emu and the cost of capital - home - bbc...

EMU study

EMU and the cost of capital

EMU and the cost of capitalEMU study

This study has been prepared by HM Treasury toinform the assessment of the five economic tests

© Crown copyright 2003

The text in this document (excluding the Royal Coat of Arms and departmental logos)may be reproduced free of charge in any format or medium providing that it isreproduced accurately and not used in a misleading context. The material must beacknowledged as Crown copyright and the title of the document specified.

Any enquiries relating to the copyright in this document should be sent to:

HMSOLicensing DivisionSt Clements House2-16 ColegateNorwichNR3 1BQ

Fax: 01603 723000

E-mail: [email protected]

Printed by the Stationery Office 2003 799397

This study has benefited from comments by Bank of England officials. All content,conclusions, errors and omissions in this study are, however, the responsibility of HMTreasury alone.

This is one of a set of detailed studies accompanying HM Treasury’s assessment of thefive economic tests. The tests provide the framework for analysing the UKGovernment’s decision on membership of Economic and Monetary Union (EMU).The studies have been undertaken and commissioned by the Treasury.

These studies and the five economic tests assessment are available on the Treasurywebsite at:

www.hm-treasury.gov.uk

For further information on the Treasury and its work, contact:

HM Treasury Public Enquiry Unit1 Horse Guards RoadLondonSW1A 2HQ

E-mail: [email protected]

CO N T E N T S

Page

Executive summary 1

1. Introduction 5

2. Theory of the cost of capital 9

3. Evidence on the impact of EMU on the credit risk-free rate 15

4. Implications of EMU entry for the market risk premium 27

5. Implications of EMU entry for UK SME financing 43

6. The structure of UK corporate financing and EMU 47

7. Conclusions: the impact of EMU on the cost of capital 51

References 53

Annex A: Econometric studies of equity market integration 57

Annex B: Corporate bond spreads 63

1 The third of the UK Government’s five economic tests for EMU entry asks whetherjoining EMU would create better conditions for firms making long-term decisions to investin the UK. To inform the assessment, this study considers the potential impact of EMU on thecost of capital for UK firms. Economic theory and evidence suggests a firm will invest if theexpected returns from the investment exceed the cost of the investment. The cost of capitalis therefore an important component of a firm’s investment decision.1

2 Historically, private sector investment levels in the UK have lagged behind those in othermajor economies. A possible explanation is that the cost of capital in the UK has been higherthan it could be, perhaps due to economic inefficiencies, for example in capital markets, or toinstability caused, for example, by mistakes in macroeconomic policy-making in the past. Inor out of EMU, the UK Government places a high priority on maintaining macroeconomicstability and on microeconomic reforms aimed at improving the conditions in which UK firmsraise capital.

3 Firms’ investment decisions are determined by the real cost of capital, which is thenominal cost adjusted for inflation expectations. Firms typically raise capital through eitherdebt or equity. In either case, the cost of capital can be broken down into two keycomponents: the economy-wide credit risk-free rate of return and a market risk premium.There is the potential for both of these components to fall if the UK joined EMU:

• the credit risk-free rate may fall if joining EMU reduces macroeconomicvolatility and lowers inflation expectations. This was an important economicbenefit of EMU for many of the current euro area countries, particularly thosewith histories of high and volatile inflation; and

• the market risk premium component of the cost of capital could fall as theintegration of EMU financial markets has the potential to reduce risk forinvestors in financial assets such as equities and bonds.

4 The credit risk-free rate for major industrial countries which have sustainable debt-to-GDPlevels can be proxied by the yields on government bonds. This reflects the virtually credit risk-freestatus of government debt in these circumstances.

5 Analysis of trends in government bond yields suggests that in euro area countries such asSpain and Italy, where inflation expectations have historically been relatively high, there was asignificant decline in nominal credit risk-free rates in the run up to EMU. Nominal risk-freerates in these countries converged to those of low inflation countries such as Germany andFrance, largely driven by falling inflation expectations. There is no evidence that credit risk-freerates fell in large low-inflation countries in the run-up to EMU as a consequence of prospectivemembership.

6 The expectation that EMU would deliver a more stable macroeconomic environmentmay also have reduced the inflation risk premium, and therefore the real cost of capital, inpreviously high inflation countries. However, the inflation risk premium is unlikely to be animportant influence on UK real interest rates given that the market expects the UKmacroeconomic framework to maintain stable and low inflation. This is in contrast to thesituation in 1997, when UK credit risk-free rates were higher than those of countries such asGermany due to the UK’s history of high and volatile inflation.

1 The analysis in this study of developments in EMU financial markets also provides information that is relevant to theEMU test on financial services, which asks what impact would entry into EMU have on the UK’s financial services industry.

Implications ofEMU for the

credit risk-freerate

EX E C U T I V E SU M M A RY

1


7 While the gap between the UK and euro area credit risk-free rates is no longer as largeas it was in 1997, there would still be some implications for UK credit risk-free rates were theUK to enter EMU, because this may reduce market segmentation between the UK and euroarea government bond markets. This would be driven by the elimination of currency riskbetween the two markets, and the shared official short-term interest rates in the UK and theeuro area. Given this, if all other things were equal, the UK government bond yield curve inEMU would be likely to closely match those of other large AAA-rated government bondmarkets such as Germany, France and the Netherlands.

8 The move to closer convergence with euro area bond yield curves may involve twoshifts. Long-duration UK government yields may rise to euro area levels, while short-durationyields may move down to match lower short-term euro area yields. However, thesemovements are unlikely to have a significant impact on the real corporate cost of capital.Short-term differences in the euro and UK yield curve probably reflect predominantly cyclicalfactors, although joining EMU would remove any premium or discount linked to expectedchanges in the exchange rate. Corporate bond yields at the long end of the curve tend to bedominated by credit risk, limiting the impact of an increase in long UK risk-free rates on thecorporate cost of capital. Overall, this means UK entry is unlikely to have a significant impacton the real corporate cost of capital through changes in the credit risk-free rate.

9 The second component of the real cost of capital is the market risk premium. Severalstatistical studies and surveys of market participants have concluded that the euro areafinancial market has become more integrated since EMU. This has the potential to lower thecost of capital for euro area firms, as the euro area market risk premium could be lower thandomestic market risk premia.

10 The market risk premium is composed of credit risk – the risk of default – and of liquidityrisk – the risk of not finding a seller or buyer at a reasonable price. The credit and liquidityrisks for corporates raising capital in the larger EMU financial market could be expected to belower than in the smaller UK market. Credit risk may be lower as investors are able to spreadrisk by investing in a diversified portfolio of assets across a large market. A larger market willreduce liquidity risk as buying and selling assets becomes easier.

11 There is evidence of growth and integration in the euro area financial market since thestart of EMU. Euro area corporate bond issuance grew strongly after 1999. Euro area equityissuance also grew up to 2000, tailing off with the fall in global equity markets. There havebeen changes in the financial infrastructure in Europe, with mergers between stockexchanges, the establishment of pan-European bond trading platforms, and mergers ofsettlement systems. There is evidence of greater portfolio diversification and a fall intransactions costs within the euro area, suggesting that the integration necessary for a fall inthe market premium is taking place.

12 Increased access to the large and integrated euro area financial market could affect thesize of the risk and liquidity premia on the UK cost of capital. At present large UK firms canaccess the euro market from outside EMU at relatively low cost. However, the removal ofexchange rate risk and transactions costs that EMU would bring, alongside the removal ofsome institutional constraints on foreign currency holdings, would increase access at themargin.

2

Implications ofEMU entry for

the market riskpremium

3

13 For UK borrowers to gain the full advantages of lower financing costs from a singleEuropean financial market there will need to be significant progress on lowering theremaining legal, regulatory and cultural barriers to full integration. These include theretention of regulations restricting the holdings of foreign assets by pension funds and otherinvestment funds, higher transactions costs involved in cross border activity due to the lackof fully-integrated financial infrastructure and the informational costs still faced by fundmanagers investing overseas. Removal of these barriers will benefit borrowers whether or notthe UK enters EMU.

14 The impact of EMU entry on the cost of capital for small and medium-sized enterprises(SMEs) could be very different from that experienced by larger firms. In principle, the removalof currency costs on cross-border financial transactions would be relatively more importantfor SMEs. However, information and monitoring costs are also an important reason whySMEs tend to raise funds through local retail finance. Bank lending is the largest source ofSME finance in the UK, with over 60 per cent of the total. Venture capital is much lessimportant in volume terms, accounting for just 1 per cent of external financing used by SMEs,but it can be an important source of finance in high-risk and high-growth areas.

15 In EMU, smaller SMEs in particular would be likely to remain reliant on local retailfinance. Over the longer term, EMU entry could potentially increase competition in the UKretail market for bank lending to SMEs. It could also increase the size of the venture capitalmarket.

16 UK firms are typically characterised as having a different capital structure from those inthe euro area: ownership is equity-orientated and highly diversified. Large UK firms rely moreon equity to raise capital, while in the euro area bank lending is more important. Someanalysts suggest the UK’s structure leads to capital market imperfections which raise the costof capital, though evidence on this is far from clear.

17 Many indicators suggest the euro area is moving more towards an equity-orientatedstructure. If EMU and other financial developments promote the development of a moreequity-orientated finance structure in the euro area, then EMU entry would be unlikely toalter the structure of UK corporate finance. If different ownership structures continue to existside by side in EMU, and they are augmented by lower barriers to cross border incorporation,this could enable UK firms to utilise different financing structures inside EMU, were the UKto decide to join.

18 Overall, the study finds little scope for UK credit risk-free rates to fall significantly werethe UK to enter EMU. The market risk premium for corporate borrowers raising capital in thelarger EMU financial market could be expected to be lower than in the smaller UK market. UKfirms can access the euro financial market from outside EMU at relatively low cost, but entrywould increase access at the margin. These issues are considered in the assessment of theinvestment test – the third of the Government’s tests for EMU entry.


Implications ofEMU entry for

the structure ofcorporatefinancing

Conclusions

Implications ofEMU entry forSME financing

The 1997assessment

1 IN T R O D U C T I O N

5

1.1 The third of the UK Government’s five economic tests for EMU entry asks:

“would joining EMU create better conditions for firms making long-termdecisions to invest in Britain?”

1.2 To inform the assessment of the five tests, this study considers the potential impact ofEMU on the cost of capital. Economic theory and evidence suggests a firm will invest if theexpected returns from the investment exceed the cost of the investment. The cost of capitalis therefore an important component of a firm’s investment decision.1

1.3 Historically, private sector investment levels in the UK have lagged behind those in othermajor economies. A possible explanation is that the cost of capital in the UK has been higherthan it could be, perhaps due to economic inefficiencies, for example in capital markets, or toinstability caused, for example, by mistakes in macroeconomic policy-making in the past. Inor out of EMU, the UK Government places a high priority on maintaining macroeconomicstability and on microeconomic reforms aimed at improving the conditions in which UK firmsraise capital. For example, the Government has an active programme of reforms to domesticcapital markets, following the Cruickshank, Myners and Sandler reviews. The Government hasalso introduced reforms aimed at increasing the supply of risk capital to UK enterprises.Moreover the Government sees macroeconomic stability as a central objective, a platformfrom which microeconomic policy reforms can be delivered and market productivityimproved.2

1.4 The 1997 assessment of the Government’s five economic tests highlighted theimportance of this issue for many of the current euro area countries. It noted that for manycountries “a lasting fall in nominal and real interest rates is one of the main economic reasonsfor joining EMU.” Several euro area countries benefited from a significant fall in nominalinterest rates in the run up to EMU entry, on the expectation that EMU would deliver low andstable inflation. This was particularly the case for countries which, in the past, hadexperienced high and volatile inflation.

1.5 The 1997 assessment also noted that in the case of the UK there was at the time adifferential between UK and German interest rates which suggested “there would be acredibility gain for the UK from joining EMU” possibly leading to lower interest rates and alower cost of capital. The situation now is very different. As a result of the reforms to the UK’smacroeconomic framework introduced in 1997, the UK has a stable macroeconomicenvironment with low inflation. The differential between UK and German interest rates hasnarrowed sharply. Indeed for longer maturity interest rates the differential has reversed, sothat UK rates are now lower than those in Germany. In effect, the potential increase incredibility referred to in the 1997 assessment has been achieved outside of EMU.

1.6 From this starting point the objective of this study is to consider how, against thisbackdrop, possible EMU entry would affect the UK cost of capital. The study does notattempt to directly measure an average cost of capital in the UK or the euro area. As isdiscussed in Section 2, for a number of reasons this is a very difficult task. It is also notnecessary for the purposes of this study. Instead, the approach taken is to break down thecost of capital into its component parts and then analyse the implications of EMU for each

The key issuesconsidered in

this study

1 The analysis in this study of developments in EU financial markets also provides information that is relevant to the EMUtest on financial services, which asks what impact would entry into EMU have on the competitive position of the UK’sfinancial services industry.2 See successive Budgets and Pre-Budget Reports for full details on this reform programme.

IN T R O D U C T I O N

part. A brief summary of this approach is given below and is explained in more detail inSection 2, which sets out the analytical framework for the study. The nature of this topicmeans that it covers some technical and complex economic issues. Some of the key technicalterms used throughout are set out below.

