on these considerations and also on theinstitutional factors discussed above. We next investigate the relationempirically.

4. Firms and countries in our sample

4.1. Economic variables

In Table 1 we summarize some important facts about the economic develop-ment of the countries in our sample. (The sources for the variables discussed inthis section are given in the appendix.) Per Capita GDP is a broad indicator ofdi!erences in wealth in each country. In 1991, per capita GDP in the sampleranges from $27,492 in Switzerland to $359 in Pakistan. Thus, the sampleincludes some of the richest and poorest countries in the world.

Three additional macroeconomic indicators are presented in Table 1. Theaverage annual growth rate in GDP over the sample period is an indicator of the"nancing needs of "rms. On an individual "rm level, the growth rate is a proxyfor the investment opportunity set faced by "rms (Smith and Watts, 1992) and itse!ect on the optimal "nancing of projects (Myers, 1977). The average in#ationrate over the sample period, shown in the third column, provides both anindicator of the government's management of the economy and evidence onwhether the local currency provides a stable measure of value to be used inlong-term contracting. There are major variations in the average rate of in#a-tion in the sample countries. The average annual rate of in#ation is highest inBrazil, at 327.6%, and lowest in Japan, at 1.5% per annum.

The "nal economic indicator shown in Table 1 is a measure of the govern-ment's subsidies to the corporate sector in each country. Government subsidiesa!ect "nancial structure decisions because implicit or explicit backing of cor-porations by the government distorts market incentives and permits some "rmsto obtain long-term loans on favorable terms. Our measure of the government'ssubsidies is the level of government grants as a percentage of GDP. Moreprecisely, we measure the sum of grants on current account by the public


Table 1Economic indicators

GDP/Cap is the real GDP per capita in US$ in 1991, obtained from World Bank NationalAccounts. Growth rate is the average annual growth rate in GDP/Cap for the period 1980}1991.Average annual in#ation is given for the period 1980}1991. It is the annual in#ation of the GDPde#ator, obtained from World Bank National Accounts. Government subsidies are de"ned as grantson current account by the public authorities to private industries and public corporations as well asto government enterprises. The "gures are a percent of GDP averaged over 1983}1991. Data areobtained from various issues of the World Competitiveness Report, The World Economic Forum,IMD International, Geneva, Switzerland.

GDP/Cap(US $)

Growth1980}1991(%)

In#ation1980}1991(%)

Governmentsubsidies toprivate and publicenterprises1983}1991

Switzerland 27,492 1.7 3.8 1.4Japan 23,584 3.9 1.5 0.6Norway 19,664 1.7 5.2 5.9Sweden 19,649 1.6 7.4 4.8United States 18,972 1.9 4.2 0.6Finland 18,046 1.6 6.6 3.0France 17,365 1.8 5.7 2.4Austria 17,288 2.2 3.6 1.3Netherlands 16,479 2.3 1.8 2.6Germany 16,439 1.8 2.8 2.0Canada 16,098 2.0 4.3 1.9Belgium 16,051 2.2 4.2 3.5Italy 14,570 2.5 9.5 2.9Australia 13,095 1.6 7.0 3.0United Kingdom 12,585 2.3 5.8 1.5New Zealand 10,643 1.0 10.3 1.2Singapore 10,294 4.9 1.9 1.9Hong Kong 9820 5.8 7.5 n.a.Spain 8752 3.3 8.9 2.4Korea 4259 6.8 5.6 6.3Malaysia 2465 3.6 1.7 4.6South Africa 2198 1.0 14.4 n.a.Brazil 2073 2.1 327.6 10.7Mexico 1801 1.0 66.5 2.3Turkey 1375 3.1 44.7 2.2Jordan 1372 !2.1 1.6 n.a.Thailand 1362 7.0 3.7 1.4Zimbabwe 630 1.7 12.5 n.a.India 375 3.3 8.2 5.8Pakistan 359 3.9 7.0 5.4


authorities to private industries and public corporations as well as to govern-ment enterprises. These grants presumably compensate for losses that are theconsequence of policies of the public authorities. Thus, this variable measuresrealized expenditures but not direct instructions to business or the level of exante commitments made by each government. Over a period, we would expecta correlation between commitments and expenditures. As the last column ofTable 1 reveals, the level of government subsidies is signi"cant is some countries,and exceeds 10% of GDP in the case of Brazil.

4.2. Legal and xnancial institutions

We explore the relation between "rms' "nancing choices and the state ofdevelopment of both the legal and "nancial institutions in our sample ofcountries. The principal indicators of legal and "nancial development are givenin Table 2.

4.2.1. Legal institutionsAs an indicator of the e$ciency of the legal system in each country, we use

a commercially available index of the level of law and order in each country.This index, prepared by the International Country Risk Guide, is scored ona scale of ten to six and aggregates annual reports by a panel of more than 100analysts. It measures the extent to which citizens of a country are willing to relyon established institutions to make and implement laws and to adjudicatedisputes. Low levels of the index denote less reliance on the legal system tomediate disputes. A second indicator, the index of legal e$ciency, produced byBusiness International Corporation, is also presented for comparison. Thissecond indicator is an index of the e$ciency and integrity of the legal environ-ment as it a!ects business, and particularly foreign "rms. This index is scoredfrom zero to ten, with lower scores indicating lower e$ciency.

LLSV, argue that legal systems based on common law o!er investors di!erentprotections than those based on civil law. Watson (1974) also discusses di!er-ences in legal traditions based on common law and civil law, and argues thelatter provide individual investors better protections. Such di!erences translateinto di!erences in the optimal contracts between "rms and investors. To test forthis relation, we follow LLSV in de"ning an indicator variable that equals one ifthe country's legal system is based on common law and zero if it based on civillaw. As Table 2 reveals, the legal systems of 13 countries in our sample are basedon common law and those of 17 countries are based on civil law.

Financial structure choices should also be a!ected by the speci"c provisionsof each country's commercial laws. To investigate further the e!ect of di!erencesin legal systems we use the indicators of creditor and shareholder rights com-piled and discussed in detail in LLSV. LLSV classify countries according towhether they provide creditors with four speci"c protections. The "rst is


whether the bankruptcy laws prohibit an automatic stay on assets, which wouldprevent automatic liquidations of insolvent "rms by secured creditors. Theexistence of an automatic stay gives managers and shareholders of a distressed"rm greater bargaining power over secured creditors. This provision alsobene"ts unsecured creditors over secured creditors. The second is whethersecured creditors are permitted to repossess their collateral in bankruptcy orwhether some third party claims, such as those of the government or theemployees, take priority. The third is whether the bankruptcy law prohibitsborrowers from unilaterally obtaining court protection from creditor demands.If distressed borrowers can obtain such protection unilaterally, their bargainingpower is increased. The fourth is whether creditors can dismiss managers andreplace them with administrators when a "rm becomes bankrupt. In addition,LLSV note whether or not the law of each country requires all "rms to maintaina reserve of equity capital. In countries where this requirement exists, "rms thatdo not ful"ll it can be dissolved.

In principle, the creditor rights identi"ed are important in de"ning feasiblecontracts between "rms and investors. However, there need be no direct statist-ical relation between the existence of a speci"c right and a speci"c "nancialcontract, such as long-term debt, even when that right is important in enforcingthe contract. For example, if the existence of a speci"c right is necessary, but notsu$cient, to make a "nancial contract enforceable, the statistical relationbetween that right and the use of the contract will be weak. The relation betweena particular creditor protection and a particular debt contract is also a!ected bythe existence of spillover e!ects of the creditor protection on other contracts.For example, strong creditor rights increase the incentives of "nancial institu-tions to monitor "rms, thereby also making stock investments in those "rmsmore attractive. The size of these spillovers depends on the development of thestock market and "nancial institutions and on the precise provisions of theinvestor-protection laws. Spillovers also work in the opposite direction. In somecountries, "nancial intermediaries hold both the equity and debt of corpora-tions. Hauswald (1996) shows that ownership of both stock and equity byintermediaries increases their incentives to reorganize "rms in distress. Inter-mediaries with an equity stake in a "rm should be willing to make loans evenwhen creditor protection is relatively weak.

Our examination of the relations between speci"c creditor protections and"nancial structure is thus exploratory in nature. We give each country a score onan empirically de"ned &index of creditor rights' based on whether its laws grantcreditors the legal protections identi"ed above. Speci"cally, we give each coun-try a score of one for each of the following conditions that its bankruptcy lawsatis"es: (i) does not permit an automatic stay on assets, (ii) does not allowborrowers to unilaterally seek bankruptcy protection, (iii) assures secured credi-tors the right to collateral, and (iv) does not grant the managers tenure pendingresolution of bankruptcy. If corporations are required to maintain a capital


Tab

le2

Inst

itutiona

lin

dic

ators

Thela

wan

dor

der

indic

ator,

pro

duc

edby

theIn

tern

atio

nal

Count

ryR

isk

rating

agen

cy,r

e#ec

tsth

ede

gree

tow

hich

theci

tize

nsof

aco

untr

yar

ew

illin

gto

acce

ptth

ees

tablis

hed

inst

itution

sto

mak

ean

dim

ple

men

tla

ws

and

adju

dica

tedi

sput

es.I

tra

nges

from

zero

tosix

with

high

ersc

ores

indic

atin

gso

und

pol

itic

alin

stitutions

and

ast

rong

court

syst

eman

dlo

wer

scor

esin

dica

ting

atr

aditio

nof

depen

ding

on

phys

ical

forc

eor

illeg

alm

eans

tose

ttle

clai

ms.

Val

uesre

por

ted

are

1985}19

91av

erag

es.T

hele

gale$

cien

cyin

dica

tor,

pro

duce

dby

Bus

ines

sIn

tern

atio

nalC

orp

ora

tion,i

san

asse

ssm

entoft

he

e$ci

ency

and

inte

grity

ofth

ele

gale

nvironm

entas

ita!

ects

busines

s,par

ticu

larly

fore

ign"rm

s.It

range

sfrom

zero

tote

nw

ith

low

ersc

ore

sfo

rlo

wer

e$ci

ency

leve

ls.

An

aver

age

valu

efo

r19

80}19

83is

avai

labl

e.The

com

mon

-law

dum

my

equal

sone

for

com

mon

-law

coun

trie

san

dze

rofo

roth

ers.

Cre

dito

rrigh

tsis

anin

dex

that

rang

esfrom

zero

to4.

5an

dag

greg

ates

cred

itor

righ

ts.S

har

ehol

der

righ

tsis

anin

dex

that

rang

esfrom

zero

to"ve

and

aggr

egat

essh

areh

olde

rrigh

tsas

desc

ribe

din

thete

xt.T

hese

thre

eva

riab

lesar

eob

tain

edfrom

La

Por

taet

al.(

1998

).M

arke

tcap

ital

izat

ion/G

DP

isth

est

ock

mar

ket

capi

taliz

atio

ndiv

ided

by

GD

P.T

urn

ove

ris

the

tota

lval

ue

ofsh

ares

trad

eddiv

ided

by

mar

ket

capi

taliza

tion.S

tock

mar

ket

dat

aar

efrom

IFC's

Em

ergi

ng

Mar

ketD

ata

Bas

e.V

alues

are

1980}19

91av

erag

es.B

ank/G

DP

isth

eto

talas

sets

ofth

ede

pos

itm

one

yban

ksdi

vide

dby

GD

Pob

tain

edfrom

IMF,In

tern

atio

nal

Fin

anci

alSta

tist

ics.