1.7 The starting point for the analysis is to consider the ways in which firms raise capital forinvestments. A significant quantity of investment is funded with internal finance fromretained profits. EMU could affect the funds available for internal finance, for examplethrough its impact on growth, an issue considered in the Government’s fifth economic test forEMU. However, this study focuses on the implications of EMU for the cost of external finance.Firms typically raise external finance in one of two ways: debt or equity. The components ofthe cost of debt and equity finance are very similar.

1.8 In both cases, the base is the credit risk-free rate of return in the economy. This dependson the balance of aggregate savings and investment in the economy, and so reflects firms’ andconsumers’ preferences between current and future consumption. In major industrialisedcountries with sustainable public debt levels, the interest rate on government bonds can beused to provide an indication of the credit risk-free rate, and this is the central approach takenin this study. However, there are a number of reasons why government bonds are notnecessarily an accurate proxy for the credit risk free-rate. These issues are explored in detailin later sections, and alternative proxies are considered.

1.9 On top of the credit risk-free rate of return, the cost of capital for firms includes a marketrisk premium, which investors demand to reflect the perceived risk of investing in aparticular market, sector, firm or project. The market risk premium might be affected by EMUentry because it depends in part on the size and efficiency of financial markets.

1.10 The degree to which EMU affects both of these components of the cost of capitaldepends in large part on the degree of financial market segmentation across countries. Ifmarkets were fully integrated then the cost of capital faced by firms of equal riskiness wouldbe equal across all countries. However, if there are barriers to cross-border investment, whichprevent the free flow of capital across borders, then the cost of capital can be different indifferent countries. A well-known puzzle in financial economics is that investors have a homebias – they invest heavily in domestic assets and less than might be expected in overseasassets. A key issue for this analysis is whether the exchange rate acts as a barrier to cross-border flows of capital, or whether market segmentation and home bias are primarily due toother factors.

1.11 To address these questions the analysis in this study is split into six sections:

• Section 2 outlines the theoretical framework used in the analysis;

• Section 3 considers the possible impact of EMU on the UK credit risk-free rateof return;

• Section 4 considers the possible impact of EMU on the UK market riskpremium;

• Section 5 considers the implications of developments in EMU financialmarkets for UK small and medium-sized enterprises (SMEs);

• Section 6 considers how changes in the structure of corporate finance in EMUmight affect the UK cost of capital; and

• Section 7 concludes.

6

1

The market riskpremium

The credit risk-free rate of

return

The degree ofmarket

segmentationand home bias

Structure of thestudy

7

1.12 In order to provide a base point for the analysis, it is important to set out the appropriatecounterfactual. In other words, what environment does the analysis assume firms raisingcapital would face if the UK remained outside EMU?

1.13 In terms of the credit risk-free rate and UK government bond yields, the counterfactualis that outside EMU the UK macroeconomic policy framework put in place in 1997 continuesto maintain a low and stable level of inflation in line with the Government’s objectives. Thequestion that is addressed in this study is how entry to EMU might affect the credit risk-freerate against this base.

1.14 In terms of financial markets, the counterfactual is that the UK financial sectormaintains and builds on the strong links that it already has with those in the other EUMember States, and that the integration of EU financial markets continues, driven in part bythe Financial Services Action Plan (FSAP). The potential gains from greater financialintegration within Europe are large. Recent estimates for the European Commission (LondonEconomics, 2002) suggest full market integration could lead to a fall in the EU average cost ofequity and bond capital of around 50 and 40 basis points respectively, which, it is estimated,could boost the level of GDP by 1.1 per cent across Member States in the long run. Thequestion addressed in this study is whether, were the UK to enter EMU, there would beadditional cost of capital savings for UK firms on top of these estimated savings.

1.15 Two other EMU studies are particularly relevant to this analysis. The EMU study by HMTreasury The location of financial activity and the euro considers the development of financialmarkets in the euro area. The EMU study by HM Treasury EMU and business sectors examinesdevelopments in the business environment in Europe in recent years and considers theimpact of EMU on these trends. It also considers the potential impact of EMU on the businessenvironment for SMEs. The implications of EMU entry for the cost of capital faced by SMEs isthe focus of Section 5 of this study.

IN T R O D U C T I O N1The counter-

factual for theanalysis

The study in thecontext of the

five testsassessment and

other EMUstudies

1 Tobin (1969).

Investmenttheory

2 TH E O RY O F T H E CO S T O F CA P I TA L

9

2.1 This section sets out the theoretical structure used to analyse the cost of capital, andhighlights some of the key difficulties that are faced in considering the impact of EMU. Thestarting point is an explanation of why the cost of capital is an important component of afirm’s decision to invest. The analysis then considers the components of the cost of debt andequity, and how these components could be affected by EMU entry.

Why the cost o f capi ta l matters2.2 Economic theory points to an important role for the cost of capital in investmentdecisions. For example, the simple neo-classical theory of investment suggests that a firm willinvest until the marginal return from investment equals the marginal cost of capital (forexample see Jorgensen, 1963). Two central factors drive investment in the neo-classical model:the return from investment, which is governed by the price and volume of output; and the costof capital, which is determined by factors such as the interest rate, depreciation and tax.

2.3 A shortcoming of the simple neo-classical model is that there is no explicit forward-looking element, for example, there is no direct consideration of expectations of futureprofits. This is addressed in dynamic models of investment, of which Tobin’s Q model1 is anexample. Tobin’s Q is the ratio of the forward-looking stock market valuation of the firm(which approximates to the market’s estimate of the present value of new investment) to theprice of new equipment (which approximates to the marginal cost of capital). Thisrelationship can be viewed as the ratio of the present value of marginal investment to themarginal cost of the investment. If Tobin’s Q is greater than one, i.e. if the marginal value ofinvestment exceeds marginal cost, then it makes sense for firms to invest more. The optimallevel of investment is where Tobin’s Q equals one.

2.4 Empirical studies using these approaches have tended to find that quantity variablessuch as output dominate the relationship, and that there is only a weak link betweeninvestment and the cost of capital (though see OECD, 2002 for a recent example of where asignificant relationship between investment and the cost of capital has been found). Onereason may be the difficulty in accurately measuring the cost of capital facing firms, an issuewhich is discussed further below.

2.5 This discussion highlights that in both the neo-classical and dynamic models the costof capital plays a key role in determining investment levels, which is why it is an importantissue to examine in the context of the investment test for EMU entry.

2.6 The analysis in this study is focused on the implications of EMU entry for the cost ofcapital raised externally by firms. However, a significant quantity of investment is fundedwith internal finance from retained profits. The cost of capital will influence both internallyand externally financed investment as firms will consider the opportunity cost of usingavailable funds in terms of returns available elsewhere. In addition, EMU could affect thefunds available for internal finance, for example through its impact on growth. The impact ofEMU on growth, stability and jobs is considered in the Government’s fifth test formembership of EMU.

The components o f the cost o f capi ta l2.7 There are two general forms of external finance available to firms: debt and equity.Firms raise debt finance by borrowing from a financial institution or by issuing a bond. Firms

TH E O RY O F T H E CO S T O F CA P I TA L

raise equity capital by selling a share in future profits, either through a public offering orthrough a private deal. The overall cost of finance to a firm is the weighted average cost of itsdebt and equity. There is an extensive literature on issues involved in measuring the weightedaverage cost of capital and numerous variations on the basic methodology have been devised(for example, see Brealey and Myers, 2000). This study does not aim to measure the cost ofcapital, and therefore does not consider these methodological issues in detail.2 The objectiveof the study is to consider how the cost of capital for UK firms may be affected by EMU entry.The framework used is to break down the cost of capital into its component parts.

2.8 Chart 2.1 illustrates the components of the cost of debt. The interest rate on borrowingor bond issuance represents the overall cost to the borrower of debt (there may be additionalcosts such as one-off arrangement fees). The interest rate can be divided into a number ofdifferent components. First, is the credit risk-free rate which can be proxied, under certainconditions, by the rate of return on government bonds of major industrialised countries withsustainable debt levels.3 In addition to this is a market risk premium, which is the returninvestors demand for holding a risky asset. The market risk premium can be divided into twofurther components. Liquidity risk is the risk involved in finding a counterpart for a desiredtransaction at a reasonable price. Credit risk is the risk that the borrower will default on thedebt.

2.9 The components of the cost of equity are similar to those of the cost of debt, in thatthey can be characterised as the credit risk-free rate of return plus the market risk premium.One of the most widely used methodologies for calculating the cost of equity is the capitalasset pricing model (CAPM). In this model the cost of equity is defined as the economy-widerisk-free rate of return plus the individual risk premium the market attaches to an investment:

Cost of equity capital = credit risk-free rate + stock’s beta X market risk premium.

10

2

2 Accurate measurement of the cost of capital at a country level is difficult. Figures for the cost of debt, particularly bankfinancing, are often not widely available. Measuring the cost of equity using a model such as CAPM requires estimates ofbeta and of the equity market premium. These can be estimated on the basis of past data, but this assumes no changegoing forward, which may not be realistic. If accurate measures of the cost of debt and equity can be found, the next stepwould be to put together a weighted average cost of capital using the proportion of debt and equity used by firms. Thisrequires data on the stock of debt and equity which is often not available. There are also difficulties in estimating the realcost of capital. This should properly be derived using ex ante measures of expected inflation, but these are often notavailable.3 An alternative, examined later in the paper, is swap rates, which can be useful where institutional or regulatory factorsmay be skewing the yields on government bonds.

The componentsof the cost of

debt

Chart 2.1: Components of the external cost of debt

Liquidity risk – the risk offinding a counterpart for

a transaction at areasonable price

Credit risk – the risk thata borrower will default

Credit risk-free rate – therate of return on a risk-free

asset

Market risk premium

Interest rate

The componentsof the cost of

equity

11

2.10 The equity market risk premium is made up of two components:

• The market risk premium is the premium that investors require for the risk ofputting money into a market portfolio of equities rather than into a risk-freeasset. It is possible to estimate the historical market premium on the basis ofthe degree to which the market has outperformed risk-free assets in the past.Over the long run, investors should drive prices of shares towards the levelneeded to yield the premium they require to compensate them for the risk ofholding shares.

• Each individual stock contributes to the risk of a portfolio, depending on thestock’s correlation to the general market movement, known as the beta. Astock which moves more than the market has a beta of greater than one; astock which moves less than the market will have a beta of less than one.Picking a portfolio from stocks which all have betas of two would produce aportfolio with a beta of two, i.e. that moves twice as much as the market.

2.11 The analysis in this study focuses on the potential impact of EMU on the two keycomponents of the cost of capital: the credit risk-free rate and the market risk premium. Atheoretical basis for the analysis of these two components is now outlined.

The credi t r i sk- free rate2.12 In major industrial countries with sustainable debt-to-GDP levels, government bondsare assumed to be virtually credit risk-free. This makes it possible to use interest rates ongovernment bonds as a proxy for credit risk-free rates, and this is the approach taken in thepaper.4 However, there are also a number of reasons why government bond yields may notalways be an accurate proxy for credit risk-free rates. These are discussed in more detail inSection 3 where alternative proxies for the risk-free rate, such as the swap rate, are presented.5

2.13 Box 2.1 explains the economic drivers of countries’ government bond yields. Thissuggests two reasons why credit risk-free rates may differ across countries:

• differences in inflation expectations and the inflation risk premium; and,

• differences in the real rate of return.

2.14 The nominal interest rate compensates investors for the expected reduction in the realvalue of an asset through future inflation. The EMU study by HM Treasury Policy frameworksin the UK and EMU explains that the inflation targets of the monetary authorities in the UKand euro area are similar in practice. This means there is unlikely to be a significant differencein the expected rate of inflation between the UK and the euro area. This is discussed furtherin Section 3.

2.15 In addition to inflation’s influence on the nominal rate, it can also influence the realcomponent of the credit risk-free rate. If investors believe there is a risk that inflation will behigher than expected, they may demand an inflation risk premium.

TH E O RY O F T H E CO S T O F CA P I TA L2

Analysis of thecost of capital in

the study

Inflationexpectations

The inflation risk premium

4 Government bond yields also include a liquidity premium, which reflects the ease with which the bond can be traded.For the purposes of using the government bond yield as a proxy for the risk-free rate, this is ignored in the subsequentanalysis. It is looked at again in analysis in Section 3 of likely developments in UK government bond yields were the UK toenter EMU.5 See also Cooper and Scholtes (2001) for a similar discussion.


2.16 In general, the real rate of return can be thought of as the price which equates savingsand investment. In the absence of any restrictions on capital flows, the real rate of returnshould be equal across countries, as any differences would be removed through internationalarbitrage. There would be a single ‘world’ real interest rate equating world saving andinvestment. Alternatively, in a world where national markets are completely segmented, theneach national real interest rate would be the price which equated domestic savings andinvestment. In this case, the real interest rate would not necessarily be equal across countries.

12

2Box 2.1: Economic determinants of international government bondyields

Three economic relationships are relevant to the differentials in international bond yields1:Uncovered Interest Rate Parity (UIP), Purchasing Power Parity (PPP), and the FisherEquation.

UIP hypothesises that in a world of freely floating exchange rates and perfect capitalmobility, interest rates and exchange rates should be such that an investor would beindifferent between holding an interest-bearing asset denominated in domestic currency,and an equivalent one denominated in foreign currency. Formally, UIP can be defined as:

i*t – it = se

t+1 – st + ρt [1]

where: i*t and it are the foreign and domestic one-period nominal interest rates

respectively; st is the spot exchange rate (foreign currency price of domestic currency);se

t+1 is the market’s one period ahead forecast of the spot exchange rate; and ρt is the riskpremium.

PPP states that the differential between the expected inflation rates between twocountries is equal to the expected movement of the exchange rate between the twocountries’ currencies:

∏e*t+1 – ∏e

t+1 = set+1 – st [2]

where: ∏e*t+1 and ∏e

t+1 are the market’s forecasts for the change in the foreign anddomestic price levels respectively between t and t+1. Combining equations [1] and [2]gives:

i*t – it = ∏e*

t+1 – ∏et+1 + ρt [3]

Equation [3] states that the nominal interest rate differential between the bonds of twocountries is equal to the expected inflation differential between the two countries plus ainflation risk premium. That is, differences in international bond yields can be explained,all other things remaining equal, by the difference between expected inflation rates, aswell as by the inflation risk premium.

Finally, the Fisher Equation states that the relationship between the nominal return ofa bond is equal to the expected rate of inflation over the holding period of the bond plusits real rate of return, rt:

it = ∏et+1 + rt

The Fisher equation can be generalised, as investors may be interested in the externalvalue of the currency as well as inflation, by substituting the expected outcome of theinvestor’s targeted value, e.g. the exchange rate, for the expected rate of inflation, ∏e

t+1.For euro area investors, their target will have changed with the introduction of EMU, sothat those targeting inflation will need to consider whether it is euro area inflation ornational inflation that matters.1 Based on Brooke et al. (2000).

The real rate ofreturn

13

2.17 There are a number of reasons why national markets might be segmented, such asinstitutional or regulatory restrictions on cross-border capital flows or a home bias toinvestment strategies which could be caused by information asymmetries, or perhaps bycurrency risk. Evidence suggests that national real interest rates do differ, suggesting thatcapital markets are at least partially segmented (for example, see Breedon et al., 1999). Thediscussion in Section 3 considers reasons why real rates of return on long-dated bonds maydiffer in the UK and the euro area.

2.18 Even under the assumption of completely integrated markets, real risk-free rates maydiffer in the short run for cyclical reasons. If a country’s real exchange rate is away from itsmedium or long-term equilibrium level, perhaps because output is away from trend, thenshort-term real interest rates may diverge because of expected real exchange rate movements.For example, if investors expect a country’s real exchange rate to depreciate, then they woulddemand a higher real interest rate. However, in the long term, with the real exchange rate atequilibrium and free capital flows, real rates should be equal. This study is primarilyinterested in the long run implications of EMU entry, though of course the short runimplications are also important and are considered.

The market r isk premium2.19 The market risk premium is the risk attached to investing in a risky asset over a risk-free asset. As the discussion of the cost of debt and equity above indicated, there are two keycomponents of this risk. First is liquidity risk, the risk that an investor is unable to find a buyeror seller for an asset at an acceptable price. Second is the risk that the firm will default on itsobligations, the credit risk.

2.20 One of the most significant potential benefits of EMU comes from the creation of adeep and broad capital market across the euro area. A deeper market, defined as a marketwhere assets are heavily traded, would reduce market participants’ liquidity risk. A broadermarket, defined as a market where a wide range of assets are traded, would allow participantsto diversify their holdings, so reducing credit and sector risk. A larger market is also likely tobring reduced transactions costs. These effects have the potential to reduce risk and lowercosts for investors in market assets such as equities and bonds. The EMU market premiummay therefore be lower than the market premium faced in national markets.

2.21 As Stulz (1999) explains, in a world of fully segmented national capital markets, eachcountry’s investors would have to bear the full risk of their country’s economic activity. Themarket risk premium they would demand for this would increase with national market risk.With capital market integration, domestic investors are able to diversify their risk profile byholding foreign assets. Because some of the unique credit risk of domestic and foreign assetsis likely to offset each other, investors can hold an international portfolio which has the sameexpected returns as previously but with lower risk. In practice, investors who invest purely intheir national securities markets are likely to enjoy some exposure to the internationaleconomy, as some domestically-listed firms will operate across borders.

TH E O RY O F T H E CO S T O F CA P I TA L2

Implications ofmarket

segmentation forthe market risk

premium


Market segmentat ion and EMU

2.22 As the discussion above has highlighted, the degree of market segmentation will affectboth the credit risk-free rate and the market risk premium. There are a number of reasons whyEMU might reduce financial market segmentation among euro area countries. The removalof exchange rates within the euro area removes a transaction cost to investing abroad; and italso reduces exchange rate risk on European investments. Furthermore, many Europeaninvestment funds are constrained by regulation to keep a certain proportion of their assetsdenominated in domestic currency. It is possible that UK EMU entry would provoke a one-offchange in net flows, as institutional investors diversify into UK assets and UK institutionsmove into euro area assets. If UK institutions have already diversified more than their euroarea counterparts, then flows into the UK from the euro area may exceed flows out of the UK,possibly increasing the availability of funds for UK firms.

2.23 Research underlines that currency risk on its own does not explain home bias (forexample, see Brealey et al. 1999). Uncertainty about exchange rates can be hedged through avariety of financial products. Currency risk may not be undesirable; it may actually enableinvestors to diversify risk in their portfolio. Another explanation for home bias is that there aresignificant transaction costs and information costs to holding foreign assets. EMU in itselfwill only reduce transaction costs to the extent that it removes currency exchange costs; andit will only reduce information costs through the increased price transparency that comesfrom having a single currency. EMU could indirectly reduce transaction costs if it acts as acatalyst for the development of a more efficient regulatory environment in the EU financialmarkets, or promotes integration of financial market infrastructure such as trading andsettlement systems.

2.24 The assumption in this study is that the integration of EU financial markets continues,driven in part by the Financial Services Action Plan (FSAP), and that this helps to reduce thecost of capital for EU firms. The question addressed in this study is whether if the UK were toenter EMU, there would be additional cost of capital savings for UK firms.

The proport ions o f debt and equity used by f i rms2.25 EMU may also affect the proportions of debt and equity used by firms, which mayaffect the weighted average cost of capital. Large firms in euro area countries such asGermany have typically used greater quantities of bank lending than large companies in theUK, who have tended to issue more equity. Structural change and the ongoing process offinancial market integration in the euro area, which EMU is one part of, are promoting greateruse of equity funding in some euro area countries. Section 5 considers this issue.

Summary: the key components o f the analys is 2.26 This theoretical review sets out three key elements for the subsequent analysis:

• the implications of EMU for the credit risk-free rate;

• the implications of EMU for the market risk premium; and

• the implications of EMU for the proportions of debt and equity used by firms.

14

2

EMU countries where inflation expectations and inflation risk have historically beenrelatively high witnessed a decline in nominal credit risk-free rates in the run-up to EMU.In part, this can be attributed to the expectation that EMU would deliver a more stablemacroeconomic environment.

If the UK were to enter EMU, it is likely that UK credit risk-free rates would convergeclosely with those of the euro area. However, because the UK macroeconomic frameworkis expected to maintain stable and low inflation, these rates are already very similar.Moreover, short-term differences which do exist probably reflect cyclical factors ratherthan underlying structural differences. This means UK entry is unlikely to have asignificant impact on the real corporate cost of capital through changes in the creditrisk-free rate.

3 EV I D E N C E O N T H E IM PAC T O F EMUO N T H E CR E D I T R I S K-FR E E RAT E

15

3.1 This section considers the impact of EMU on credit risk-free rates of return. It firstconsiders developments in the existing euro area countries, and then looks at the potentialimplications of UK entry for the UK credit risk-free rate.

Impact o f EMU on credi t r i sk- free rates in the euro area

3.2 In major industrial countries with sustainable debt-to-GDP levels, government bondsare virtually credit risk-free – there is almost no risk that such governments will default ontheir debt commitments. Of course, no government bond will ever be completely free ofcredit risk; as discussed below, small differences in credit risk are one reason for theremaining small differences in euro area government bond yields. However, for this analysis,international government bond yields are used as a convenient proxy for credit risk-freerates. But for completeness, another potential proxy – the swap rate – is also considered. Thisalso addresses any bias in the analysis stemming from some specific features in the UKgovernment bond market, discussed in more detail below. A similar analysis can be found inthe article by Willem Buiter in the EMU study Submissions on EMU from leading academics.

3.3 Chart 3.1 presents the development of nominal euro area government bond yields inthe run-up to the start of EMU. The chart shows that nominal yields converged over thisperiod and were very similar by the start of EMU in 1999, reflecting the common monetarypolicy within EMU. Remaining differences are due to small variations in the credit risk ofnational governments and in liquidity differentials. Chart 3.1 shows a steep decline innominal yields in Spain and Italy, countries which have had histories of higher inflation (seeChart 3.2). As discussed in Section 2, inflation expectations are one of the factors explainingdifferences in international bond yields.

3.4 This suggests that an expectation that EMU would provide a low inflationenvironment has driven down nominal yields in these countries. In addition to a fall ininflation expectations, Spain and Italy will have experienced a decline in the inflation riskpremium. If the macroeconomic environment in EMU is seen as more credible there is lessrisk of inflation being higher than expected.

Developments innominal euro

area governmentbond yields

EV I D E N C E O N T H E IM PAC T O F EMU O N T H E CR E D I T R I S K-FR E E RAT E

3.5 The decline in nominal yields in Germany, France and the Netherlands in Chart 3.1 isless steep and is unlikely to reflect lower inflation expectations in EMU – these countriesalready had relatively low inflation in the 1990s. Chart 3.3 indicates that the pattern of fallingyields over the 1990s is not restricted to EMU countries. Yields in the UK and the US have alsodeclined over the 1990s, after a period of relatively high and volatile yields in the 1970s and1980s. Brooke et al. (2000) find that much of the relative fall in UK yields over the past 25 yearscan be attributed to a decline in relative UK inflation expectations.1 A number of UK marketspecific factors have also contributed to this fall and are considered in more detail insubsequent sections.

16

3

Source: Ecowin.

0

2

4

6

8

10

12

14

16

Chart 3.1: 10-year government bond yields in euro area countries (annual average)

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Per cent

Germany France Italy Spain Netherlands

Source: HM Treasury.

Chart 3.2: Inflation in euro area countries (annual average)

-2

0

2

4

6

8

10

Per cent

Germany France Italy Spain Netherlands

1986 1988 1990 1992 1994 1996 1998 2000 2002

1 These issues are also considered in Cooper and Scholtes (2001).

17

3.6 It is important to make clear the distinction between real and nominal rates of return.Nominal interest rates are equal to the real interest rate plus the expected rate of inflation andthe inflation risk premium (for more detail see the discussion of the Fisher Equation in Box2.1). Ultimately it is real interest rates, and the real cost of capital, which will influence firms’investment decisions.

3.7 This analysis suggests that high inflation countries experienced a decline in nominalyields in the run-up to EMU. Much of this fall in nominal rates reflects a decline in inflationexpectations and so does not translate into a decline in the real cost of capital which mattersto firms. However, part of the fall can be attributed to a decline in the inflation risk premiumassociated with these countries’ assets. The inflation risk premium reflects the risk thatinflation will be higher than expected, and tends to be higher when actual inflation is highand volatile. A fall in the inflation risk premium is a real gain in the credit risk-free rate ofreturn in these countries. Assuming that the cost of capital for firms in these countries isclosely linked to the credit risk-free rate, this translates into a real fall in the cost of capitalfor firms.

3.8 As discussed in Section 2, the underlying real interest rate is driven by the supply anddemand for savings and investment, and in the absence of any restrictions on capital flows,the real rate of return should be equal across countries, as any differences would be removedthrough international arbitrage. There would be a single world real interest rate equatingworld saving and investment. Alternatively, with fully segmented capital markets eachcountry would have a national risk-free real interest rate which may or may not be equal tothat in other countries. International capital flows within the euro area were relativelyunrestricted before EMU, which suggests persistent and significant differences in real interestrates would be unsustainable. However, exchange rate risk and regulatory restrictions oncurrency denomination of asset holdings may have created barriers to full capital marketintegration, allowing for the possibility of differences in real interest rates. To the extent thatEMU removes these barriers, it may have created the potential for further convergence of realinterest rates.

3.9 Deriving real interest rates from nominal interest rates requires an estimate of inflationexpectations. This is difficult to measure. One simple approach is to assume that inflation

EV I D E N C E O N T H E IM PAC T O F EMU O N T H E CR E D I T R I S K-FR E E RAT E3

Developments ineuro area realrisk-free rates

USUKSpainItalyFranceGermany

Per cent

20021998199419901986198219781974197019661962

Chart 3.3: 10-year government bond yields (annual average)

0

5

10

15

20

25

Source: Ecowin.

Nominal and realyields


expectations are based on past inflation, and so use an average of past inflation as an estimateof expected future inflation. Chart 3.4 presents estimates of real credit risk-free interest ratesin EMU countries by subtracting inflation (measured over the previous three years) from thenominal yield on ten-year government bonds. These should be approached with care as usingbackward-looking inflation expectations in an environment of declining actual inflation canbias down the real interest rate. The chart shows a decline in real credit risk-free rates in theeuro area countries in the years leading up to and since the start of EMU. A similar decline hasbeen seen in the UK, which makes it difficult to attribute the decline to EMU itself.