Ban

kis

the

sum

mat

ion

ofIF

Sline

s22

ath

roug

h22

f.V

alue

sar

e19

80}19

91av

erag

es.G

DP

valu

esar

efrom

Wor

ldBan

kN

atio

nal

Acc

ounts

.

Law

and

orde

rIn

dica

tor

Leg

alE$

cien

cyIn

dica

tor

Com

mon

Law

Dum

my

Cre

ditor

Rig

hts

Inde

x

Shar

ehold

erR

ight

sIn

dex

Mar

ket

Cap

ital

izat

ion/

GD

P

Turn

over

Ban

k/

GD

P

Switze

rlan

d6.

0010

.00

01.

501.

000.

750.

403.

12Ja

pan

5.00

10.0

00

2.25

3.00

0.96

0.52

2.41

Norw

ay6.

0010

.00

02.

203.

000.

180.

411.

50Sw

eden

6.00

10.0

00

2.20

3.00

0.44

0.24

1.37

United

Stat

es6.

0010

.00

11.

005.

000.

600.

580.

96Fin

land

6.00

10.0

00

1.00

2.00

0.18

0.18

1.41

Fra

nce

5.00

8.00

00.

102.

000.

230.

311.

91A

ustr

ia6.

009.

500

3.10

2.00

0.08

0.55

2.26

Net

herlan

ds

6.00

10.0

00

2.00

2.00

0.44

0.39

1.86

Ger

man

y5.

579.

000

3.10

1.00

0.24

1.22

2.07

Can

ada

6.00

9.25

11.

004.

000.

460.

280.

95Bel

gium

6.00

9.50

02.

100.

000.

310.

121.

14It

aly

5.00

6.75

02.

200.

000.

150.

241.

05A

ustr

alia

6.00

10.0

01

1.00

4.00

0.50

0.29

1.01


United

Kin

gdom

4.43

10.0

01

4.00

4.00

0.85

0.38

1.62

New

Zea

land

6.00

10.0

01

3.00

4.00

0.38

0.15

0.88

Singa

pore

5.00

10.0

01

4.00

3.00

0.93

0.32

1.88

Hong

Kon

g4.

7110

.00

14.

004.

001.

190.

40n.

a.Sp

ain

4.00

6.25

02.

202.

000.

220.

301.

80K

ore

a2.

176.

000

3.50

2.00

0.22

0.69

0.92

Mal

aysia

3.86

9.00

14.

003.

000.

790.

151.

54So

uth

Afric

a1.

716.

001

3.00

4.00

1.35

0.05

0.76

Bra

zil

4.00

5.75

01.

203.

000.

110.

480.

45M

exic

o3.

006.

000

0.20

0.00

0.10

0.69

0.41

Turk

ey2.

674.

000

2.20

2.00

0.05

0.08

0.49

Jord

an2.

008.

660

n.a.

1.00

0.47

0.13

1.32

Thai

land

3.57

3.25

13.

103.

000.

180.

591.

19Zim

bab

we

2.00

7.50

14.

003.

000.

100.

080.

33In

dia

1.71

8.00

14.

002.

000.

070.

590.

63Pak

ista

n2.

005.

001

4.00

4.00

0.04

0.11

0.66


reserve, then the size of that reserve as a proportion of assets is added to theindex. The index is presented in Table 2. Scores range from a high of four toa low of 0.1. In addition to using our empirical index of creditor rights, in theregressions below we also test separately for the e!ect of each component of thecreditor-rights index.

We proxy for the rights of shareholders using an index developed by LLSV.This index is scored on a scale of one to "ve. It is obtained by adding a scoreof one for each of the following elements: (i) shareholders are allowed to voteby mail, (ii) shareholders are not required to deposit their shares with atrustee prior to voting, (iii) the law allows cumulative voting for directors, (iv)the law gives minority shareholders special protection, and (v) the mini-mum percentage of share capital that entitles a shareholder to call foran extraordinary general meeting is less than or equal to 10%. Thisindex measures the costs faced by minority investors who want to in#uencedecision-making within the "rm and is presented in Table 2. The index issubject to the same caveats as the index of creditor rights presented earlier.Whether the costs faced by small shareholders when exercising their rights areimportant in determining "rms' "nancial structure decisions will depend onwhether there also exist large investors or "nancial intermediaries who canenforce investor rights. If these large investors exist, then costs faced by smalloutside investors are not necessarily material in determining "nancing patterns.Note that both indexes are additive and linear. Thus, we do not address thepossibility that these factors are not equally important or that they interact ina more complicated way.

4.2.2. Financial institutionsWe measure access to publicly traded equity markets by the ratio of stock

market capitalization to gross domestic product (MCap/GDP). DemirguK c7 -Kuntand Levine (1996) discuss alternative indexes of stock market development. Thestatistics on "nancial markets and intermediaries quoted in this paragraph arecompiled in that paper. Within our sample there is considerable variation in thisratio, ranging from 1.35 in South Africa to 0.04 in Pakistan. Interestingly, insome of the more developed countries, such as Italy, the MCap/GDP is lowerthan is some of the developing countries, such as Malaysia (0.15 compared to0.88, respectively).

In addition to size, we also measure the level of activity in the stock markets ofeach country by the turnover ratio (TOR), computed by dividing the total valuetraded by the market capitalization. Higher values of the turnover ratio indicatea higher level of liquidity. As noted above, a high turnover also increases theincentives for investors to become informed. Thus, a high turnover facilitatesexternal monitoring of corporations. DemirguK c7 -Kunt and Levine (1996) andDemirguK c7 -Kunt and Maksimovic (1996,1998) "nd the turnover ratio to bea good indicator of stock market development.


Access to "nancial intermediaries by "rms is measured by the ratio of thedomestic assets of deposit banks to GDP. Again, there are wide variationsacross countries, both within developed countries (for example, Japan hasa ratio of 2.3 while the U.S. has a ratio of 0.94) and developing countries(compare Malaysia at 1.37 with Turkey at 0.46). Comparable data on other"nancial institutions, such as development banks and "nance companies, isharder to obtain. For 19 countries in our sample, including several developingcountries, we obtain the sum of the deposits of all deposit-taking "nancialinstitutions that do not o!er checking or demand accounts (lines 42a}f fromIMF's International Financial Statistics). The correlation between the size ofthese other "nancial institutions relative to GDP and Bank/GDP for the 19countries in our subsample is 0.24, which is not signi"cant (p"0.32). Thus, thereis no evidence that Bank/GDP is a proxy for other "nancial institutions.

4.3. Firm-specixc characteristics

An important consideration in the choice of "nancial structure by "rms is thereduction of agency costs. The particular types of agency costs to which the "rmis exposed and their magnitude will in general vary from "rm to "rm. Thus, theobserved di!erences in "nancial structures in our sample of countries depend inpart on the characteristics of the population of "rms in each economy. Wecontrol for the di!erences in "rm characteristics between countries by introduc-ing "rm-speci"c variables that are suggested by theory and that are empiricallyuseful in explaining "nancial structure decisions of individual "rms in a subset ofour sample (DemirguK c7 -Kunt and Maksimovic, 1995).

Two of these "rm-speci"c variables are descriptors of the "rm's operatingcharacteristics. The ratio of net "xed assets to total assets, NFA/TA, is anindicator of the structure of the "rm's assets. Fixed assets can be used ascollateral. Thus, "rms with a high ratio of "xed assets should have greaterborrowing capacity. Moreover, since "rms in the U.S. have been found to matchthe maturity of assets with that of liabilities (Stohs and Mauer, 1996), NFA/TAshould be correlated with long-term leverage for "rms in our sample. Oursecond "rm-speci"c variable is a descriptor of the "rm's operating cycle: mea-sured by the ratio of net sales to net "xed assets, NS/NFA. A "rm with highNS/NFA is more likely to need short-term "nancing to support sales. It is likelyto generate short-term assets, such as accounts receivable and notes from itscustomers. Thus, if "rms match the maturity of their assets and liabilities, a highratio of sales to assets will be associated with short-term indebtedness.

Two variables measure the cash constraints of "rms. A high ratio of dividendsto total assets, Div/TA, suggests that the "rm has a cash surplus relative to itsinvestment needs. Such "rms would be expected to reduce their leverage. Thesecond indicator of liquidity is the ratio of earnings before interest and taxes tototal assets. Several studies "nd a strong negative relation between this variable


and leverage, both in the U.S. (e.g., Spence, 1985) and in developing countries(DemirguK c7 -Kunt and Maksimovic, 1995). Firms' capital structures also dependon the tax advantages of debt and equity "nancing. The complexity of taxsystems, with both federal and local taxes, makes it di$cult to compare thebene"ts of debt across a large sample of countries. And e!ective tax rates candi!er signi"cantly from statutory tax rates (see Graham et al., 1996). Theimplications of di!erent tax systems for the composition of debt and for debtmaturity are not clearcut. As a result, we do not include tax variables in ourcross-sectional regressions.

The use of accounting data requires that accounting rules are similar enoughthat the numbers are comparable across countries. Fitzgerald et al. (1979)compare the principal reporting requirements across countries. A direct com-parison across countries of the key requirements shows that the principles onwhich they based are similar enough in stated intent to make them comparable.However, as Ball (1995) points out, countries di!er in the extent to which theiraccounting systems emphasize the importance of public disclosures useful toinvestors. Accounting systems in the Anglo-Saxon common-law tradition tendto emphasize the importance of strict matching of revenues and expenses,whereas systems in the Continental European tradition place a greater stress onconservatism, and allow corporations to smooth pro"ts using hidden reserves.Watts and Zimmerman (1986) describe how accounting systems are shaped bythe market for accounting information and the politics of accounting. Thesedi!erences in the way principles are applied can translate into signi"cantdi!erences in the timing of reported pro"ts, as occurred when Daimler-Benzdecided to report its pro"ts using U.S. rules (Ball, 1995).

Because we use averages over time in forming the variables for our regres-sions, our research design shields us from some of the problems posed by thesedi!erences in emphasis and timing. However, a signi"cant concern in interpret-ing the "nancial statements from the sample of countries pertains to di!erencesarising from di!erent levels of in#ation and the di!erence in how in#ation istreated in "nancial statements. This problem is likely to be most severe forBrazil, Mexico, and Turkey, which, as Table 1 shows, have the highest rates ofin#ation during the sample period. Whereas in most countries in the sample"xed assets are stated at their historical cost, the "nancial statements of "rms inMexico and Brazil are adjusted during part or all of the sample period. Since1984, listed "rms in Mexico have been required to use current replacement costsfor valuing inventories and property, plant, and equipment. Other nonmonetaryassets and stockholders' equity are restated using speci"ed consumer priceindexes. Any gains or losses resulting from in#ation adjustments are reported inthe income statement. The "nancial accounts of Brazilian "rms have beenadjusted for in#ation throughout the sample period, although speci"c require-ments were modi"ed in 1987 and again in 1989. Permanent assets and share-holder equity are adjusted using speci"ed government indexes. As in the case of


Mexico, the adjustment is re#ected in the income statement. However, observersnote that the increases in the speci"ed index do not fully re#ect the realizedin#ation. Turkey, which has the third highest average in#ation rate, 24%, doesnot permit in#ation-adjusted accounting (Price Waterhouse, 1993). The highaverage return on assets reported by Turkish "rms could be caused by thisunderreporting. Care must be exercised in comparing the results for Brazil andTurkey with those of other countries in the sample.

The preliminary evidence presented in Figs. 1}4 suggests that the "nancingdecisions of large and small "rms are di!erently determined. Accordingly, weanalyze them separately in the regressions reported below. For each economywe divide our panel of "rms into quartiles based on asset size. We de"ne as &large"rms' the "rms in the largest quartile in their respective country. These "rms arelikely to have the best access to "nancial markets and institutions in theirrespective countries. In selecting the sample of &small "rms' in each country, weattempt to maintain comparability of "rm sizes across countries. Since what isperceived as a &small "rm' di!ers across countries, to obtain a standardized sizewe divide "rm asset size by the country's GDP. By this measure, the smallest"rms in the dataset for each of the developing countries are of approximatelysimilar size. For these countries, we de"ned as &small "rms' those "rms in thesmallest quartile in their respective countries. For every other country in thesample, &small "rms' are "rms in that quartile of "rms in each country whichmost closely approximates in size the smallest quartile of "rms in the developingcountries, where size is measured by the ratio of "rm's assets to the economy'sGDP. For each country, the "rm-speci"c variables are constructed by takingequally weighted averages of the annual values for the whole period for thesubsamples of large and small "rms separately. Weighting by "rm size does notalter the results.

We present correlation matrices in Table 3. Simple correlations betweencountry means of the leverage variables of interest and the explanatory variablesare shown in Panel A of Table 3 for large and small "rms separately. The twovariables measuring the e!ectiveness of the legal system are signi"cantly corre-lated with all of our "nancial structure variables. The e!ectiveness of the legalsystem is highly correlated with greater reliance on long-term debt and lessreliance on short-term debt. The signs of the correlations between the "nancialstructure variables and per capita GDP parallel those between the "nancialstructure variables and the legal-e!ectiveness variables. However, the correla-tions with the legal-e!ectiveness variables are higher and more statisticallysigni"cant. Of the other legal variables, the most interesting correlations arewith the creditor-rights index. High scores on the index of creditor rights areassociated with a greater reliance on short-term debt over long-term debt andlower absolute levels of the ratio of long-term debt to total debt. This isconsistent with the argument by Diamond (1991) that lenders who engage inmonitoring have an incentive to make short-maturity loans. The correlations


Tab

le3

Corr

elat

ion

mat

rice

s

Thede

pend

entv

aria

bles

are

long

-ter

mde

btto

tota

lass

ets(L

TD

/TA

),sh

ort-te

rmdeb

tto

tota

lass

ets(S

TD

/TA

),an

dlo

ng-

term

debt

toto

tald

ebt(L

TD

/TD

).The

indep

enden

tva

riab

les

arede"

ned

asfo

llow

s:N

FA

/TA

isnet"xe

das

sets

divi

ded

by

tota

lass

ets.

Pro"tis

inco

mebef

orein

tere

stan

dta

xesdi

vide

dby

tota

lass

ets.

NS/

NF

Ais

nets

ales

divi

ded

bynet"xe

das

sets

.D

iv/T

Ais

div

iden

dsdiv

ided

byto

tala

sset

s.G

DP/C

apis

GD

Ppe

rca

pita

.Gro

wth

isth

egr

owth

rate

ofr

ealG

DP

per

capita.

In#at

ion

isth

ein#at

ion

rate

ofth

eG

DP

de#

ator.

TO

Ris

stock

mar

ket

turn

ove

rde"

ned

asth

eto

talva

lue

ofsh

ares

trad

eddiv

ided

by

mar

ket

capital

izat

ion.

MC

ap/G

DP

isth

est

ock

mar

ketca

pital

izat

ion

ofth

eco

untr

ydi

vide

dby

itsG

DP

.Ban

k/G

DP

isth

eto

tala

sset

sof

thedep

osit

money

bank

sdi

vide

dby

GD

P.G

ov.S

ubs./G

DP

areth

egr

ants

on

curr

entac

count

byth

epubl

icau

thor

itie

sto

pri

vate

indu

stries

and

pub

licco

rpora

tions

asw

ellas

togo

vern

men

ten

terp

rise

s,div

ided

by

GD

P.L

awO

rder

,whi

chra

nge

sfrom

zero

tosix,

isan

indi

cato

rofth

ede

gree

tow

hich

citize

nsofa

count

ryar

eab

leto

utili

zeth

eex

isting

lega

lsys

tem

tom

edia

tedi

sput

esan

den

forc

eco

ntra

cts.

Leg

alis

anin

dex

rang

ing

from

zero

tote

nas

sess

ing

the&e$

cien

cyan

din

tegr

ity

oft

hele

gale

nviron

men

tas

ita!

ects

bus

ines

s,par

ticu

larly

fore

ign"rm

s'.F

orbo

thin

dexe

slo

wer

score

sin

dic

ate

low

eren

forc

emen

t/e$

cien

cyle

vels.C

omm

onis

adum

my

variab

leth

ateq

uals

one

forco

mm

on-law

count

ries

and

zero

foroth

ers.

Shar

ehol

der

Rig

htsis

anin

dex

rang

ing

from

zero

to"ve

that

aggr

egat

essh

areh

old

errigh

tsan

dC

redi

torR

ight

sis

anin

dex

rangi

ng

from

zero

to4.

5th

atag

greg

ates

cred

itorrigh

tsas

desc

ribe

din

thete

xt.A

llva

riab

les,

exce

ptth

ela

stfo

ur,

areav

erag

edov

erth

e19

80}19

91pe

riod

when

avai

labl

e,so

that

each

coun

try

has

one

obse

rvat

ion.I

nPan

elA

aver

ages

of"rm

leve

lva

riab

les

(LTD

/TA

,ST

D/T

A,L

TD

/TD

,N

FA

/TA

,Pro"t,

NS/

NFA

,D

iv/T

A)

are

calc

ula

ted

for

the

larg

est

and

smal

lest"rm

sse

para

tely

.Pan

elB

calc

ula

tesav

erag

esusing

all"

rmsin

the

sam

ple.

Cor

rela

tions

repo

rted

are

Pea

rson

corr

elat

ion

coe$

cien

ts.P

-val

ues

are

give

nin

ital

ics.

Num

ber

sof

obse

rvat

ions

are

report

edun

der

the

resp

ective

p-va

lues

Pan

elA

:C

orre

lati

ons

ofD

epen

dent

and

Inde

pend

ent

Var

iabl

es

NFA

/TA

Pro"t

NS/

NFA

Div

/TA

GD

P/

Cap

Gro

wth

In#.

TO

RM

Cap

/G

DP

Ban

k/G

DP

Gov.

Sub./

GD

P

Law

& Ord

er

Leg

alC

om

mon

Shr.

Rig

hts

Srd.

Rig

hts

Larg

ex

rms

LTD

/TA

!0.

127

!0.

132

!0.

269

!0.

437

0.67

6!

0.22

6!

0.37

20.

192

!0.

108

0.29

0!

0.20

50.

639

!0.

460

0.19

40.

061

!0.

337

0.51

10.

488

0.15

80.

018

0.00

00.

230

0.04

30.

308

0.56

90.

127

0.31

60.

000

0.01

10.

304

0.74

80.

073

2930

2929

3030

3030

3029

2630

3030

3029

STD

/TA

!0.

520

0.15

90.

553

0.03

5!

0.09

40.

262

!0.

229

!0.

108

!0.

286

0.07

40.

072

!0.

338

!0.

353

!0.

334

!0.

282

0.16

60.

004

0.40

20.

002

0.85

50.

620

0.16

20.

223

0.56

90.

125

0.70

30.

725

0.06

80.

056

0.07

10.

132

0.39

129

3029

2930

3030

3030

2926

3030

3030

29

LTD

/TD

0.15

2!

0.17

3!

0.47

0!

0.29

00.

547

!0.

284

!0.

214

0.19

30.

141

0.18

6!

0.29

10.

658

0.53

30.

092

0.20

9!

0.34

30.

432

0.36

00.

010

0.12

70.

002

0.12

80.

257

0.30

60.

458

0.33

40.

149

0.00

00.

003

0.62

90.

268

0.06

929

3029

2930

3030

3030

2926

3030

3030

29


Smal

lx

rms

LTD

/TA

!0.

184

!0.

227

!0.

079

!0.

473

0.65

5!

0.14

6!

0.23

10.

215

!0.

205

0.20

8!

0.04

40.

556

0.40

2!

0.30

3!

0.05

2!

0.33

10.

339

0.22

80.

684

0.01

00.

000

0.44

30.

219

0.25

30.

277

0.27

80.

832

0.00

10.

028

0.10

30.

783

0.08

029

3029

2930

3030

3030

2926

3030

3030

29ST

D/T

A!

0.18

60.

414

0.22

10.

023

!0.

320

0.53

9!

0.07

30.

070

!0.

289

!0.

121

0.16

8!

0.44

8!

0.43

6!

0.08

3!

0.10

30.

309

0.33

50.

023

0.24

90.

907

0.08

50.

002

0.70

00.

714

0.12

10.

533

0.41

30.

013

0.01

60.

664

0.58

70.

103

2930

2929

3030

3030

3029

2630

3030

3029

LTD

/TD

!0.

163

!0.

318

!0.

133

!0.

462

0.72

6!

0.25

0!

0.23

20.

211

!0.

099

0.25

1!

0.17

60.

685

0.52

1!

0.19

10.

009

!0.

397

0.39

80.

087

0.49

10.

012

0.00

00.

183

0.21

80.

263

0.60

10.

189

0.39

10.

000

0.00

30.

313

0.96

20.

033

2930

2929

3030

3030

3029

2630

3030

3029

Pan

elB

:C

ross

-Cor

rela

tion

sof

Inde

pend

ent

Var

iabl

es

Pro"t

NS/

NFA

Div

/TA

GD

P/

Cap

Gro

wth

In#.

TO

RM

Cap

/G

DP

Ban

k/

GD

PG

ov.

Sub./

GD

P

LA

Wan

dO

rder

Leg

alC

om

mon

Shr.

Rig

hts

Srd.

Rig

hts

NFA

/TA

0.12

2!

0.61

20.

224

!0.

607

!0.

232

0.63

9!

0.05

0!

0.06

9!

0.62

10.

461

!0.

471

!0.

440

0.17

10.

172

!0.

109

0.52

70.

000

0.25

20.

001

0.22

60.

000

0.79

80.

723

0.00

00.

018

0.01

00.

017

0.37

40.

371

0.58

3

Pro"t

0.10

80.

818

!0.

423

0.03

4!

0.08

0!

0.29

20.

095

!0.

378

!0.

090

!0.

504

!0.

494

0.22

70.

177

0.32

40.

576

0.00

00.

020

0.85

80.

675

0.11

80.

619

0.04

40.

663

0.00

50.

006

0.22

70.