3.10 A different approach to analysing real interest rates is to look at the yields on index-linked government bonds. Chart 3.5 presents yield curves for index-linked bonds in the UKand France – the only large euro area country to issue this type of bond.2 These show UK ratesconsistently below French rates over all maturities. However, it is difficult to use these yieldcurves to directly compare UK and French real rates because:

• inflation measures differ. So for example, using HICP measures of UK inflationrather than RPI would suggest a lower interest rate differential;

• the French index-linked bonds are new instruments and real yields will tendto be high as the market matures; and

• index-linked gilts are illiquid instruments, with activity in the marketconcentrated around auction dates. The major buyers of index-linked debt,pension funds and insurance companies, tend to buy and hold them in orderto hedge their indexed liabilities.

18

3

2 A yield curve maps yields against different bond maturities.

Chart 3.4: Real interest rate on 10-year government bonds, using backward-looking inflation expectations1 (annual average)

1 Inflation calculated as three-year moving average of past inflation.Source: Ecowin and HM Treasury calculations.

0

1

2

3

4

5

6

7

8

9

10

UK Netherlands Spain Italy France Germany

2002200120001999199819971996199519941993199219911990

Per cent

19

3.11 Evidence suggests that the decline in risk-free rates in the euro area has been passedthrough to commercial rates. Since EMU began in 1999, money market rates in EMU haveequalised, due to the single monetary policy operated by the European Central Bank (ECB)and the ability of banks to transfer euro liquidity around the EU via the TARGET paymentsystem. Again, this resulted in a large decline in rates in Spain and Italy. Chart 3.6 illustratesthis fall in the inter-bank lending market in EMU countries.


0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Chart 3.5: Real yields on UK and French index-linked government bonds1

5 10 15 20 25 30 35

Per cent

Years to maturity

1 As of December 2002.Source: Bloomberg and Agence France Trésor.

UK (linked to UK RPI) France (linked to French CPI, excluding tobacco)

Pass-through tomoney market

rates

Source: Datastream and HM Treasury calculations.

Chart 3.6: 12-month money market interest rates (annual average)

Per cent

Germany SpainFrance Italy Netherlands

0

2

4

6

8

10

12

14

20022001200019991998199719961995199419931992


3.12 In conclusion, the impact of EMU on credit risk-free rates in the euro area has been tocontribute to a lowering of nominal rates in those countries, especially in Spain and Italy,where inflation expectations and inflation risk have historically been relatively high. As theanticipation that these countries would join the euro grew, their nominal yields converged tothose of low-inflation countries, such as Germany, on the expectation that EMU woulddeliver a stable macroeconomic environment. To the extent that this fall reflected a decline inthe inflation risk premium, this translated into a decline in real yields. There is evidence thatthese declines have fed through to commercial rates, and so lowered the cost of capital inthese countries.

3.13 As Chart 3.4 illustrates, the UK also experienced a fall in credit risk-free rates during the1990s. The implications of possible UK entry for the UK credit risk-free rate are explored inmore detail below.

Impl icat ions of EMU entry for the UK credit r isk- free rate

3.14 Following the 1997 reforms of the UK macroeconomic framework, current marketexpectations are that the UK will maintain stable and low inflation; this means nominal creditrisk-free rates in the UK are already low and are close to those of the euro area. In fact, as Chart3.7 shows, the differential between UK government bonds (gilts) and German governmentbonds (bunds) at the 30-year maturity level has reversed since 1997. Table 3.1 summarises theevolution of these differentials. In May 1997, 30-year gilts were priced at 63 basis points overbunds; in February 2003, 30-year bunds were 31 basis points over gilts. Declines in the UK giltspread over bund yields have also been noticeable at five and ten-year durations.

3.15 This suggests that the UK will not see a decline in credit risk-free rates to anything likethe degree experienced by countries such as Spain and Italy. However, there would beimportant implications for UK government bond yields of entry into EMU. Many of thesereflect structural factors specific to the UK bond market which would be affected by UK entryto EMU.

20

3

Source: Bloomberg.

Chart 3.7: Difference between yields on UK and German government bonds

-200

-150

-100

-50

0

50

100

150

200

250

300

350Basis points

Feb-9

6

Feb-9

7

Aug-9

6

Aug-9

7

Feb-9

8

Aug-9

8

Feb-9

9

Aug-9

9

Feb-0

0

Aug-0

0

Feb-0

2

Aug-0

2

Feb-0

3

Feb-0

1

Aug-0

1

10-year5-year 30-year

21

Table 3.1: Summary of differential between UK and German governmentbonds

Basis points 5-year 10-year 30-year

May 1997 +226 +133 +63

October 1997 +145 +88 +28February 2003 +71 +23 –31

Note: A positive differential indicates higher UK yields.

Source: Bloomberg

3.16 A number of factors are responsible for 30-year UK gilt rates falling beneath those ofbunds.3 Economic theory predicts that inflation expectations are a key driver of yields, andthere is evidence that the narrowing of the gap between inflation expectations in Germanyand the UK, due to the greater stability provided by the new UK macroeconomic framework,has reduced gilt rates at all maturities.

3.17 However, a number of UK market-specific factors have also played a role:

• on the supply side, improving government finances, and a limited supply ofsubstitute corporate securities has led to a tightening of gilts supply;4 and

• on the demand side, there is strong demand for gilts as a result of a number ofUK-specific institutional factors relating to the wider coverage of privatepension funds in the UK. For example, the Minimum Funding Requirement,Guaranteed Annuity Options and FRS17 have created incentives for UKpension funds to hold long-dated UK government bonds.

3.18 Entry into EMU would eliminate currency risk between UK government bond marketsand those of the euro area. It would also mean that the UK and euro area would share acommon official short-term interest rate. Given this, if all other things were equal, one wouldexpect the UK government bond yield curve to closely match those of other large AAA-ratedgovernment bond markets such as Germany, France and the Netherlands. In theory therecould be a negative impact on the UK’s credit rating from joining the single currency, asparticipant countries can no longer use central bank borrowing to finance governmentspending. However, countries retain the ability to meet debt payments with tax revenue andno participant has experienced a downgrade in credit rating since the introduction of thesingle currency.

3.19 By entering the single currency, UK investors would have access to a wider pool ofassets to choose from. For example, institutional investors who currently hold long gilts maywish to purchase instead, without any currency risk, German or French long-maturitygovernment bonds if the yields are higher. Equally, euro area investors will have a bigger poolof assets if the UK enters EMU. This could lead, based on the assumptions in Chart 3.8 below,to net flows into UK assets at the short end. Overall supply and demand of savings andinvestments should be unaffected, but there may be distributional effects. The impact ondifferent durations of debt is discussed below.


Implications ofUK entry for the

gilt-bunddifferential and

the UK yieldcurve

3 For a fuller explanation see Brooke et al. (2000).4 The UK’s fiscal deficit of 2.9 per cent of GDP in 1996-97 became a surplus of 3.9 per cent in 2000-01.


3.20 How close the UK’s yield curve would map onto those of other AAA-rated governmentswould be in part dependent upon the liquidity and benchmark status of the UK’s governmentbonds in relation to other AAA-rated government counterparts.5 The status of UK governmentbonds in the euro area market would depend upon a number of factors including:

• the size of the market. The UK has a longer average maturity of debt thanother EMU issuers. If the UK were to join EMU, its bonds would constitutearound 32 per cent of the resulting total market stock above ten years.6 Thisdominance should have a favourable impact on the liquidity and thebenchmark status of gilts of long-term maturity, and could lead to the longend of the UK yield curve lying beneath the bund curve;

• investor inertia. Currently, UK pension and life funds are the largest holdersof long-dated gilts. If the UK enters EMU, these institutions could enjoy a yieldpick-up by switching to other EMU government bonds without an increase incurrency risk (as mentioned above, institutional factors currently createincentives for UK pension funds to hold long-dated gilts). This would result inswitching from gilts to bunds and the UK yield curve matching closely thelong end of the bund curve. However, movement by UK pension and life fundsaway from long-dated gilts may be slowed by their unfamiliarity withdiversifying into other markets and investor inertia; and

• better long-term prospects for UK creditworthiness. The UK Government hasbetter long-run financial prospects than its AAA-rated German and Frenchpeers, as it has a lower net debt-to-GDP ratio and a significantly lower level ofunfunded pension liabilities in comparison to Germany and France. Giltsshould therefore have slightly better credit risk than these countries’ bonds.

22

3

Chart 3.8: Hypothetical shape of UK government bond yield curve, if the UK were to join EMU1

2.5

3.0

3.5

4.0

4.5

5.0

5.5

252015105 30

Per cent

Years to maturity

UK New UK Germany

1 As of December 2002.Source: HM Treasury.

5 Until now the liquidity premium on government bonds has been ignored. However, as mentioned in the introduction,government bond yields include a liquidity premium, reflecting the risk of being unable to find a buyer or seller on themarket.6 Against 14 per cent for Germany, 17 per cent for Italy and 13 per cent for France.

23

3.21 If the UK were to enter EMU, the factors above suggest gilts might become thebenchmark on bonds of a maturity of over 15 years, leading to a negative differential betweenbund and gilt yields at the long end of the bond market. UK yields might be higher thanGermany’s at the ten-year maturity area due to the liquid bund futures contract, which givesbunds benchmark status in this maturity area.

3.22 For the purposes of illustration, Chart 3.8 projects the UK yield curve if the UK entersEMU on the assumption that over the one-to-ten year maturity period UK yields would fallfrom their current levels, as expectations of short rates harmonise to those of the euro(proxied by German bond yields).7 UK yields are then projected to rise relative to their currentlevels out to 30-years as UK pension and life funds sell long gilts in order to diversify into longeuro government bonds. In essence, upon joining EMU, this projection suggests that the UKyield curve would steepen. This projection is based on the assumption that UK entry has noimpact on bund yields. If this does not hold, then the change in the UK yield curve would beless strong than projected (demand for long duration euro area sovereign bonds from UKfunds could result in the flattening of the euro area yield curve).

3.23 A steepening of the UK yield curve, such as is projected here, would lead to an increasein yields on medium and long-dated gilts, and a decrease for short-dated gilts. The impact ofthis on the cost of new corporate debt will be dependent upon corporate issuance strategy. Ifcorporate fundraising is focused at the short end of the curve, then the reduction in short-term yields would provide an opportunity to reduce the cost of financing, assuming thatcorporate issuance is priced at a constant spread over the government rate. These benefitswould be tempered if corporates issue along the medium and long parts of the curve where,on this projection, yields would rise if the UK joins EMU. While most UK corporates issuebonds at or below 15-year maturities in sterling, a significant number (46 per cent) are issuedat maturities over 15 years. In contrast, the average maturity of euro-denominated corporatebonds, whether from the UK or all issuers, was five years.8 This could reflect corporates takingadvantage of the relative positions of euro area and UK government bond yield curves,although corporates issuing at long maturities face credit risk premia that heavily influenceyields.

3.24 As there is currently substantial issuance at the shorter end, this projection suggeststhat EMU entry could lead to a lower cost of capital, at least in nominal terms. Plus, thecurrent pattern of issuance in euros and sterling suggests that corporates would react to a fallin UK government yields at the short end by increasing issuance at shorter maturities andmoving away from long-maturity issues. The benefits of this would have to be traded offagainst the increase in roll-over risk (i.e. the risk involved in continually issuing new debt)that might arise from this strategy.

3.25 In a historical context the possible gains from a fall in UK short-term yields describedhere are not large. For example, Chart 3.3 shows that over the 1990s, long-term UK governmentbond yields fell from around 12 per cent in 1990 to current levels of around 4.5 per cent.

3.26 Moreover, the current relative position of euro area and UK short-term interest rates islargely determined by the cyclical position of the economy. While current yields point to anominal decline in short rates if the UK were to enter EMU, this picture will change with theprevailing economic circumstances. Furthermore, this would only represent a decline innominal yields. There is little to suggest that the real short-term risk-free rate in the UK woulddecline in EMU. The EMU study Policy frameworks in the UK and EMU by HM Treasury,explains that the inflation targets of the monetary authorities in the UK and euro area are


Implications forthe corporatecost of capital

7 The hypothetical UK gilt short-rates under EMU entry in this chart (3.9 per cent for 5 year bonds) are still below euroshort rates currently observed in the market for 5 years’ time (4.8 per cent, see Chart 3.9), meaning that Chart 3.8assumes that the current market predictions of euro short rates are not fulfilled.8 Data as of end 2002.


probably very close. This means there is unlikely to be a significant difference in the currentexpected rate of inflation whether the UK remains outside or were it to join EMU. Equally, themacroeconomic reforms of 1997 are expected to maintain low and stable inflation in the UK,which means that there is likely to be little significant difference in the inflation risk premiumfactored into the rate of return on UK and euro area assets.