350

0.08

7

NS/

NF

A0.

001

0.21

70.

255

!0.

431

0.10

30.

144

0.28

1!

0.33

20.

035

0.07

50.

219

0.23

40.

231

0.99

80.

258

0.18

20.

020

0.59

50.

457

0.14

70.

098

0.85

80.

698

0.25

30.

222

0.23

7

Div

/TA

!0.

489

!0.

044

!0.

106

!0.

387

0.35

0!

0.34

9!

0.29

9!

0.46

5!

0.35

10.

361

0.20

40.

339

0.00

70.

822

0.58

50.

038

0.06

30.

069

0.14

60.

011

0.06

20.

055

0.28

90.

077

GD

P/C

ap!

0.17

6!

0.32

30.

196

0.17

90.

697

!0.

404

0.87

20.

709

!0.

350

!0.

053

!0.

429

0.35

30.

082

0.30

00.

343

0.00

00.

041

0.00

00.

000

0.05

80.

781

0.02

0

Gro

wth

!0.

089

0.16

40.

004

0.05

40.

142

!0.

084

!0.

255

0.19

40.

176

0.42

60.

641

0.38

70.

983

0.77

90.

489

0.65

80.

173

0.30

50.

051

0.02

1

In#at

ion

0.66

!0.

232

!0.

391

0.64

9!

0.17

9!

0.36

3!

0.19

7!

0.01

3!

0.25

80.

728

0.21

80.

360.

000

0.34

50.

049

0.29

60.

948

0.17

6TO

R!

0.14

00.

262

!0.

013

0.14

00.

025

!0.

211

!0.

187

!0.

040

0.45

90.

170

0.95

00.

460

0.89

40.

263

0.32

20.

835


Tab

le3

(con

tinu

ed)

Pro"t

NS/

NFA

Div

/TA

GD

P/

Cap

Gro

wth

In#.

TO

RM

Cap

/G

DP

Ban

k/

GD

PG

ov.

Sub./

GD

P

LA

Wan

dO

rder

Leg

alC

om

mon

Shr.

Rig

hts

Srd.

Rig

hts

MC

ap/G

DP

0.36

7!

0.43

60.

101

0.41

40.

370

0.41

20.

216

0.50

0.02

60.

595

0.02

30.

044

0.02

40.

259

Ban

k/G

DP

!0.

391

0.53

10.

521

!0.

51!

0.18

9!

0.03

80.

048

0.00

30.

004

0.06

20.

325

0.84

7G

ov.S

ub.

/G

DP

!0.

381

!0.

286

!0.

210

!0.

056

0.14

20.

050.

156

0.30

30.

786

0.48

8

Law

&O

rder

0.69

9!

0.20

60.

036

!0.

432

0.00

00.

275

0.83

70.

019

Leg

al0.

026

0.14

9!

0.06

00.

891

0.43

30.

757


Com

mon

0.72

20.

473

0.00

00.

010

Shar

ehold

erR

ight

0.23

50.

220

Var

iable

de"ni

tions

and

sour

ces

Firm

-lev

eldat

aG

lobal

Van

tage

de"

nitions

:V

aria

bles

are

from

the

indu

stry

/com

mer

cial

tape

ofG

loba

lV

anta

gedat

aba

se,f

roze

nas

ofD

ecem

ber

1995

.LTD

/TA"

(tota

llia

bili

ties!

curr

entlia

bili

ties

)/to

talas

sets"

(DA

118!

DA

104)

/DA

89.

STD

/TA"

curr

ent

liabili

ties

/tot

alas

sets"

DA

104/

DA

89.

LTD

/TD"

(tot

allia

bilit

ies!

curr

ent

liabi

lities

)/to

talliab

ilities"

(DA

118!

DA

104)

/DA

118.

TD

/TA"

tota

lliab

ilities

/tota

las

sets"

DA

118/

DA

89.

NFA

/TA"

net"xe

das

sets

/tot

alas

sets"

DA

76/D

A89

.Pro"t"

(inco

me

befo

rein

com

eta

xes#

inte

rest

expe

nse

s)/tota

las

sets"

(DA

21#

DA

15)/D

A89

.N

S/N

FA"

net

sale

s/ne

t"xe

das

sets"

DA

1/D

A76

.D

iv/T

A"

tota

ldi

vide

nds/

tota

las

sets"

DA

34/D

A89

.For

the

eigh

tco

untr

ies

that

are

from

IFC's

corp

ora

te"nan

ceda

taba

se,va

riab

les

follow

the

de"

nition

sgi

ven

above

.O

ther

data

sour

ces

In#at

ion

isth

ean

nua

lin#at

ion

ofth

eG

DP

de#

ator

obt

ained

from

World

Ban

kN

atio

nal

Acc

ounts

.R

ealG

DP

per

capi

taan

dits

grow

thra

tear

eobt

ained

from

World

Ban

kN

atio

nalA

ccoun

ts.

Ban

k/G

DP

isth

ede

pos

itm

oney

ban

kdo

mes

tic

asse

tsto

GD

P,ob

tain

edfrom

IMF,In

tern

atio

nalFin

anci

alSta

tist

ics,

variou

sye

ars.

Dep

osit

mone

ydom

est

icas

sets

are

the

sum

mat

ion

ofIF

Slin

es22

ath

roug

h22

f.St

ock

mar

ket

variab

les(m

arket

capi

taliz

atio

nan

dtu

rnove

r)ar

efrom

IFC's

Em

ergi

ng

Mar

kets

Dat

aBas

e.The

orig

inal

sourc

efo

rde

velo

ped

coun

trie

sis

Morg

anSta

nley

Cap

ital

Inte

rnat

ional

.The

Law

and

orde

rin

dic

ator

isobt

aine

dfrom

the

Inte

rnat

iona

lC

ountr

yR

isk

Gui

de.

Itha

sbee

nus

edby

Knac

kan

dK

eefe

r(1

995)

.The

Leg

ale$

cien

cyin

dica

tor

isfrom

Busine

ssIn

tern

atio

nalC

orpo

ration.

Oth

erle

galin

dica

tors

(com

mon-

law

dum

my,

cred

itor

righ

ts,an

dsh

areh

old

errigh

tsin

dica

tors

)ar

eobt

aine

dfrom

La

Port

aet

al.(1

998)

.G

over

nm

ent

subsidi

esto

priva

tean

dpu

blic

ente

rprise

sda

taar

eob

tain

edfrom

various

issu

esof

the

World

Com

pet

itiv

enes

sR

eport

,The

World

Eco

nom

icF

oru

m,&

IMD

Inte

rnat

ional

,Gen

eva,

Sw

itze

rlan

d.


between the "nancial structure variables and two other legal variables, the indexof shareholder rights and the dummy for common law, are of smaller magnitude.Finally, correlations involving the institutional and "rm-speci"c control vari-ables show less evidence of statistical signi"cance.

Panel B of Table 3 explores the raw correlations between the explanatoryvariables using data for all "rms in the sample, regardless of size. The legal-e!ectiveness variables are highly positively correlated with income per capita aswell as with the existence of a large banking sector. The relation between thesevariables and the other institutional variables is mixed. However, "rms incountries with an e!ective legal system, as measured by both these variables,tend to have a lower ratio of net "xed assets to total assets, are on average lesspro"table, and pay out lower dividends than "rms in countries with less e!ectivelegal systems.

The three legal variables that measure speci"c characteristics of the legalsystem show fewer signi"cant correlations. As pointed out by LLSV, countrieswith a common-law tradition have better shareholder and creditor rights.However, in these countries the correlation with shareholder rights is stronger,indicating a relative predilection for protecting shareholders. This may be one ofthe explanations for the positive correlation that we observe between thecommon-law dummy variable and the ratio of stock market capitalization toGDP and the negative correlation between the common-law dummy and theratio of bank assets to GDP.

Inspection of the table also reveals that countries with large banking systemstend to have higher ratios of market capitalization to GDP. This "nding hasbeen reported by DemirguK c7 -Kunt and Levine (1996). Large banking systems arealso negatively correlated with in#ation. Finally, it is interesting to note that"rms in countries with larger banking systems have lower ratios of net "xedassets to total assets. This would be consistent with the hypothesis that "nancialintermediaries have a greater willingness to lend against short-term assets,perhaps as a result of their ability to monitor corporations.

Table 4 presents summary statistics for the variables we use in the regressionsreported below.

5. Determinants of 5nancial structure

Di!erences in the legal systems and "nancial institutions between countriesa!ect the borrowing of "rms in developed and developing countries in twoways. First, these di!erences a!ect the absolute levels of long-term andshort-term borrowing. Second, they create incentives to alter the mix oflong-term and short-term debt. Accordingly, we analyze both the ratio oflong-term to total debt and the levels of long-term and short-term debt relativeto total assets.


Table 4Summary statistics

Firm-level variables are reported separately for large and small "rms. These are de"ned as follows:LTD/TA is long-term debt to total assets, STD/TA is short-term debt to total assets, and LTD/TDis long-term debt to total debt. NFA/TA is net "xed assets divided by total assets. Pro"t is incomebefore interest and taxes divided by total assets. NS/NFA is net sales divided by net "xed assets.Div/TA is dividends divided by total assets. Institutional variables are available at the country level.GDP/Cap is the GDP per capita in thousands of US$. Growth is the growth rate of real GDP percapita. In#ation is the in#ation rate of the GDP de#ator. TOR is stock market turnover de"ned asthe total value of shares traded divided by market capitalization. MCap/GDP is the stock marketcapitalization of the country divided by its GDP. Bank/GDP is the total assets of the deposit moneybanks divided by GDP. Gov. Subs./GDP are the grants on current account by the public authoritiesto private industries and public corporations as well as to government enterprises, divided by GDP.Law & Order, which ranges from zero to six, is an indicator of the degree to which citizens ofa country are able to utilize the existing legal system to mediate disputes and enforce contracts. Legalis an index ranging from zero to ten assessing the &e$ciency and integrity of the legal environment asit a!ects business, particularly foreign "rms'. For both indices lower scores indicate lower enforce-ment/e$ciency levels. Common is a dummy variable that equals one for common-law countries andzero for others. Shareholder Rights is an index ranging from zero to "ve that aggregates shareholderrights and Creditor Rights is an index ranging from zero to 4.5 that aggregates creditor rights asdescribed in the text. All variables, except the last four, are averaged over the 1980}1991 period whenavailable, so that each country has one observation

N Mean Std. Dev. Minimum Maximum

¸arge ,rmsLTD/TA 30 0.269 0.103 0.105 0.528STD/TA 30 0.326 0.099 0.170 0.489LTD/TD 30 0.446 0.125 0.248 0.706NFA/TA 29 0.412 0.104 0.227 0.704Pro"t 30 0.100 0.032 0.070 0.202NS/NFA 29 4.687 3.738 0.883 19.632Div/TA 29 0.019 0.012 0.002 0.052

Small xrmsLTD/TA 30 0.175 0.115 0.056 0.500STD/TA 30 0.338 0.088 0.227 0.538LTD/TD 30 0.321 0.146 0.158 0.664NFA/TA 29 0.355 0.088 0.149 0.534Pro"t 30 0.110 0.044 0.048 0.273NS/NFA 29 6.236 3.545 1.436 16.997Div/TA 29 0.028 0.022 0.002 0.088

Economic and institutional variablesGDP/Cap 30 10.234 7.737 0.299 26.348Growth 30 0.024 0.020 !0.023 0.070In#ation 30 0.141 0.341 0.016 1.866TOR 30 0.364 0.248 0.053 1.221MCap/GDP 30 0.419 0.358 0.043 1.351Bank/GDP 29 1.307 0.669 0.325 3.123Gov. Subs./GDP 26 3.130 2.262 0.600 10.686Law & Order 30 4.447 1.597 1.714 6.000Legal 30 8.247 2.069 3.250 10.000Common 30 0.433 0.504 0.000 1.000Shr. Rights 30 2.533 1.332 0.000 5.000Srd. Rights 29 2.453 1.224 0.100 4.000


3We also estimate panel regressions in which each "nancial structure variable of interest isregressed on the explanatory variables and country and year dummies. This speci"cation ispotentially misspeci"ed because it treats cross-sectional and time-series variation equally. Moreover,annual observations may not be independent. However, it has the advantage of using all the data,and was reported in an earlier version of this paper.