3.27 These issues can be explored further by using an alternative measure of expectationsof government bond yields – the forward yield curve. Chart 3.8 plots a spot rate curve – theinterest rate applicable today on government bonds maturing on dates from one year to 30years ahead. The forward curve in Chart 3.9 plots the implied short-term interest rate that themarket expects to apply across a range of dates from six months to 30 years ahead. The twocurves are closely related. The yields in the spot rate chart, for example for ten years ahead,will be the average of the rates in the forward curve from now until ten years hence. Forwardrates are not known with certainty so there will often be a ‘term premium’ between forwardand short rates, reflecting perceptions of the risk that actual short rates will be different fromforward expectations.

3.28 On the forward curve, the point at which UK and euro rates cross and UK short ratesbecome lower than euro rates is around five years. This compares to around ten to 15 years inthe spot curve. The higher UK forward rates until around five years in Chart 3.9 contribute tohigher UK spot rates to around ten years in Chart 3.8 (as future spot rates are the average offorward rates). Out to around five years expectations of higher UK rates may reflect cyclicalfactors. However, the market is unlikely to have a view on relative cyclical positions furtherthan five years from now. Instead, the forward curve at longer time horizons may reflectstructural factors related to the UK gilt market, such as are discussed above. In short, thisanalysis suggests that there is no expectation that the current cyclical factors leading to lowereuro area short-term rates relative to the UK will be sustained into the longer term.

24

3

1Instantaneous forward rates as of December 2002.Source: Bank of England.

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

6.5Chart 3.9: UK and euro area government forward curves1

5 10 15 20 25

Per cent

Years to maturity

Euro area UK

Using forwardrates to compare

governmentyields

25

3.29 These comparisons of UK and euro area government bond yields are used on theassumption that government bonds are an accurate proxy for risk-free rates. However, asdescribed above, supply and demand in the gilt market is currently influenced by UK-specificinstitutional factors, such as inelastic demand from pension funds. It can be argued that thesefactors depress the yield on gilts below the ‘true’ credit risk-free rates, making UK governmentbond yields a less accurate proxy of risk-free rates. Government yields may also be lower thanthe risk-free rate to the private sector because holding bonds can facilitate cheaper borrowingin the repo market.

3.30 The swap rate is the rate at which a variable interest rate cashflow is swapped for a fixedrate. Cooper and Scholtes (2001) argue that swap rates in the inter-bank market act as a moreconsistent measure of the credit risk-free rate, as the market is very large and does not facethe supply constraints seen in the UK government bond market. Moreover, it is easy to takelong or short positions using swap contracts, and there are no obvious regulatory distortionsaffecting swap pricing. Swaps do however bring some exposure to counterparty and generalbanking credit risk, though this is generally small in the inter-bank market.

3.31 Using this alternative measure of the risk-free rate reduces, but does not eliminate, thedifferences between UK and euro risk-free rates. Comparing the forward curve for UK andeuro swap rates (Chart 3.10) with the forward curve for government bond yields (Chart 3.9),indicates that the gap between long-term rates is much reduced in the swap curve. This is dueto the removal of much of the distortion to long-term gilt yields from the institutional factorsdiscussed above.


Swap rates as analternative proxy

for risk-freerates

Note: Swap forward rates as of December 2002.Source: Bank of England.

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

6.5

Chart 3.10: UK and euro area bank liability (swap) forward curves

Per cent

Years to maturity

Euro area UK

5 10 15 20 25

Conclus ion on the impact o f EMU entry on the UK credi tr isk- free rate

3.32 Through macroeconomic reform, the UK has established a credible low-inflationenvironment which has contributed to a steady decline in long-term credit risk-free rates.This means EMU entry would not offer the UK the significant falls in nominal credit risk-freerates seen in EMU countries with historically high inflation. It is likely that if the UK were toenter EMU, there would be close convergence of the UK government bond yield curve withthat of existing EMU countries. This may lead to a fall in short-term UK yields and a rise inlong-term yields. However, these changes are relatively small in historic terms. Moreover, thecurrent differential in short-term yields is largely a result of the current cyclical positions ofthe UK and euro area economies. It is difficult to argue that it would translate into asignificant fall in the real cost of capital for UK firms.

26


The size and efficiency of financial markets will affect the size of the market risk premiumcomponent of the cost of capital. There are signs that the euro has prompted increasedintegration in the euro financial market. However, there is not yet sufficient evidence to beable to determine the impact of this on the market risk premium.

Large UK firms can access the euro market from outside EMU at relatively low cost.However, the removal of exchange rate risk and transactions costs that EMU entry wouldbring, alongside the removal of some institutional constraints on foreign currency holdings,would increase access at the margin.

The potentialimpact of EMUon credit risk

4 IM P L I C AT I O N S O F EMU EN T RY F O R

T H E MA R K E T R I S K PR E M I U M

27

4.1 This section considers whether EMU has led to a reduction in the market riskpremium component of the cost of capital. There are two key ways in which EMU mightreduce the market risk premium. First, if EMU encourages investors to diversify portfoliosacross national borders this could reduce the credit risk component of the market riskpremium. Second, liquidity risk, which is the risk of being unable to find a counterparty fora transaction, may be lower in a larger and more integrated market.

4.2 As Stulz (1999) explains, in a world of fully-segmented national capital markets, eachcountry’s investors would have to bear the full risk of their country’s economic activity. Themarket risk premium they would demand for this would increase with national market risk.With capital market integration, domestic investors are able to diversify their risk profile byholding foreign assets. Because some of the credit risk of domestic and foreign assets is likelyto offset each other, investors can hold an international portfolio which has the sameexpected returns as previously but with lower risk. Therefore investors will demand lower riskpremia from firms when they are able to diversify. In practice, investors who invest purely intheir national securities markets are likely to enjoy some exposure to the internationaleconomy, as some domestically-listed firms will operate across borders.

4.3 Aside from potentially reducing investors’ credit risk exposure, EMU may reduceliquidity risk, which represents the risk of being unable to find a buyer or seller for an assetat a reasonable price. Liquidity risk in an integrated EMU financial market is likely to be lowerthan in a segmented national market, simply because the size of market is larger and therange of market participants is broader. Lower liquidity risk should translate into a lowermarket premium and so a lower cost of capital for firms.

4.4 The gains from financial market integration are potentially large.1 This is why it isimportant that financial market integration in the EU continues to develop, whether the UKis inside EMU or not, through efforts such as the Financial Services Action Plan (FSAP).Recent analysis for the European Commission (London Economics, 2002) suggests that fullintegration would lead to a fall in the cost of EU equity capital of around 50 basis points, andof the cost of EU debt of around 40 basis points, which it is estimated would boost the levelof EU GDP by 1.1 per cent in the long run.

4.5 The question addressed in this section is whether were the UK to enter EMU, therecould be additional cost of capital savings for UK firms through the market risk premium.This clearly depends on the degree to which a separate exchange rate acts as a barrier tointegration. Research generally suggests other factors are more important constraints, butthat at the margin the exchange rate is likely to act as a barrier.

The potentialimpact of EMU

on liquidity risk

The potentialgains from

marketintegration

Will EMUpromote

financial marketintegration…

1 See Stulz (1995) and (1999) for further analysis of the potential gains from international financial market integration.

IM P L I C AT I O N S O F EMU EN T RY F O R T H E MA R K E T R I S K PR E M I U M

4.6 For example, a puzzle in financial economics is that investors hold financial portfolioswhich are heavily weighted to domestic assets, and have much lower levels of internationalassets than optimal portfolio theory suggests they should. This home bias means they seemto miss out on the benefits of international diversification discussed above.

4.7 There are a number of reasons why EMU might reduce home bias among euro areainvestors. The removal of exchange rates within the euro area removes a transaction cost toinvesting overseas; and it also reduces exchange rate risk on European investments.Furthermore, many investment funds are constrained by regulation to keep a certainproportion of their assets denominated in domestic currency. The institutional set up ofEMU, such as the European Central Bank’s (ECB) cross-border money market paymentsystem (TARGET), may also promote integration.

4.8 It is possible that UK EMU entry would provoke a one-off change in net flows in linewith the one-off reduction in some of the barriers to cross-border investment, as institutionalinvestors diversify into UK assets and UK institutions move into euro area assets. If UKinstitutions have already diversified more than their euro area counterparts, flows into the UKmight exceed out-flows. Where investors are holding UK assets as a result of currencydiversification, there may also be a one-off impact if joining EMU provokes a sell-off of someof these assets.

4.9 Research has found that currency risk on its own does not explain home bias (forexample, see Brealey et al., (1999). Uncertainty about exchange rates can be hedged througha variety of financial products. Currency risk may not be undesirable; it may actually enableinvestors to diversify risk in their portfolio. A more likely explanation for home bias is thatthere are significant transaction costs and information costs to holding foreign assets. EMUin itself will only reduce transaction costs to the extent that it removes currency exchange andhedging costs; and it will only reduce information costs through the increased pricetransparency that comes from having a single currency. EMU could indirectly promotegreater integration if it acts as a catalyst for the development of a more efficient regulatoryenvironment in the EU financial markets, or promotes integration of financial marketinfrastructure, such as trading and settlement systems. Both these developments would belikely to lead to lower cross-border transaction costs.

4.10 The impact of integration on the cost of capital is not all one way. If EMU promoteseconomic convergence in Member States, this could reduce the benefits of portfoliodiversification. If national economies move together more closely, then domestic and foreignrisk is less likely to be offset. In such circumstances, investors may find that portfolioallocation strategies based on sectoral rather than country distributions are superior, an issuewhich is examined in further detail in later sections.

4.11 To address these issues, this section first examines the trends towards integration inthe euro area financial markets which could lead to lower credit and liquidity risk. It thenlooks at two important indicators of integration, transactions costs and portfoliodiversification, which could help to bring down the cost of capital in the euro area. Thesection ends by comparing estimates of the market risk on bonds and equities in the euroarea and the UK, as a measure of whether the UK could benefit from a reduction in marketrisk by joining EMU.

Integrat ion o f euro area f inanc ia l markets4.12 A number of indicators suggest the euro financial market has become more integratedsince EMU, though regulatory and structural constraints are still a hurdle to many cross-border transactions. The remainder of this section examines the key indicators in more detail(for more detail see Bank of England, 2002).

28

4

Structure of theremainder of

this section

…related tohome bias

29

4.13 The introduction of the euro is generally seen to have had a positive impact on theeuro-dominated corporate bond market. The market has grown strongly since 1999 (seeChart 4.1), driven primarily by issuances of commercial bonds, although these have mostlybeen issued by the financial sector. The average size of issues has also increased, as has therange of credit ratings offered. Factors other than the introduction of the euro have played apart in these trends – other international markets also witnessed increases in issuance in1999. One important factor was a market rebound after the Russian debt default and theLong-Term Capital Management (LTCM) crisis in late 1998. In addition, the sector received aboost from several large issues from the telecoms sector, used to finance the purchase of thirdgeneration licences. However, in 2002 the corporate bond market fell back due to thedeteriorating global economic environment.

4.14 One area in which EMU may have a significant impact on the cost of capital is byincreasing access to finance for smaller, higher risk firms. A well-developed market for high-yield bonds would increase access and potentially lower the cost of capital for firms whichmay not previously have been able to borrow on the bond market.

4.15 The evidence thus far on the development of the euro high-yield market is notconvincing. Table 4.1 presents the volume of euro-denominated bond issuances by rating.This indicates an increase in the share of the middle range of issues with credit ratings of Aand BBB. An important factor behind this has been the large volume of issues from thetelecoms companies, as well as some downgrading of previously higher rated bonds. Bycontrast, there has been little sign of much activity at the high-yield end of the market to date.It may be that UK entry to EMU would have beneficial spillover effects for smaller, higher riskfirms in the euro area as the UK has a relatively well-developed venture capital industry (seeSection 5).

IM P L I C AT I O N S O F EMU EN T RY F O R T H E MA R K E T R I S K PR E M I U M4Developments in

EMU bondmarkets

Chart 4.1: Gross issuance of international bonds and notes

Source: Bank for International Settlements (BIS), 2002.

SterlingEuro US dollar

US$ billions

0

50

100

150

200

250

300

350

2001 Q42000 Q41999 Q41998 Q41997 Q41996 Q41995 Q41994 Q41993 Q4

High-yield bondsand venture

capital


Table 4.1: Shares of euro-denominated bond issues by rating of issue(based on volume)

Per cent Bond rating

Year AAA AA A BBB BB B CCC

1996 64.5 25.6 8.3 0.7 0.3 0.7 0.01997 58.5 28.5 9.2 1.2 0.8 1.5 0.01998 55.0 25.0 15.5 2.1 1.2 1.4 0.01999 35.8 26.9 25.7 9.5 0.3 1.5 0.22000 40.3 27.3 27.0 3.9 0.2 1.0 0.320011 32.4 24.0 35.3 7.9 0.7 0.7 0.11Up to third quarter.Note: Data complied from Thomson Financial Securities’ IFR Platinum bond database, which covers mainly international issues which are self-described assuch in the prospectus, stating that the issuer targets investors outside its home country, or that are denominated in a different currency than the homecurrency of the issuer, or that the issue is underwritten or co-managed by an international investment bank. See Santos and Tsatsaroni (2002) for moreinformation.Source: Santos and Tsatsaroni, 2002.