Signi"cant changes in the legal systems of countries from year to year are rare,and the indicators of investors' legal protections do not vary over time. Asa result, our investigation of the determinants of "nancial structure reliesprimarily on cross-sectional analysis across countries, taking as our observa-tions the time-series country means of each variable.3

The "rst speci"cation, estimated using ordering least squares (OLS) andWhite's adjustment for heteroskedasticity, is reported in Columns 1 and 4 ofPanels A}C of Table 5. The dependent variables in the three panels are the ratiosof long-term and short-term debt to total assets and long-term debt to totaldebt, respectively. The explanatory variables are the "rm-speci"c characteristicsand descriptors of the legal system and "nancial institutions.

We follow the conventional approach of interpreting a signi"cant positive(negative) coe$cient in the regressions reported below as evidence of a positive(negative) relation between the dependent variable and the corresponding ex-planatory variable. This approach enables us to describe associations within oursample. As pointed out by Barclay et al. (1997), claims that these relations aremore general require speci"c restrictions on the functional form describing howthe "rm's value depends on the explanatory variables.

A potential problem in explaining di!erences in "nancial structures acrossnations by institutional factors is that some of these institutions can themselvesbe in#uenced by "rms' "nancing decisions or by the development of otherinstitutions. At the country level it is unlikely that the major legal and "nancialinstitutions develop completely independently. In such cases, a causal inter-pretation requires a model of dependence between institutions.

In the present context a key concern is whether a country's "nancial institu-tions are an independent determinant of "rms' "nancial structures. While anindividual "rm takes the size of the banking sector as given, that size is a!ectedby the aggregate decisions of all "rms. More importantly, characteristics of thelegal system that facilitate enforceable contracts between "rms and their credi-tors also facilitate contracting between banks and other borrowers.

In our second statistical speci"cation we allow for the possibility that the sizeof a country's banking sector is in#uenced by a country's legal system and the"nancing choices of "rms. We use a two-stage estimator to control for thisdependence. In the "rst stage we obtain a predicted size of the banking sector,given the country's level of development and its legal system. In the second stagewe replace the ratio of the total assets of deposit banks to GDP, Bank/GDP, byits predicted value in the cross-sectional regressions.


Speci"cally, we regress Bank/GDP on the legal-e!ectiveness variables, percapita GDP, and on In#ow, the sum of short-term and long-term capital #owsplus foreign direct investment and portfolio #ows into the country dividedby GDP. The last two variables proxy for the level of development of thecountry and for foreign funds #owing into the country's "nancial system,respectively. Since these variables are not used in the regressions reported below,they also provide identifying restrictions for the coe$cients. The regressioncontrols for potential simultaneity bias when both the "nancial structures of"rms and the size of the banking sector depend on the legal system. As most ofthe descriptors of the legal system are time-invariant, the predicted values ofBank/GDP in the "rst-stage regression are estimated cross-sectionally, using themeans of the time-varying explanatory variables (t-statistics are reported inparentheses):

Bank/GDP"0.32(0.66)

#0.01Law & Order(0.06)

!0.11Shareholder Rights(~1.03)

#0.22Creditor Rights(2.58)

!0.11Common-Law dummy(~0.35)

!0.02Inflow(~0.00)

#0.073Per Capita GDP(2.77)

RM 2"0.57, n"26.

The null hypothesis that the regression has no explanatory power is rejected atless than the 1% level. As expected, more-developed countries have largerbanking sectors. The banking sector is also larger in countries in which creditorrights are better protected. Interestingly, we do not "nd a relation betweenBank/GDP and the e$ciency of the legal system or any of the other descriptorsof the legal system. Of course, causality can also run in the other direction, i.e.,"nancial and non-"nancial "rms in#uence commercial laws, but the incentivesfor such behavior in a general equilibrium context have not been explored andare beyond the scope of this paper.

We also "nd a relation between legal variables and the ratio of "rms' net "xedassets to total assets. (We do not "nd legal variables useful in explaining stockmarket development.) However, replacing the "xed asset ratio by its predictedvalue does not cause us to change any of our interpretations of the legal orinstitutional variables.

5.1. Long-term debt

Panel A of Table 5 presents OLS regressions explaining the ratio of long-termdebt to total assets (LTD/TA) for the large and small quartiles of "rms ineach country over the sample period. Two results pertaining to the legalenvironment are of particular interest. First, reliance on long-term debt by large


Table 5Determinants of debt maturity

Panel A: Long-term debt* The regression equation estimated is: LTD/TA"b0#b

1NFA/TA

#b2

Pro"t#b3

NS/NFA#b4

Growth#b5

TOR#b6

Bank/GDP#b7

Gov. Subs./GDP#

b8

In#ation#b9

Law & Order#b10

Shareholder Rights#b11

Creditor Rights#b12

Common-Law dummy. The dependent variable (LTD/TA) is the long-term debt to total asset ratio. NFA/TAis net "xed assets divided by total assets. Pro"t is income before interest and taxes divided by totalassets. NS/NFA is net sales divided by net "xed assets. Growth is the growth rate of real GDP percapita. TOR is stock market turnover de"ned as the total value of shares traded divided by marketcapitalization. Bank/GDP is the total assets of the deposit money banks divided by GDP. Gov.Subs./GDP are the grants on current account by the public authorities to private industries and publiccorporations as well as to government enterprises, divided by GDP. In#ation is the in#ation rate of theGDP de#ator. Law & Order, which ranges from zero to six, is an indicator of the degree to whichcitizens of a country are able to utilize the existing legal system to mediate disputes and enforcecontracts. Shareholder Rights is an index ranging from zero to "ve that aggregates shareholder rightsand Creditor Rights is an index ranging from zero to 4.5 that aggregates creditor rights as described inthe text. The common-law dummy equals one for common-law countries and zero for others. Inspeci"cations (1) and (2) all variables, except last three, are averaged over the 1980}1991 period so thateach country has one observation. In speci"cation (2) Bank/GDP is replaced with its predicted valuefrom the following regression: Bank/GDP"b#b

1In#ow#b

2GDP/Cap#b

3Law & Order#

b4


Creditor Rights#b6/Common-Law dummy. In#ow is the sum of short-

term and long-term capital #ows plus foreign direct investment and portfolio #ows into the countrydivided by GDP. GDP/Cap is real GDP per capita in US$. For all speci"cations White's hetero-skedasticity-consistent standard errors are given in parentheses

Large "rms Small "rms

(1) (2) (1) (2)

NFA/TA 0.287" 0.358# 0.469" 0.506"

(0.130) (0.118) (0.208) (0.212)

Pro"t 0.281 0.716# 0.189 0.610!(0.336) (0.248) (0.398) (0.315)

NS/NFA !0.002 !0.001 0.007 !0.006(0.002) (0.002) (0.006) (0.006)

Growth !0.434 !0.292 !0.411 !0.923(0.501) (0.668) (0.959) (0.863)

TOR 0.111# 0.094" 0.113! 0.088(0.035) (0.037) (0.062) (0.075)

Bank/GDP !0.026 0.022 !0.047 0.160!

(0.020) (0.052) (0.026) (0.095)

Gov. Subs./GDP 0.019# 0.024# 0.026# 0.035#(0.006) (0.005) (0.007) (0.004)

In#ation !0.275# !0.268# !0.263# !0.233#

(0.028) (0.034) (0.043) (0.057)

Law & Order 0.043# 0.042# 0.052" !0.004(0.010) (0.016) (0.019) (0.028)


Table 5 (continued)


(1) (2) (1) (2)

Shareholder Rights 0.023" 0.012 0.015 0.028(0.009) (0.009) (0.013) (0.019)

Creditor Rights !0.014 !0.023 !0.011 !0.050(0.009) (0.016) (0.015) (0.033)

Common-Law dummy !0.074" !0.033 !0.106# !0.001(0.031) (0.033) (0.041) (0.062)

Adj. R2 0.62 0.62 0.28 0.30No. of observations 26 25 26 25

Panel B: Short-term debt* The regression equation estimated is: STD/TA"TA"b0#

b1

NFA/TA#b2

Pro"t#b3

NS/NFA#b4

Growth#b5

TOR#b6

Bank/GDP#b7

Gov.Subs./GDP#b

8In#ation#b

9Law & Order#b

10Shareholder Rights#b

11Creditor Rights#

b12

Common-Law dummy. The dependent variable (STD/TA) is the short-term debt to total assetratio. NFA/TA is net "xed assets divided by total assets. Pro"t is income before interest and taxesdivided by total assets. NS/NFA is net sales divided by net "xed assets. Growth is the growth rate ofreal GDP per capita. TOR is stock market turnover de"ned as the total value of shares tradeddivided by market capitalization. Bank/GDP is the total assets of the deposit money banks dividedby GDP. Gov. Subs./GDP are the grants on current account by the public authorities to privateindustries and public corporations as well as to government enterprises, divided by GDP. In#ationis the in#ation rate of the GDP de#ator. Law & Order, ranging from zero to six, is an indicator of thedegree to which citizens of a country are able to utilize the existing legal system to mediate disputesand enforce contracts. Shareholder Rights is an index ranging from zero to "ve that aggregatesshareholder rights and Creditor Rights is an index ranging from zero to 4.5 that aggregates creditorrights as described in the text. The common-law dummy equals one for common-law countries andzero for others. In speci"cations (1) and (2) all variables, except the last three, are averaged over the1980}1991 period so that each country has one observation. In speci"cation (2) Bank/GDP isreplaced with its predicted value from the following regression: Bank/GDP"b#b

1In#ow#

b2

GDP/Cap#b3

Law & Order#b4


Creditor Rights#b6

Common-Law dummy. In#ow is the sum of short-term and long-term capital #ows plus foreign directinvestment and portfolio #ows into the country divided by GDP. GDP/Cap is real GDP per capitain US$. For all speci"cations White's heteroskedasticity-consistent standard errors are given inparentheses


(1) (2) (1) (2)

NFA/TA !0.601# !0.502# 0.035 0.207(0.110) (0.090) (0.113) (0.157)

Pro"t 0.738" 1.016# 0.286 0.473(0.365) (0.330) (0.302) (0.304)

NS/NFA 0.005! 0.004! 0.019# 0.018#(0.003) (0.002) (0.005) (0.005)


Table 5 (continued)


(1) (2) (1) (2)

Growth 0.274 0.434 2.838# 2.361#

(0.598) (0.595) (0.638) (0.640)

TOR !0.035 0.046 !0.005 0.006(0.028) (0.030) (0.045) (0.047)

Bank/GDP !0.022 0.058 !0.047" !0.015(0.017) (0.047) (0.013) (0.048)

Gov. Subs./GDP 0.007 0.008 !0.006 !0.001(0.007) (0.006) (0.008) (0.008)

In#ation !0.037 !0.019 0.016 !0.013(0.031) (0.033) (0.032) (0.035)

Law & Order !0.033# !0.051# !0.008 !0.015(0.010) (0.014) (0.008) (0.017)

Shareholder Rights 0.012 0.014 !0.006 0.000(0.008) (0.014) (0.010) (0.014)

Creditor Rights 0.001 !0.013 0.006 0.002(0.007) 0.011 (0.009) (0.014)

Common-Law dummy !0.103" !0.067 !0.068 !0.065(0.036) (0.042) (0.042) (0.052)

Adj. R2 0.75 0.75 0.69 0.64No. of observations 26 25 26 25

Panel C: Debt maturity* The regression equation estimated is: LTD/TD"b0#b

1NFA/TA

#b2

Pro"t#b3

NS/NFA#b4

Growth#b5

TOR#b6

Bank/GDP#b7

Gov. Subs./GDP#

b8

In#ation#b9

Law & Order#b10


Creditor Rights#b12

Common-Law dummy. The dependent variable (LTD/TD) is the long-term debt to total debt ratio. NFA/TA isnet "xed assets divided by total assets. Pro"t is income before interest and taxes divided by total assets.NS/NFA is net sales divided by net "xed assets. Growth is the growth rate of real GDP per capita.TOR is stock market turnover de"ned as the total value of shares traded divided by marketcapitalization. Bank/GDP is the total assets of the deposit money banks divided by GDP. Gov.Subs./GDP are the grants on current account by the public authorities to private industries and publiccorporations as well as to government enterprises, divided by GDP. In#ation is the in#ation rate of theGDP de#ator. Law & Order, ranging from zero to six, is an indicator of the degree to which citizens ofa country are able to utilize the existing legal system to mediate disputes and enforce contracts.Shareholder Rights is an index ranging from zero to "ve that aggregates shareholder rights andCreditor Rights is an index ranging from zero to 4.5 that aggregates creditor rights as described in thetext. The common-law dummy equals one for common-law countries and zero for others. Allvariables, except last three, are averaged over the 1980}1991 period so that each country has oneobservation. In speci"cation (2) Bank/GDP is replaced with its predicted value from the followingregression Bank/GDP"b#b

1In#ow#b

2GDP/Cap#b

3Law and Order# b

4Shareholder

Rights#b5

Creditor Rights#b6

Common-Law dummy. In#ow is the sum of short-term and long-term capital #ows plus foreign direct investment and portfolio #ows into the country divided by GDP.GDP/Cap is real GDP per capita in US$. For both speci"cations White's heteroskedasticity-consistent standard errors are given in parentheses


"rms is clearly higher in countries with an e!ective legal system. In the reportedequations the e$ciency of the legal system is measured by the InternationalCountry Risk Guide index. The Business International Corporation index is notused since this index is targeted towards foreign "rms; its coe$cient is of thesame sign, although not always signi"cant (and if so, at a lower level).The magnitude of the coe$cient on legal e!ectiveness in the cross-sectional

Table 5 (continued)


(1) (2) (1) (2)

NFA/TA 0.603# 0.592# 0.420 0.373(0.116) (0.101) (0.257) (0.237)

Pro"t !0.610 !0.351 0.201 0.629(0.449) (0.379) (0.482) (0.423)

NS/NFA !0.005 !0.005 0.000 !0.015!(0.003) (0.003) (0.007) (0.008)

Growth !0.417 !0.562 !1.491 !1.876"(0.687) (0.696) (1.301) (0.965)

TOR 0.108" 0.105" 0.154! 0.116(0.039) (0.039) (0.079) (0.093)

Bank/GDP !0.022 !0.034 !0.033 0.209"(0.018) (0.046) (0.029) (0.104)

Gov. Subs./GDP 0.004 0.009 0.030# 0.038#(0.008) (0.006) (0.008) (0.007)

In#ation !0.210# !0.222# !0.283# !0.229#(0.062) (0.040) (0.044) (0.061)

Law & Order 0.055# 0.067# 0.069" 0.002(0.011) (0.019) (0.024) (0.031)

Shareholder Rights 0.012 0.001 0.014 0.026(0.010) (0.012) (0.017) (0.022)

Creditor Rights !0.013 !0.011 !0.021 !0.068"(0.011) (0.015) (0.020) (0.025)

Common-Law dummy 0.015 0.026 !0.041 0.091(0.035) (0.036) (0.044) (0.072)

Adj. R2 0.73 0.73 0.39 0.47No. of Observations 26 25 26 25

!indicate signi"cance levels of 10%."indicate signi"cance levels of 5%.#indicate s igni"cance levels of 1%.


regressions suggests that the marginal e!ect of the di!erence between a verye$cient legal system (e.g., Switzerland) and an ine$cient one (e.g., Turkey) is toincrease the ratio of large "rms' long-term assets to total assets by 0.25. Forsmall "rms the coe$cient is only signi"cant in the OLS speci"cation. Second,there is only limited evidence that indexes of speci"c protections or common-lawtraditions directly a!ect a "rm's long-term indebtedness. Thus, while both largeand small "rms in common-law countries have less long-term debt, this e!ectdisappears when we adjust for the endogeneity of the banking system. Thissuggests that the legal tradition a!ects "nancial structures through its e!ect oninstitutions. Interestingly, high values of the index of creditor rights are notcorrelated with the use of long-term debt. We explore the role of creditor rightsin more detail below.

Of the "nancial institution variables, for large "rms the only signi"cantcoe$cient is that of the stock market's turnover ratio, TOR. Thus, for large"rms an active stock market is associated with higher long-term leverage.However, the marginal e!ect of the variation of stock market activity on thelong-term debt of large "rms is relatively smaller than that of the e$ciency of thelegal system. Comparing countries with relatively high turnover (such as theU.S. or Thailand) and countries with low turnover (such as Pakistan) indicatesthat the higher turnover in the former increases long-term debt by about5 percentage points. The stock market size variable, MCap/GDP, is not in-cluded in the reported speci"cations since it is not signi"cant, consistent with theresults in DemirguK c7 -Kunt and Maksimovic (1998).

The coe$cient of the size of the banking sector is not signi"cant in thelarge-"rm equations. By contrast, for small "rms the coe$cients of the bankingvariable exhibit an interesting pattern. The size of the banking sector by itself isnot signi"cant. However, when the predicted values of Bank/GDP are usedinstead of actual values in Column 5, the coe$cient changes sign and issigni"cant at the 10% level. Since creditor rights are an important determinantof the predicted value of Bank/GDP, this provides some evidence that strongrights promote access to long-term credit for small "rms, albeit indirectly byincreasing the size of the banking system. This is consistent with the hypothesisthat small "rms are &marginal' and that their long-term borrowing is correlatedwith year-to-year changes in the size of the banking sector. Large "rms are likelyto be inframarginal borrowers, whose needs are satis"ed even when the bankingsystem is undeveloped.

Four of the control variables are also signi"cant for both large and small"rms. Extensive government subsidies are associated with a high ratioof long-term debt to assets. High average rates of in#ation are negativelyassociated with the use of long-term debt for both large and small "rms. Anincrease of 5% in the in#ation rate reduces the ratio of long-term to total assetsby approximately 1.25 percentage points. High average ratios of net "xed assetsto total assets are associated with a higher ratio of long-term debt to total


assets. This is consistent with the notion that "xed assets serve as good collateralfor long-term debt. Pro"table "rms have more long-term debt. The ratio ofdividends to total assets is not signi"cant when entered into the regressionequation, and we drop it from subsequent regressions. An additional variable,the mean standard deviation of "rm-level cash #ows in each country, is alsoinsigni"cant and are not reported.

Taken together, the results suggest that for large "rms the observed variationin the levels of long-term debt across countries is related to the e!ectiveness ofthe legal system and the liquidity of the stock market. For all "rms, highpro"tability, the availability of collateral, a low rate of in#ation, and a high levelof government subsidies are associated with long-term "nancing. There is someevidence that increased creditor protection increases the amount of long-termdebt used by large "rms indirectly by increasing the size of the banking sector.

We also estimate these equations using a panel that allows for both time-series and cross-sectional variation. The results suggest that year-to-year with-in-country variations in the explanatory variables a!ect large and small "rmsdi!erently. Within countries, changes in the use of long-term debt by large "rmsare related to changes in the e$ciency of the legal system, the level of govern-ment subsidies, and the activity level in the stock market. These factors are notassociated with changes in the use of long-term debt by small "rms in the panelregressions. A likely explanation for this di!erence is that small "rms, with theirreduced access to "nancial markets, governmental subsidies, and the legalsystem, are less likely to be a!ected by marginal improvements in "nancialmarkets and the legal system and changes in the level of subsidies. This isconsistent with the additional "nding that within-country variation in long-termborrowing by small "rms is more strongly related to yearly changes in the size ofthe banking sector. By contrast, the cross-sectional di!erences between coun-tries that we measure in the paper are likely to be of a greater magnitude. Thus,they a!ect small "rms as well as large "rms.

5.2. Short-term debt

Panel B of Table 5 examines cross-country variation in the use of short-termdebt. For large "rms, the coe$cients of the legal-e$ciency variables and thecommon-law dummy are signi"cant. Large "rms in countries with more e!ectivelegal systems have less short-term debt. This e!ect is economically signi"cant andis of the same magnitude as the positive e!ect of legal e$ciency on long-term debt.However, we do not "nd that the use of short-term debt by small "rms is a!ectedeither by the e!ectiveness of the legal system or by whether or not the legal systemis based on common law. The other indexes of shareholder or creditor rights donot help explain the use of short-term debt for either large or small "rms.

Only one of the coe$cients of "nancial system variables is statisticallysigni"cant. Small "rms in countries with a large banking sector have less


short-term debt. However, this e!ect disappears when the Bank/GDP variable isreplaced by its predicted value. Cross-country variation in the stock market or invalues of the indexes of investor protection do not help explain di!erences in theusage of short-term debt. Similarly, the indicator of government subsidies asa fraction of the GDP does not improve the explanatory power of the regressions.

Inspection of Panel B reveals that several "rm-speci"c variables help explainshort-term leverage. One of these variables, the ratio of net sales to net "xedassets, is positively associated with short-term borrowing for both small andlarge "rms. A high ratio of net "xed assets to total assets is associated with lowerlevels of short-term borrowing for large "rms. This is consistent with the notionthat such "rms more easily match the maturity of borrowing with the maturityof their assets. Thus, large "rms with "xed assets prefer long-term over short-term borrowing. Interestingly, more-pro"table large "rms have more short-termdebt, whereas we do not obtain a similar signi"cant relation for small "rms.Small "rms that grow fast rely more heavily on short-term debt. This "nding isconsistent with Myers' (1977) hypothesis that growth options are not "nancedusing long-term debt.