4.16 Equity issuance in the euro area rose sharply in 1999 and 2000 compared to the pre-europeriod. New equity worth $86 billion was issued in 2000 Q1-Q3, compared with $108 billion infull year 1999 and an annual average of $56 billion between 1995 and 1998.2 Factors aside fromthe introduction of the euro explain this increase. Until 2000, the value of European equitymarkets was rising sharply, mirroring those in the US, and this encouraged equity issuance. Inaddition, a substantial proportion of new issuance was in the telecoms and technology sectors,propelled by rapid worldwide growth in these areas over the past couple of years. Increasingmerger and acquisition (M&A) activity in the euro area (which is, in part, a consequence of theintroduction of the euro) has also led to increased equity issuance. Since 2000, equity marketsin the euro area, US and UK have fallen substantially, especially in the telecoms andtechnology sectors, confirming the importance of sectoral developments in the market, andmaking it all the more difficult to isolate any effect from the start of EMU.

4.17 The growth of new equity issuance coincided with the continued expansion of EUmutual funds: in 1992 EU mutual fund assets equalled 8.2 per cent of EU GDP, by 1998 thishad risen to 31 per cent.3 The growth expected in personal savings through pension funds andmutual funds should continue to provide finance available for investment in the euro areaequity market in the future.

4.18 Running alongside the general process of financial market integration has been arestructuring and rationalisation of the banking and financial services industry. This has notonly been in response to the euro, but to rapid developments in liberalisation, technology andglobalisation in the financial services industry. There have been a significant number of majorbanking mergers and mergers between banks and other financial companies such asinsurance firms; the merger between Allianz and Dresdner is a recent example. These issuesare considered in more detail in the EMU study by HM Treasury The location of financialactivity and the euro.

4.19 Most of these mergers have been between companies from the same country. Cross-border mergers have so far been limited. The wholesale banking sector is already veryconcentrated, limiting the possibilities for cross-border mergers. More scope exists for cross-border mergers between commercial banks, and a few of these have taken place, butsignificant barriers remain:

• legal and regulatory rules vary considerably between countries;

• for cultural reasons, customers may prefer local banks;

30

4

Developments inEMU equity

markets

Developments inEMU financial

institutions

2 Bank of England (2000).3 Bank of England (1999).

31

• savings may not be as great as for domestic mergers where overlappingservices can be reduced; and

• availability of information on credit history and current financialcharacteristics may be limited.

4.20 For example, in Germany the complex cross-ownership structure of the major bankslargely insulates them from the threat of consolidation. The tax position of capital gains alsoaffects this structure.

4.21 Increased concentration in financial services could take the market in one of twodirections. If mergers are driven by efficiency gains then they should result in lower costs and,if these savings are passed on to customers, to a reduced cost of capital. Cross-border mergersmay increase perceived contestability in European markets, which should also reduce thecost of capital. On the other hand, concentration might reduce competitive pressures, inwhich case margins could rise, leading to an increase in the cost of capital. Concentrationmay also have different effects on different market participants. If the development of largecommon trading platforms makes it more difficult for smaller issuers to tap the euro markets,their costs could rise.

4.22 Corvoisier and Groop (2001) examine the impact of deregulation and concentration inthe EU banking market over the period 1993 to 1999. They find that M&A activity has increasedbanking concentration over this period, and that this increased concentration is correlatedwith higher interest margins for bank loans. This contrasts with the market for savings depositsand time deposits, where increased concentration is correlated with lower margins. Corvoisierand Groop argue that the market for loans is more open to collusion as it is an informationintensive product. Firms familiar with the local market may have a comparative advantage thatallows them to extract a rent and reduce competition.

4.23 The ECB (2000) finds that competitive pressures have forced euro area bank lendingmargins down, for both the household and corporate sectors. The ECB concludes that inperiphery countries, such as Ireland, Italy, Portugal and Greece, part of this fall is due to theconvergence of market rates to EMU levels. The ECB believes an increase in bankingcompetition is also an important factor. There are significant levels of foreign entrants tobanking in Ireland, the UK and Nordic countries in particular, with entry facilitated by newtechnology such as Internet banking. In so far as this indicates low barriers to entry already,the additional competitive impact of EMU membership would be muted.

4.24 Financial market infrastructure has been highlighted as a significant factor affecting thedevelopment of the euro financial market. Without a well-functioning euro financial marketinfrastructure it would be difficult for the UK, whether in or out of EMU, to benefit from thelower cost of capital associated with lower credit risk and liquidity risk premia. There are threemain infrastructure elements: trading platforms, clearing houses and settlement systems.

4.25 Trading platforms bring together buyers and sellers to agree on the volume and priceof transactions. Integration of trading platforms could reduce liquidity risk by increasingmarket size; potentially it could also reduce transaction costs through economies of scale. Inthe equity markets, the main development since the start of EMU has been the establishmentof Euronext, the merger of the Paris, Brussels and Amsterdam stock exchanges. ThePortuguese exchange has now joined Euronext; and in early 2002 Euronext purchased theLondon derivatives exchange, LIFFE. In 1999, the Stockholm and Copenhagen exchangesagreed to create a common trading platform. They have subsequently been joined by Osloand Reykjavik. In the bond market, Euro-MTS is a common platform which has been createdfor trading the government bonds of Belgium, Finland, Germany, Greece, Spain, France, Italy,Ireland, Portugal, the Netherlands and Austria.

IM P L I C AT I O N S O F EMU EN T RY F O R T H E MA R K E T R I S K PR E M I U M4

Developments ineuro financial

marketinfrastructure


4.26 A clearing house’s main function is to assume and manage risk by acting ascounterparty to both sides of a transaction. If the buyer cannot pay or the seller cannotdeliver, the clearing house has to fulfil the obligations of the defaulting party. Clearing houseshave traditionally provided their services for derivatives markets. However, recently some,including London Clearing House (LCH) and Clearnet, have also started to act as centralcounterparty for cash markets.

4.27 Finalisation of a transaction requires settlement, which involves the actual transfer of cashand securities between participants. Before EMU, European settlement systems were based on aframework of national central security depositories (CSD) in each Member State, plus twointernational depositories (ICSD), which were originally established to settle eurobonds. Twoseparate mergers so far have taken place: Cedel, a ICSD based in Luxembourg, merged withDeutsche Börse Clearing, the German CSD, to form Clearstream (with Deutsche Börsesubsequently buying out the Cedel shareholders); while Euroclear, which settles trades forEuronext’s exchanges, acquired the national CSDs of France, Belgium and the Netherlands, andmore recently merged with London’s Crest. The new Euroclear Group will handle about 60 percent of trades in leading European equities and more than 50 per cent of fixed income trades.4

4.28 In addition to these three main infrastructure elements, there are also paymentsystems for interbank unsecured money market transactions. TARGET, the ECB’s cross borderpayment system, and the Euro Banking Association’s (EBA) EURO1 are the two pan-Europeaneuro payment systems which, by allowing commercial banks to manage euro liquidity in asingle pool, are contributing to the integration of euro trading and settlement in this area ofthe financial markets.

4.29 In its first report on clearing and settlement, the Giovannini Group5 claimed that whilethe settlement of domestic trades in the EU was similar to the analogous service in the US, thesettlement of cross-border trades was substantially less efficient than the settlement ofdomestic trades. The Group concluded that there were barriers to the efficient delivery ofcross-border settlement services arising out of differences in technical requirements/marketpractices, tax regimes and legal systems across Europe.

4.30 There is an ongoing debate about whether EMU infrastructure should move in thedirection of concentration or competition. Supporters of concentration argue that networkexternalities are generally seen at all three levels of the infrastructure: trading, clearing andsettlement. At the level of trading, a dominant exchange pools liquidity. This should make themarket-place more efficient by reducing the spreads on individual stocks, ensuring thatprices properly reflect supply and demand, and making the market more robust in the face ofshocks. On the other hand, a single concentrated supplier may be less efficient than if therewere competition between several providers of such services, particularly as competition canco-exist with liquidity. Most traders now have access to technology that allows each individualtrade to be routed to the execution venue where they will obtain the best price.

4.31 At the level of clearing, a dominant clearing house could, for example, help users saveon capital by taking account of offsetting risks across a wide range of trades in setting marginrequirements. However, a dominant clearing house would concentrate counterparty risk inone institution. It would need to manage its risk very effectively to sustain the stability of thefinancial system. At the level of settlement, the London Stock Exchange argued in its responseto the Giovanni Group’s report6 that currently there was no meaningful competition betweenCSDs. However, technological change and potential liberalisation of national laws, whichhave restricted where securities may be settled, might encourage competition with theexisting and dominant CSDs.

32

4

4 Based on pre-merger volumes.5 European Commission, DG Economic and Financial Affairs (2002).6 London Stock Exchange (2002).

33

4.32 There is also debate over whether infrastructure should be horizontally or verticallyintegrated. A horizontal structure implies that there is separation of the ownership of trading,central clearing and settlement. A vertical structure implies common ownership of all threeelements of market infrastructure.

Impact o f market deve lopments on transact ion costs andport fo l io d ivers i f i cat ion4.33 As already suggested, EMU may promote the development of a more integratedfinancial market and thus a lower risk premium due to a lowering of transaction costs, theremoval of currency costs and risk, and increased competition in a larger market. An indicatorof the degree of integration in the euro market is the extent to which investment funds arediversified across national borders. Therefore, this section considers the impact of recentmarket developments on transaction costs, hedging costs and portfolio diversification.

4.34 Research suggests that the developments in the EMU financial market outlined abovemay have helped to reduce market transaction costs. At the start of EMU, transaction costs forcross-border financial trades in the euro area were significantly higher than domestic ones.Adjoute et al. (1999) estimate that cross-border trades cost ten to twenty times more thandomestic ones in the EU. They attribute this to the complex and inefficient structuring of cross-border payments, the lack of a delivery versus payment system, which increases transactionrisk, and the longer time and greater custody risk involved between trade execution andcompletion due to the number of intermediaries involved. According to a report on the benefitsof financial integration for the European Commission (London Economics, 2002) the cost ofequity capital could fall by around 50 basis points if there was full European financialintegration, mainly through a lowering of transaction costs.

4.35 Murray (2001) asserts that the US enjoys significant cost savings from operating asingle equity settlement and clearing house, compared to the 26 which operate in the EU.There is some evidence of falling equity commissions. In the euro area, Mann and Meade(2002) find that pan-European commissions for domestic equity transactions fell andconverged between 1997 and 2001.

4.36 EMU may also lead to a reduction in transaction costs through increased competitionin the retail and investment banking sector,7 though as discussed above, increased bankingconcentration following the wave of euro area bank mergers may offset any increase incompetition. In the euro-denominated bond market, evidence from BIS (2000) suggests thatthere has been an important decline in underwriting fees. Before EMU, the bond market waslargely segmented. Investment banks faced relatively high costs of entry into each market, asthey required high levels of analytical expertise relative to the size of the market. Now theEMU bond underwriting market can be seen as more contestable.

4.37 BIS (2000) estimate that the average fee for international bonds denominated in eurosfell by around 1 percentage point (as a percentage of the nominal amount of the issue) overthe period 1997 to 2000 – very close to average gross fees in the US dollar market. Santos andTsatsaronis (2002) also find that underwriting fees for euro-denominated corporate bonds arenow very close to those of US dollar-denominated bonds.

4.38 There has been some evidence of increased transactions costs in the foreign exchangemarkets since the introduction of the euro (Detken and Hartmann, 2002), with somecommentators suggesting that the single currency increased transparency in spot trading andthereby inventory risk, forcing dealers to increase spreads. However, it appears likely that thisincrease has appeared as a result of quoting conventions on spreads, which meant that a pip8

became a larger proportion of nominal exchange rates on the dominant dollar/euro trade.


Transaction costs

7 See Jean Dermine’s contribution to the EMU study Submissions on EMU from leading academics.8 A pip is the lowest spread currently possible – one hundredth of a US cent per euro.


4.39 Accessing euro area markets may be cheaper for UK firms if the UK were in EMUbecause of the removal of hedging costs. There would also be ‘in-house’ savings in the formof lower expenditures on specialist personnel engaged in the management of exchange raterisk between the two currencies.

4.40 Exchanging foreign currency can be done either at the spot rate or at a forward rate inorder to hedge against unexpected currency movements. The cost to a firm of spottransactions is the spread between the market rate and the rate available to them through abank, plus any commission charged. For large deals this spread could be next to nothing, oreven used as a loss leader. Unlike spot rate transactions, the cost of forward rate transactionsmust also take into account credit risk, which will depend on the size of the deal and thecreditworthiness of the company concerned.

4.41 Small and medium-sized enterprises (SMEs) may find hedging costs more costly thanlarger firms as:

• they do not necessarily go to the best provider because of their relative lack ofknowledge and expertise;

• the size of deals they require means they have less leverage with the providerto negotiate a better deal; and

• for forward transactions, they may be thought of as worse credit risks thanlarger firms.