In unreported regressions we also regress the ratio of total debt to total assetson our explanatory variables. For large "rms, the total amount of debt does notvary with the e$ciency of the legal system, but the coe$cients on the share-holder rights index, stock market turnover, Bank/GDP, "rm pro"tability, andof the ratio of government subsidies to GDP are positive and signi"cant. Thecoe$cients on the common-law dummy and in#ation are negative and signi"-cant. For small "rms the results are less clear: the coe$cient on the rate ofin#ation is negative and signi"cant and the coe$cient of the ratio of governmentsubsidies to GDP is positive and signi"cant.

5.3. Debt maturity

We explore how the maturity of the "rm's liabilities varies across countries inPanel C. As our indicator of maturity, we use the ratio of long-term liabilities tototal liabilities. The speci"cations parallel the regressions reported in PanelsA and B.

The results of cross-country regressions of debt maturity on "rm-speci"c andinstitutional variables are consistent with the results for long-term and short-term debt reported above. We "nd evidence that the higher the quality of legalinstitutions, the greater the proportion of long-term "nancing, particularly inthe case of large "rms. The indicator of the legal system's e$ciency in settlingdisputes is positive and signi"cant at the 1% level in the large-"rm equation.The predicted ratio of long-term to total debt is higher by about 0.35}0.4 incountries with the most e$cient legal systems than in countries with the leaste$cient systems. There is less evidence that legal e$ciency is important for small"rms * the coe$cient is positive signi"cant at the 5% level in the small-"rm


OLS equation, but the e!ect disappears when predicted values of Bank/GDPare used.

Consistent with the previously reported results, large "rms in countries withactive stock markets have longer debt maturities. There is less evidence that thelevel of market activity is related to the debt maturity of smaller "rms. Similarly,the evidence on the e!ect of the banking system on the debt maturity of small"rms is mixed. When the predicted values of Bank/GDP are used as anexplanatory variable, the debt maturity of small "rms in countries is positivelyrelated to the size of the banking sector. The predicted ratio of long-term debt tototal debt increases by approximately 30% in a country whose banking systemis in the lower end of the range (say, New Zealand) compared to a country at thehigh end (Japan). Together with the "nding that small "rms in countries withlarge banking systems have less short-term debt, this "nding suggests thata large banking sector enables small "rms to extend the maturity of their debt.However, the coe$cient of Bank/GDP is signi"cant at the 5% level in only oneof the two speci"cations.

The coe$cients of several of the control variables are of interest. The coe$c-ient of in#ation is negative and highly signi"cant for both large and small "rms.High ratios of net "xed assets to total assets are positively associated with longerdebt maturities for large "rms. High levels of government subsidies are posit-ively associated with longer debt maturities for small "rms.

5.4. Specixc legal protections

In Table 5 we "nd no evidence that the index of creditor rights helps predicteither short-term or long-term leverage or debt maturity. In this subsection weexplore the relation between creditor rights and debt levels in more detail. First,we replace the creditor-rights index by all its individual components in thelarge-"rm and small-"rm cross-sectional equations in Panels A}C of Table 5.Speci"cation (1) of Table 6 show the coe$cients of the individual components ofthe index in these equations for large and small "rms, respectively. In Speci"ca-tion (2) we replace the creditor-rights index by each component in turn. Finally,in Speci"cation (3) the index is replaced by each component in turn, but this timeBank/GDP is replaced by its predicted value.

Inspection of Table 6 reveals that the variation in most of the components ofthe creditor-rights index in this sample is not signi"cantly related to the debtcomposition of large "rms. The one exception is variation in the right of securedcreditors to be paid "rst in bankruptcy. This right is associated with increasedshort-term borrowing and a signi"cantly shorter maturity of debt for large"rms. The interpretation of a signi"cant partial correlation between any singlelegal protection and the use of a "nancial contract must be tentative because theimportance of the protection varies in di!erent systems. Moreover, because weare estimating a series of regressions, the statistical signi"cance of any single


Tab

le6

Impa

ctof

cred

itor

righ

ts

There

gres

sion

equa

tion

estim

ated

issp

eci"

cation

(1)f

rom

Tab

le5,

Pan

elsA}C

.In

(1)t

heag

greg

atecr

editorrigh

tsva

riab

leis

repl

aced

byal

lthe

indi

vidu

alin

dic

ators

,(2)

ente

rsea

chin

dica

torone

ata

tim

e.C

redi

torrigh

tva

riab

lesar

ere

strict

ions

forgo

ing

into

reorg

aniz

atio

n,au

tom

atic

stay

on

asse

ts,s

ecure

dcr

editors

paid"rs

t,m

anag

emen

tst

ays

inre

org

aniz

atio

n,an

dle

gal

rese

rve

requi

rem

ent

asa

per

cent

age

ofca

pital

.Ent

ries

are

resp

ective

coe$

cien

tes

tim

ates

forth

ein

div

idual

lega

lindic

ator

obta

ined

from

the

bas

iceq

uat

ion.I

nsp

eci"

cation

(3),

indi

vidu

alrigh

tsar

eag

ain

ente

red

one

ata

tim

e,bu

tth

ere

gres

sion

isin

two

stag

esw

ith

Ban

k/G

DP

repl

aced

by

its

pre

dict

edva

lue

obt

aine

dby

regr

essing

iton

In#ow

,G

DP

/Cap

,Law

&O

rder

,Sha

reho

lder

Rig

hts,

Com

mon

-Law

dum

my,

and

the

cred

itor

righ

tva

riab

leen

tere

din

the

seco

nd

stag

e.In

allsp

eci"

cation

sW

hite's

het

erosk

edas

tici

ty-c

ons

iste

ntst

anda

rder

rors

are

give

nin

par

enth

eses

Lar

ge"rm

sSm

all"rm

s

(1)

(2)

(3)

(1)

(2)

(3)

LTD

/TA

Res

tric

tions

for

goin

gin

tore

org

aniz

atio

n!

0.01

8!

0.02

7!

0.05

4"!

0.05

0!

0.00

1!

0.00

6(0

.036

)(0

.025

)(0

.026

)(0

.057

)(0

.037

)(0

.054

)

Aut

om

atic

stay

onas

sets

!0.

009

0.00

10.

011

0.08

3"0.

068"

0.11

3"(0

.027

)(0

.021

)(0

.025

)(0

.036

)(0

.028

)(0

.040

)

Sec

ured

cred

itors"rs

tpai

d!

0.03

9!

0.03

9!

0.06

90.

095

0.07

10.

044

(0.0

36)

(0.0

28)

(0.0

48)

(0.0

76)

(0.0

66)

(0.1

00)

Man

agem

ent

stay

sin

reorg

aniz

atio

n0.

029

0.02

20.

052

!0.

106

0.00

90.

173"

(0.0

37)

(0.0

36)

(0.0

65)

(0.0

72)

(0.0

51)

(0.0

70)

Leg

alre

serv

ere

quired

as%

ofca

pita

l0.

020

0.04

6!

0.08

2!

0.30

3!

0.15

8!

0.74

8#(0

.143

)(0

.118

)(0

.170

)(0

.248

)(0

.210

)(0

.192

)

STD

/TA

Res

tric

tions

for

goin

gin

tore

org

aniz

atio

n!

0.01

1!

0.00

10.

005

0.00

3!

0.01

50.

013

(0.0

27)

(0.0

02)

(0.0

19)

(0.0

26)

(0.0

16)

(0.0

22)

Aut

om

atic

stay

onas

sets

0.00

60.

007

0.03

30.

040!

0.01

20.

017

(0.0

21)

(0.0

16)

(0.0

24)

(0.0

21)

(0.0

24)

(0.0

33)


Secu

red

cred

itors"rs

tpa

id0.

044!

0.03

9!!

0.01

6!

0.02

5!

0.04

0!

0.04

3(0

.024

)(0

.023

)(0

.038

)(0

.026

)(0

.026

)(0

.036

)

Man

agem

entst

ays

inre

org

aniz

atio

n!

0.01

80.

003

0.04

7!

0.07

2#!

0.06

3#!

0.04

7(0

.032

)(0

.022

)(0

.040

)(0

.036

)(0

.019

)(0

.037

)

Leg

alre

serv

ere

quired

as%

ofca

pital

!0.

034

!0.

017

!0.

152

0.01

50.

046

0.00

1(0

.079

)(0

.057

)(0

.103

)(0

.102

)(0

.075

)(0

.117

)

LTD

/TD

Res

tric

tions

for

goin

gin

tore

org

aniz

atio

n0.

004

!0.

019

!0.

049

!0.

025

!0.

006

!0.

011

(0.0

38)

(0.0

29)

(0.0

32)

(0.0

70)

(0.0

63)

(0.0

58)

Auto

mat

icst

ayon

asse

ts!

0.01

8!

0.00

4!

0.01

10.

087!

0.08

0"0.

143#

(0.0

24)

(0.0

20)

(0.0

37)

(0.0

45)

(0.0

37)

(0.0

49)

Secu

red

cred

itors"rs

tpa

id!

0.07

1"!

0.05

8"!

0.01

30.

085

0.09

40.

044

(0.0

33)

(0.0

23)

(0.0

61)

(0.0

93)

(0.0

93)

(0.1

18)

Man

agem

entst

ays

inre

org

aniz

atio

n0.

065

0.03

2!

0.00

2!

0.03

50.

061

0.25

3#(0

.037

)(0

.037

)(0

.054

)(0

.087

)(0

.059

)(0

.076

)

Leg

alre

serv

ere

quired

as%

ofca

pita

l0.

086

0.07

20.

067

!0.

219

!0.

132

!0.

810#

(0.1

32)

(0.1

09)

(0.1

35)

(0.2

93)

(0.2

35)

(0.1

89)

! indic

ate

sign

i"ca

nce

leve

lsof

10%

."in

dica

tesign

i"ca

nce

leve

lsof5%

.# indic

ate

sign

i"ca

nce

leve

lsof

1%.


coe$cient can be overstated. However, the signs of the coe$cients of the right ofsecured creditors to be paid "rst in bankruptcy are consistent with the proposi-tion that short-term lenders have a greater incentive to monitor borrowers andbene"t most from an ability to repossess secured assets, as suggested by Dia-mond (1991). Interestingly, the rights of secured creditors are not similarlycorrelated with the "nancing decisions of small "rms. In the U.S., trade creditorspossess rights that other secured creditors do not possess. Thus, systematicdi!erences in the amount of trade credit used by large and small "rms wouldconfound our results.

Table 6 also provides evidence that small "rms in countries with automaticstays on the assets of bankrupt "rms have more long-term debt and a higherratio of long-term debt to short-term debt. There is no evidence that these "rmsare substituting long-term for short-term debt. Again, an interpretation must betentative. In bankruptcy, an automatic stay bene"ts borrowers at the expense ofsecured creditors. As a result, small "rms might wish to borrow more if theautomatic stay credibly commits lenders not to defraud them in the event thatthey become "nancially distressed.

6. Conclusions

We examine the maturity of "rms' liabilities in 30 developed and developingcountries during the period 1980}1991. We "nd systematic di!erences in the useof long-term debt between developed and developing countries and small andlarge "rms. In developed countries, "rms have more long-term debt anda greater proportion of their total debt is held as long-term debt. This is trueregardless of "rm size across our sample of countries. This di!erence cannot beexplained by matching of maturities of assets and liabilities as "rms in develop-ing countries have higher proportions of net "xed assets to total assets. Also,large "rms have more long-term debt as a proportion of total assets and debtcompared to smaller "rms. We attempt to explain the observed cross-countryvariation in leverage and maturity of liabilities by di!erences in the legalsystems, "nancial institutions, and government subsidies, as well as by "rmcharacteristics and macroeconomic factors (such as the rate of in#ation and theeconomy's growth rate).

We "nd strong evidence that large "rms in countries with e!ective legalsystems have more long-term debt relative to assets, and their debt is of longermaturity. Large "rms in countries with e!ective legal systems have lowershort-term liabilities, suggesting that such "rms are substituting long-term debtfor short-term debt. For small "rms, evidence of a relation between the e!ec-tiveness of the legal system and the ratio of long-term debt to assets is weaker.We also do not "nd evidence of lower short-term liabilities by small "rms incountries with more-e!ective legal systems, perhaps because small "rms tend to


use less long-term debt than do large "rms. These conclusions are consistentwith the "ndings of DemirguK c7 -Kunt and Maksimovic (1998) that a higherproportion of "rms in countries with e!ective legal systems "nance their growthexternally.

We also test the hypothesis that the tradition on which a country's legalsystem is based in#uences the optimal "nancing of "rms in that country. We "ndlimited evidence that both large and small "rms in countries with a common-lawtradition use less long-term debt, relative to their assets, than "rms in countrieswith a civil-law tradition. There is also limited evidence that large "rms incommon-law countries use less short-term debt. Debt maturity does not di!erbetween common-law and civil-law countries.

The structure of "nancial institutions is also an important determinant of"rms' "nancing choices. Consistent with DemirguK c7 -Kunt and Maksimovic's(1998) results on external "nancing of investment, we "nd that whereas thevariation in the size of the stock market relative to the country's economy is notcorrelated with "nancing patterns, variation in the level of activity of the stockmarket does have explanatory power for large "rms. In countries with activestock markets, large "rms have more long-term debt and debt of longer matur-ity. Neither the level of activity nor the size of the market is correlated with the"nancing choices of small "rms. By contrast, in countries with a large bankingsector, small "rms have less short-term debt and their debt is of longer maturity.Variation in the size of the banking sector does not have a correspondingcorrelation with the capital structures of large "rms. At the margin, largebanking sectors enable smaller "rms to substitute long-term debt for short-termdebt.

We also "nd that the magnitude of government subsidies to industry ispositively related to the use of long-term debt by both large and small "rms,perhaps as a result of implicit guarantees. In#ation is negatively related to theuse of long-term debt. Variation in several "rm-speci"c characteristics is alsorelated to the use of debt of di!erent maturities. In particular, high ratios of net"xed assets to total assets are positively related to the use of long-term debt byboth large and small "rms and to longer debt maturities for large "rms only.This "nding suggests that large "rms can more easily use their "xed assets toobtain long-term debt. By contrast, high ratios of sales to "xed assets areassociated with more extensive use of short-term debt. This would be consistentwith maturity matching if "rms with high ratios of sales to "xed assets also havehigh ratios of accounts receivable to "xed assets.

In sum, the underlying legal and institutional di!erences explain a largeportion of the variation in the use of long-term debt. While we have identi"edrelations between "nancial institutions and legal system origin and e$ciency, onthe one hand, and "nancial structures of "rms on the other, we have not beenable to consistently relate speci"c investor protections with "rm "nancing. Thisis not surprising, since the constraints that speci"c features of the legal system


impose on contracting by "rms and investors depend on the characteristics ofthe "nancial system in each country. The exact way in which this happens is anopen research question.

Our paper has several policy implications. First, it provides evidence that"rms in developing countries have less long-term debt, even after accounting fortheir characteristics. Second, the paper shows that this lack of term "nancing ismainly due to institutional di!erences, such as the extent of government subsi-dies, di!erences in the levels of development of stock markets and banks, anddi!erences in the underlying legal infrastructure. Third, the results indicate thatwhile policies that help develop countries' legal and "nancial infrastructuresmight be e!ective in increasing the access of "rms to long-term debt, di!erentpolicies would be necessary to lengthen the debt maturity of large and small"rms. Improvements in legal e!ectiveness seem to bene"t all "rms, although thisresult is much less signi"cant for the smallest "rms, which have limited access tothe legal system. Similarly, policies that would help improve the functioning andliquidity of stock markets would again bene"t mostly large "rms. In contrast,policies that would lead to improvements in the development of the bankingsystem would improve the access of smaller "rms to long-term credit.

Appendix A

Number of "rms and the sample period

No. of "rms Time period

Australia 401 1983}1991Austria 44 1983}1991Belgium 89 1983}1991Brazil! 100 1985}1991Canada 494 1983}1991Switzerland 150 1983}1991Germany 359 1983}1991Spain 116 1983}1991Finland 55 1983}1991France 544 1983}1991United Kingdom 1275 1983}1991Hong Kong 173 1983}1991India! 100 1980}1990Italy 81 1983}1991Jordan! 38 1980}1990Japan 1104 1983}1991Korea! 100 1980}1990Mexico! 100 1984}1991Malaysia 143 1983}1991


No. of "rms Time period

Netherlands 165 1983}1991Norway 52 1983}1991New Zealand 41 1983}1991Pakistan! 100 1980}1988Singapore 213 1983}1991Sweden 68 1983}1991Thailand 137 1983}1991Turkey! 45 1982}1990United States 3247 1983}1991South Africa 67 1983}1991Zimbabwe! 48 1980}1988

!The data source for the "rm level variables is IFC's corporate inance data base. Otherwise, the dataare from Global Vantage data base.

References

Ball, R., 1995. Making accounting more international. Journal of Applied Corporate Finance 8,19}29.

Barclay, M.J., Marx, L.M., Smith Jr., C.W., 1997. Leverage and maturity as strategic complements.Unpublished Working Paper, University of Rochester.

Barclay, M.J., Smith Jr., C.W., 1995. The maturity of corporate debt. Journal of Finance 50, 609}631.Bolton, P., Scharfstein, D.S., 1993. Optimal debt structure with multiple creditors. LSE Capital

Markets Group Discussion Paper No. 161.Calomiris, C., 1993. Corporate "nance bene"ts from universal banking: Germany and the United

States, 1870}1914. NBER Working Paper No. 4408.DemirguK c7 -Kunt, A., Levine, R., 1996. Stock market development and "nancial intermediaries:

stylized facts. World Bank Economic Review 10, 291}321.DemirguK c7 -Kunt, A., Maksimovic, V., 1995. Capital structures in developing countries: evidence from

ten country cases. Policy Research Working Paper 1320, The World Bank.DemirguK c7 -Kunt, A., Maksimovic, V., 1996. Stock market development and "rm "nancing choices.

World Bank Economic Review 10, 341}369.DemirguK c7 -Kunt, A., Maksimovic, V., 1998. Law, "nance, and "rm growth. Journal of Finance 53,

2107}2137.Diamond, D.W., 1984. Financial intermediation and delegated monitoring. Review of Economic

Studies 51, 393}414.Diamond, D.W., 1991. Debt maturity and liquidity risk. Quarterly Journal of Economics 106,

709}737.Diamond, D.W., 1993. Seniority and maturity of debt contracts. Journal of Financial Economics 33,

341}368.Fitzgerald, R.D., Stickler, A.D., Watts, T.R., 1979. International survey of accounting principles and

reporting practices. Price Waterhouse International, distributed by Butterworths, Ontario,Canada.

Graham, J., Lemmon, M., Schallheim, J., 1996. Debt, leases and the endogeneity of corporate taxstatus. Unpublished Working Paper, University of Utah.

Grossman, S.J., 1976. On the e$ciency of competitive stock markets where trades have diverseinformation. Journal of Finance 31, 573}585.


Grossman, S.J., Stiglitz, J.E., 1980. On the impossibility of informationally e$cient markets.American Economic Review 70, 393}408.

Harris, M., Raviv, A., 1990. Capital structure and the informational role of debt. Journal of Finance45, 321}349.

Hart, O., Moore, J., 1995. Debt and seniority: an analysis of hard claims in constraining manage-ment. American Economic Review 85, 567}587.

Hauswald, R.B.H., 1996. Banking systems, bankruptcy arrangements and institutional comp-lementarity. Unpublished Working Paper, University of Maryland.

Hoshi, Kashyap, Scharfstein, D.S., 1990. The role of banks in reducing the costs of "nancial distressin Japan. Journal of Financial Economics 27, 67}88.

Jensen, M.C., Meckling, W., 1976. Theory of the "rm: managerial behavior, agency costs, and capitalstructure. Journal of Financial Economics 3, 305}360.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1998. Law and "nance. Journal ofPolitical Economy 106, 1113}1155.

Myers, S.C., 1977. Determinants of corporate borrowing. Journal of Financial Economics 5,147}175.

Myers, S.C., Majluf, N.S., 1984. Corporate "nancing and investment decisions when "rms haveinformation that investors do not have. Journal of Financial Economics 13, 187}221.

Price Waterhouse, 1993. Doing Business in Turkey. Price Waterhouse, Istanbul, Turkey.Rajan, R.G., 1992. Insiders and outsiders: the choice between informed and arm's length debt.

Journal of Finance 47, 1367}1400.Rajan, R., Zingales, L., 1995. What do we know about optimal capital structure? Some evidence

from international data. Journal of Finance 50, 1421}1460.Rajan, R., Zingales, L., 1998. Financial dependence and growth. American Economic Review 88,

559}587.Ravid, S.A., 1996. Debt maturity * a survey. Financial Markets, Institutions and Instruments 5,

1}69.Singh, A., Hamid, J., Salimi, B., Nakano, Y., 1992. Corporate "nancial structures in developing

countries. Technical paper, No. 1, International Finance Corporation.Smith Jr., C.W., Warner, J.B., 1979. On "nancial contracting: an analysis of bond covenants. Journal

of Financial Economics 7, 117}161.Smith Jr., C.W., Watts, R.L., 1992. The investment opportunity set and corporate "nancing,

dividend and compensation policies. Journal of Financial Economics 32, 263}292.Spence, A.M., 1985. Capital structure and the corporation's product market environment. In:

Friedman, B.M. (Ed.), Crporate Capital Structures in the United States, National Bureau ofEconomic Research Project Report Series. University of Chicago Press, Chicago, IL.

Stohs, M.H., Mauer, D.C., 1996. The determinants of corporate debt maturity structure. Journal ofBusiness 69, 279}312.

Watson, A., 1974. Legal Transplants. University of Virginia Press, Charlottesville, VA.Watts, R.L., Zimmerman, J.L., 1986. Positive Accounting Theory. Prentice-Hall, Englewood Cli!s,

NJ.


institutions, financial markets, and firm debt maturity

Documents