4.42 Section 5 considers the implications of UK entry to EMU for SME financing.

4.43 Evidence suggests that portfolio diversification in the euro area has been taking place,but perhaps not as rapidly as was generally expected before EMU. There are a number offactors which could explain the lack of full diversification. First, many countries have retainedregulations which restrict the foreign holdings of pensions and other investment funds.Second, as discussed above, the lack of a fully-integrated financial market infrastructuremeans there are still often higher transaction costs involved in cross-border activity. Third,fund managers may face informational costs when investing overseas, due to greaterknowledge of local markets. Finally, as selling existing holdings is likely to crystallise capitalgains, diversification may only take place with new money. Portfolio diversification may turnout to be a longer-term phenomenon, as firms gradually restructure their operations andproducts on a pan-euro area basis. Progress towards more integrated infrastructure may alsotake time.

4.44 Surveys of market players generally indicate that EMU investors are now focusing onEMU-wide allocation strategies. For example, Tsatsaronis in BIS (2001) reports that 75 percent of European equity fund managers believe that sector portfolio allocation strategiesdominate country strategies, while only 10 per cent believe country strategies dominate. In1997, the respective preferences were 20 per cent and 50 per cent – so there has been asignificant shift since EMU.

4.45 Analysis by the ECB (2001) also finds a shift in preference away from domestic marketstowards pan-European sector-based allocation. The results of their survey are shown in Table4.2. Domestic allocation remains high but has decreased since the introduction of the euro,while funds held in pan-European or euro area funds have increased from 12 per cent to25 per cent of the total. BNP Paribas figures, in Table 4.3, which look at both equity and bondallocation, confirm the ECB’s findings of a shift away from domestic holdings.

34

4

Portfoliodiversification

Hedging costs

35

Table 4.2: Assets of euro area equity funds by area of investment

Per cent

Year Domestic Euro area only Non-euro area Europe Global Total

1998 43 – 12 45 1001999 33 8 14 45 1002001 34 10 15 41 100Source: ECB, 2001.

Table 4.3: Assets of euro area institutional investors by area ofinvestment

Per cent Domestic Foreign Domestic Foreignbonds bonds equities equities

December 1998 73 27 62 38December 1999 66 34 57 43December 2000 62 38 52 48December 2001 57 43 50 50March 2002 55 45 49 51Note: Institutional investors here include pension funds, mutual funds, insurance companies and banks.Source: BNP Paribas.

4.46 The way in which portfolio allocation strategies develop in EMU will depend in part onthe degree to which EMU promotes convergence of national euro area equity returns. Theeuro area’s single monetary policy will lead to the convergence of risk-free nominal rates ofreturn across national markets; it may also promote real economic convergence, which couldlead to convergence of firms’ expected cash flow and therefore of equity valuations. Theremoval of currency risk also removes another source of country-specific risk. These factorsmean that euro area-wide risk factors may come to dominate equity returns, making country-specific factors less important. This may make the benefits of diversification across the euroarea less strong.

4.47 Several studies have used statistical techniques to examine EMU equity market trends.The results of these studies are summarised in Annex A. These studies use two basicapproaches: (1) analysing the factors which drive equity returns, and (2) analysing equityindices’ correlations. If the euro area is becoming more economically integrated, thenindividual equity returns will be driven less by country-specific factors and more by euroarea-wide and/or sector-wide factors.

4.48 The studies generally point to increased convergence of equity markets acrossdifferent euro area countries. This may lower the benefits of diversification on a country basisin EMU. However, the studies also generally suggest that EMU-wide strategies based onsectors are superior to those based on countries and based on domestic allocations. Thismeans that despite some economic convergence, there are still significant benefits to EMU-wide portfolio diversification.

4.49 There is evidence that EMU is promoting the development of a more integratedfinancial market in the euro area, though remaining regulatory and structural constraintsmean there is still a long way to go. Evidence already suggests that these developments arereducing transaction costs and promoting increased portfolio diversification.

4.50 These developments can be expected to lead to a lower market risk premium. In a moreintegrated market investors will be able to diversify credit risk to a greater degree than waspossible in the segmented euro area markets (though increased economic convergence mayreduce the size of these diversification benefits). Investors will also find it easier to buy and sell


The influence ofeconomic

convergence onportfolio

allocation

Conclusions onEMU market risk

and liquiditypremium


financial assets, and so liquidity risk should be lower. However, after just four years of EMU itis not possible to quantify any reduction in these risk premia that derives purely from theimpact of EMU.

4.51 The increased size of the market should allow for the expansion of segments of themarket, such as high-risk bonds, which were previously constrained by a narrow investorbase, though this is not yet apparent. A well-developed market for high-yield bonds wouldincrease access and potentially lower the cost of capital for firms that may not previously havebeen able to borrow on the bond market. Other things being equal, these developments couldlower the cost of capital for firms looking to raise finance in the euro area.

4.52 Certain sectors will benefit from these changes more than others. Firms from thesmaller euro area countries potentially have the most to gain from access to an integratedeuro area market. Previously they may have been limited to a narrow financial market with asmall investor base. SMEs from all countries may also have the most to gain. Largemultinational firms generally have the resources and reputation to allow them to tap intointernational financial markets. SMEs are generally much more restricted to their nationalmarkets. These special considerations mean that the impact of EMU on SMEs is consideredseparately in Section 5.

Potent ia l impact o f EMU entry on the UK market r iskpremium4.53 Evidence suggests that EMU has promoted the growth and integration of eurofinancial markets. This may lower the market risk premium for euro area countries. However,many UK firms would be able to access this market whether or not the UK enters EMU.Indeed, research suggests that exchange rate risk may not be as large a barrier to cross-borderfinancial market provision as other factors, though it is likely that at the margin, the removalof exchange rate risk would increase access to the euro market for UK firms.9

4.54 One pointer to the size of market risk premia is the spread on commercial bonds overgovernment bond yields of similar maturities. The spread reflects the risk which the marketattaches to the corporate bond. This will reflect overall market risk and the risk attached to theparticular firm. Comparing the spreads on national indices of many different companies canprovide some evidence on national market risk. However, as discussed in Section 2, UKgovernment bond yields may not be a good proxy for the credit risk-free rate, so the study alsolooks at the spread over swap rates.

4.55 Chart 4.2 shows the five to seven-year spread over their respective benchmarkgovernment bond yields of UK and euro area AA-rated issuers. UK issuers’ nominal debt costsover the period are higher than those for euro area ones, though this gap began to decline fromthe end of 2000. Charts 4.3 and 4.4 show similar patterns, with the UK A-rated and BBB-ratedbond spreads higher than the EMU spreads, with narrowing beginning around the end of 2000.10

4.56 It is possible that one reason why UK AA-rated institutions have higher spreads thantheir euro area counterparts is lower market risk in the euro area. However, it is difficult toisolate this effect, if it is present, from other factors affecting corporate spreads, such as thegeneral economic outlook in each country, which will determine future profitability and therisk of bankruptcy. In addition, UK credit spreads may also have been wider because of therelative scarcity of government bonds in the UK and the demand for gilts from pension fundsdiscussed in Section 3. This makes it difficult to draw any firm conclusions on the basis ofthese spread indices.

36

4

Comparing UKand euro area

aggregatecorporate bond

spreads usinggovernment

rates

9 For example, see Brealey et al. (1999).10 Spreads of corporate bond yields over government bond yields may also be affected by the cyclicality of spreads, withspreads tending to fall as bond yields fall and vice versa.

37


Source: Morgan Stanley Capital International (MSCI).

0

20

40

60

80

100

120

140

Chart 4.2: Spreads of UK and euro area AA-rated bonds over government rates (5–7 year)

Sterling Euro

Basis points

Dec-98

Apr-9

9

Aug-9

9

Dec-99

Apr-0

0

Aug-0

0

Dec-00

Apr-0

1

Apr-0

2

Aug-0

2

Dec-02

Aug-0

1

Dec-01

Source: MSCI.

0

20

40

60

80

100

120

140

160

180

Chart 4.3: Spreads of UK and euro area A-rated bonds over government rates (5–7 year)

Sterling Euro

Basis points

Dec-98

Apr-9

9

Aug-9

9

Dec-99

Apr-0

0

Aug-0

0

Dec-00

Apr-0

1

Apr-0

2

Aug-0

2

Dec-02

Aug-0

1

Dec-01


4.57 As discussed in Section 3, it is possible to remove the distortion associated with thespecific influences on the UK government bond market by instead examining the corporatespread over swap rates. Chart 4.5 indicates that AA-rated sterling corporate bonds trade at ahigher spread over the swap rate than euro area corporates. This is in line with the findings ofCharts 4.2 to 4.4 above, which looked at the spread over government bonds. However, for Aand BBB-rated corporate bonds, there is no clear cost advantage for euro issuance. Theaverage maturities of sterling bonds are higher than those of the euro bonds in Charts 4.5 to4.7. As noted previously, longer maturity bonds would be expected to have a term premium,leading to larger spreads.

38

4

Source: MSCI.

0

50

100

150

200

250

300

350

400

450

Chart 4.4: Spreads for UK and euro area BBB-rated bonds over government rates (5–7 year)

Dec-98

Apr-9

9

Aug-9

9

Dec-99

Apr-0

0

Aug-0

0

Dec-00

Apr-0

1

Apr-0

2

Aug-0

2

Dec-02

Aug-0

1

Dec-01

Sterling Euro

Basis points

Comparing UKand euro area

aggregatecorporate bond

spreads usingswap rates

Note: Sterling AA-rated bonds had an average maturity of 7.7 years. Euro AA-rated bonds had an average maturity of 4.5 years.Source: Merrill Lynch.

0

10

20

30

40

50

60

70

Chart 4.5: Spreads of UK and euro area AA-rated bonds over swap rates

May-02

Jun-02

Sep-0

2

Oct-02

Jul-02

Aug-0

2

Jan-02

Feb-0

2

Mar-02

Apr-0

2

EuroSterling

Basis points

0

50

100

150

200

250

300

Note: Sterling BBB-rated bonds had an average maturity of 6.6 years. Euro -rated bonds had an average maturity of 3.5 years.Source: Merrill Lynch.

EuroSterling

Basis points

Chart 4.7: Spreads of UK and euro area BBB-rated bonds over swap rates

May-02

Jun-02

Sep-0

2

Oct-02

Jul-02

Aug-0

2

Jan-02

Feb-0

2

Mar-02

Apr-0

2

39


Note: Sterling A-rated bonds had an average maturity of 8.1 years. Euro A-rated bonds had an average maturity of 4.4 years.Source: Merrill Lynch.

0

20

40

60

80

100

120

140

May-02

Jun-02

Sep-0

2

Oct-02

Jul-02

Aug-0

2

Jan-02

Feb-0

2

Mar-02

Apr-0

2

EuroSterling

Basis points

Chart 4.6: Spreads of UK and euro area A-rated bonds over swap rates


4.58 Another potential measure of the credit risk premium attached to euro area bonds canbe seen in Chart 4.8, showing the credit default swap (CDS) indices. In theory, CDS shouldequal the difference between the yield and the credit risk-free rate on the underlying bond.The CDS price can be regarded as the cost of insurance against default occurring on theunderlying bond, although there is no need to hold that bond in order to trade in creditdefault swaps. Chart 4.8 indicates that credit risk is lower in the euro area than in the UK andthe US. However, these indices do not compare the same ratings and sectors of issuers, somust be regarded with caution. Telecoms CDS have been excluded from these series, astelecoms are more heavily weighted in the euro index and have significantly higher spreadsthan other corporate CDS.

4.59 A problem with comparing aggregate indices is that differences may reflect thedifferent proportions of different sectors and types of issuers within them. It was for thisreason that telecoms were excluded from the aggregate in Chart 4.8. An alternative is tocompare directly the euro and sterling spreads over swap rates for specific companies thatissue in both currencies.

4.60 The charts in Annex B present spreads over swaps for five large UK firms: BritishTelecom, Gallaher, Pearson, HBOS and Lloyds TSB. This comparison suggests little consistentdifference between sterling and euro risk premiums. Some firms face a lower spread overswap rates (an alternative measure of the risk-free rate) when raising capital in euro, whileothers have a lower spread in sterling. So, although the aggregate picture shows no clear costadvantage to UK firms of issuing bonds in euro, it may be cheaper for individual firms to issueeuro-denominated debt. EMU entry would allow these firms to avoid the cost of swappingtheir debt into sterling, although this is likely to be a small cost for a large issuer, with thetransaction cost of a conventional swap ranging from approximately two basis points up tofive basis points, depending on maturity and size.

4.61 Another method for comparing the cost of capital in the UK and euro area markets is to lookat the equity risk premium in both areas. However, estimating the equity risk premium isnotoriously difficult. There are four ways of looking at this, as outlined in Ibbotson and Chen (2001):

40

4Credit defaultswap indices

Basis points

Euro areaUSUK

Apr-02Mar-02Feb-02Jan-02 Jul-02 Aug-02Jun-02May-020

50

100

150

200

250

Source: CreditTrade and Bank of England.

Chart 4.8: Credit default swap spreads, excluding telecoms(5 year)

Analysing therisk premium for

specific firms

UK and euroarea equity risk

premia

41

• using an historical time series of the difference between equity market returnsand government bond yields and extrapolating forward;

• using earnings, dividends and overall productivity measures to derive theexpected equity risk premium;

• surveying finance professionals; and

• deriving equities’ expected returns through the payoff demanded by investorsfor bearing the risk of equity investments, which relies on the CAPM modeloutlined in Section 2.

4.62 Comparative estimates have been produced using the first method, i.e. by looking athistorical evidence on the difference between equity market returns and government bondyields. Dimson et al. (2002) look at returns from 1900 to 2001 and estimate a lower equity riskpremium in the UK (5.5 per cent) than in France (6.7 per cent), Germany (9.6 per cent), theUS (6.7 per cent) and the euro area (7.5 per cent).11 However, there are problems with thisapproach. Actual equity returns may be lower than this method suggests because it ignoresthe impact of firms which go bankrupt on returns. In addition, there is no reason why the pastshould be any guide to current or future risk.

4.63 A recent survey of investment managers by Merrill Lynch12 looks at expectations offuture equity risk premia, and finds that the UK equity risk premium is around 3.5 per cent,compared to 4 per cent for the euro area and 4 per cent for the US. Credit Suisse First Boston(2000) estimate equity risk premia based on expected returns of 4.4 per cent for the UK and3.7 per cent for the US over the period 1995 to 2000.13 This suggests that membership of alarge, integrated financial market could bring a reduction in the risk premium. However,Dimson et al. (2002) also look at expected risk premia, and find that UK rates are lower (at 6.2per cent) than Germany (10 per cent), France (9.5 per cent) and the US (7.5 per cent).

Conclus ion on the impact o f EMU entry on the UK marketr isk and l iqu id i ty premium4.64 Evidence suggests that the euro capital market has become more integrated since thestart of EMU, though there is still much to be done before the market is fully integrated.However, after only a few years of EMU it is difficult to isolate an EMU effect on market riskpremia, either from looking at corporate bond spreads or equity risk premiums.

4.65 Moreover, many UK firms are able to access the euro market at relatively low cost.However, for smaller companies transactions and hedging costs are likely to be a significanthurdle to raising capital in the euro area, or indeed in any overseas market. Hedging costs willbe higher as a proportion of funds raised when financing needs are lower, as much of the costinvolved in acquiring the financial expertise needed to undertake such operations is fixed.This means that for SMEs UK entry to EMU could remove a quite significant barrier toaccessing the euro markets, potentially reducing the cost of capital for these companies.However, there would remain significant informational barriers and transactions costs tosmaller companies raising funds in EMU. These issues are discussed in more detail in the nextsection.


11 The euro area estimate is a GDP weighted, seven-country average (Germany, France, Italy, Netherlands, Spain, Belgiumand Ireland) derived from Dimson et al. (2002). The GDP data are for 2001, European Commission, DG Economic andFinancial Affairs (2001).12 Merrill Lynch (2002). Weighted average taken of fund managers views on appropriate equity risk premium for individualareas.13 Credit Suisse First Boston (2001) estimate an equity risk premium of 4.3 per cent for the UK and 4.1 per cent for theeuro area over the period 1995 to 2001. Comparable estimates for the US over this period are not available.

5 IM P L I C AT I O N S O F EMU EN T RY F O R

UK SME FI N A N C I N G

43

5.1 SMEs are defined as those firms with less than 250 employees. They constitute themajority of UK firms, accounting for 99 per cent of UK enterprises and 43 per cent of non-government employment.1 (Further details about the structure of UK industry and potentialimplications of EMU entry can be found in the EMU study EMU and business sectors, by HMTreasury.) Removing exchange rate transaction costs and currency risk through EMU entrymay be particularly beneficial to UK SMEs.

5.2 At present, all UK firms accessing euro capital markets will face search costs, currencytransaction costs and currency hedging costs. For large companies these costs can be spreadover high-value transactions and so will be fairly small as a proportion of finance raised. Butfor SMEs these costs are likely to be a significant hurdle to raising euro finance. As discussedin Section 4, hedging costs may be higher for SMEs than for large firms in relative terms.Information and monitoring costs are also an important factor in SME financing. Financeproviders to SMEs require local knowledge of firms and economic conditions, and oftenmonitor the use of finance after it has been provided. Chart 5.1 provides a breakdown ofsources of external finance for SMEs. By far the greatest source of external finance is banklending, accounting for 61 per cent of total external finance.

Chart 5.1: Sources of external finance for SMEs, 1997-1999

Note: Factoring provides business with finance against outstanding invoices. Asset-based finance includes leasing and hire purchase.Source: Bank of England, 2001.

Other sources7%

Banks61%Asset-based

23%

Venture capital1%

Trade customers1%

Factoring3%

Partner/shareholder4%

The impact of EMU entry on the cost of capital for small and medium-sized enterprises(SMEs) could be very different from that experienced by larger firms. SMEs tend to raisefunds locally and so are unlikely to access the euro bond market. They also tend to relymore heavily than large firms on bank lending and venture equity.

The reduction in currency costs on cross-border finance would be relatively moreimportant for SMEs. However, information and monitoring costs would be likely to remaina constraint on SME’s accessing the euro area financial market. Smaller SMEs in particulartend to use primarily local retail finance. Over the long run EMU entry could increasecompetition in the domestice UK market for bank lending to SMEs, and could increase thesize of the venture capital market.

1 Small Business Service (2002).

IM P L I C AT I O N S O F EMU EN T RY F O R UK SME FI N A N C I N G

5.3 Asset-based finance and factoring account for the second greatest share. Venture capitalis much less significant in volume terms, accounting for just 1 per cent of the total, but is aparticularly important potential source of finance for high-growth SMEs.

5.4 The importance of information in SME financing means that even if UK entry to EMUwere to reduce the costs of accessing financial markets for SMEs, the importance of localknowledge would remain a constraint on market access. This will particularly be the case forsmaller SMEs. This section considers in more detail the possible impact of EMU on banklending and venture capital.

Impact o f EMU on UK bank lending to SMEs

5.5 By removing exchange rate risk and boosting price transparency, EMU entry couldpotentially increase competition in UK banking by encouraging euro area banks to enter theUK market. The recent Competition Commission (2002) investigation into small businessbanking concluded that a complex monopoly exists in the supply of banking services to SMEsand identified practices carried out by the eight main clearing banks in the UK which restrictor distort competition and operate against the public interest. The report also proposed aseries of remedies, including measures to impove transparency and reduce the costs ofswitching accounts.

5.6 As outlined in Section 4, evidence from the euro area on the impact of EMU on bankingcompetition is mixed. The dominant trend in the euro area in recent years has been towardincreased concentration in the banking sector. Merger activity has predominantly involvedbanks from the same country and retail financial services in the euro area are still largelysegmented on national lines. Cross-border mergers, which the euro might have beenexpected to stimulate, have so far been limited mostly along distinct regional lines such asScandinavia (e.g. Nordea NB). However, there is some evidence2 of increasing cross-borderprovision of banking services, which may improve the competitive environment.

5.7 It is not surprising that cross-border retail banking services have been slow to developin the euro area, as national regulations act as a barrier and local knowledge of domesticeconomic conditions is very important in lending to SMEs. New entrants are also likely to beput off by reluctance on the part of SMEs to switch banks. The Competition Commission(2002) investigation found that SMEs were very reluctant to change banks because of theperceived complexity of switching for little financial benefit, the significance of maintainingrelationships with a particular bank or relationship manager, and the ability of the existingbank to negotiate lower charges if there is a threat of switching. The behavioural remedies aredesigned to tackle the barriers to switching, but will take some time to fully take effect, so inthe short term new entrants are unlikely to capture a sizeable share of the UK market.

5.8 Overall, it seems unlikely that EMU entry would lead to an immediate and significantincrease in competition in SME bank lending. The importance of local knowledge and thereluctance of SMEs to switch banks will constrain new euro area entrants into the UK market.This does not mean that EMU entry would not promote competition over the longer run.Despite the constraints outlined above, EMU entry could promote increased pricetransparency and so spur competition in the sector. However, the impact is unlikely to belarge in the short run.

44

5

2 European Central Bank (2000).

Impact o f EMU on UK venture capi ta l and other equityf inanc ing

5.9 Although representing a small component of the financing of SMEs, venture capital andrelated forms of equity financing can be a vital source of capital for high-risk and high-growthventures. Such firms may be unsuitable candidates for debt financing due to a lack ofcollateral, the lack of a track record, and long time horizons until positive cash flow isexpected.3 Equity financing through venture capital may be more suitable in these instances.

5.10 As well as providing finance, venture capitalists often have technical, financial andmanagement expertise which allows them to better judge higher risk projects than a banklender. Venture capitalists have an incentive to use any relevant expertise they may have toboost returns.

5.11 The UK venture capital industry is the largest in Europe, accounting for 29 per cent oftotal European venture capital investment in 2001. It accounted for the equivalent of 0.7 percent of UK GDP in 2001. Total funds raised by UK private equity firms in 2001 reached £12.7billion, up by 18 per cent from 2000. An important characteristic of the UK venture capitalmarket is that it is still skewed towards later-stage investments. Only around 9 per cent of totalfunds were allocated to start-up or early-stage projects in 2001, although this represents a 30per cent increase on 2000 levels (British Venture Capital Association, 2002). There is also astrong regional bias with the South East and London receiving 52 per cent of UK venturecapital investment.

5.12 In most of the rest of Europe the venture capital industry is much less developed than inthe UK or the US. The size of the sector has increased rapidly in recent years. Funds of €38.2billion were raised in Europe4 in 2001, compared with just €5.5 billion in 1995 (EuropeanVenture Capital Association (EVCA), 2002). The UK accounted for €20.5 billion of the 2001European figure. The market in the US is much larger, though since the bursting of theInternet bubble in 2000, funds raised from US venture capital have declined sharply. Theamount invested fell to $40.6 billion in 2001, from $106 billion in 2000 but remainssignificantly higher than the $7.37 billion invested in 1995.5 Moreover, of the US total, 22 percent went to early-stage investments in 2001.

5.13 Large UK institutional investment funds, in particular insurance companies, seemaverse to venture capital and to start-ups in particular. Those institutional investment firmswhich invested in private equity in the US in 2001 allocated 7.5 per cent of their funds to it,compared to just 3.7 per cent in equivalent UK funds. Indeed much of the growth in venturecapital in the UK in recent years has been from US institutions. That said, the amountinvested by UK pension funds in venture capital doubled from £0.8 billion in 2000 to £1.6billion in 2001.

5.14 Could financial integration in EMU promote the creation of a much larger venturecapital industry in the EU, perhaps similar to that in the US? In part, this depends on thefactors that make the US market so much larger than the EU.

IM P L I C AT I O N S O F EMU EN T RY F O R UK SME FI N A N C I N G5

3 For a review of SME financing issues and trends in the UK, see Lund and Wright (2001).4 Data are for the UK, France, Germany, Italy, Spain, the Netherlands, Switzerland, Belgium, Denmark, Finland, Norway,Ireland, Poland, Austria, Portugal, Hungary, Greece, the Czech Republic, Iceland and Slovakia.5 These figures are from the US National Venture Capital Association, which uses a narrower definition of venture capitalthan the EVCA.

The UK and EUventure capital

industries

45

The implicationsof EMU for

venture capitalin the EU

IM P L I C AT I O N S O F EMU EN T RY F O R UK SME FI N A N C I N G

5.15 One reason may be that it is easier for venture capitalists in the US to exit theirinvestments through public share offerings. There is a divide in the EU between countrieswith equity-orientated financial systems, such as the UK and Netherlands, which supportrelatively large venture capital industries, and other non-equity-orientated markets where itis less developed. The rise in private pension funds, growth of institutional investors, anincrease in equity market integration, and the development of technology-based equityexchanges have all promoted the development of the EU equity market in recent years. Thiswill also promote the development of the EU venture capital market. Other factors may holdback development, for example, the regulatory environment for venture capital in the US isseen to be more favourable than the EU.

5.16 A crucial question is whether EMU has promoted an integrated market for venturecapital in the euro area, or whether it is still largely fragmented on national lines. If the latterwere the case then there would be little to gain from euro area entry. There are obviousreasons why venture capital may remain fragmented. It relies on the availability ofinformation on the risks of a project, and involves close co-operation between the venturecapitalist and the SME. This suggests that venture capitalists may have a strong home bias,preferring to deal with SMEs that operate in local markets with which they are familiar. Someevidence for this comes from the strong South East bias to venture capital in the UK wheremost private equity firms are based. In addition, the differing regulatory environments forprivate equity in Europe may also act to sustain a fragmented market.

5.17 EMU may also affect expected returns. If EMU promotes trade integration, investorsmay view UK start-ups and small businesses as entrants to the euro area market rather thanthe domestic market. Expected returns may rise as a consequence. Higher expected returnsmay be one explanation for the strength of the US venture capital market; US start-ups entera large integrated market and if they prosper then growth can be very rapid.

5.18 The UK venture capital sector currently accounts for 28 per cent of the EU total. EMUentry could potentially more than double the pool of funds which UK SMEs could accesswithout currency risk. Plus growth in EU equity markets may promote the furtherdevelopment of the EU industry.

5.19 However, the flow of funds could also go the other way, with UK venture capitalistsswitching funding to a more accessible euro area market. EMU entry could encourage UKfirms to build on their existing advantage in the European market. In 2001, the UK invested€2.8 billion of venture capital in other European countries, while Germany invested€0.9 billion and the Netherlands €0.3 billion. EMU could boost the quantity of UK venturecapital that goes to non-UK EU SMEs, supporting an outflow of financing from the UK.

Conclus ion on the impact o f EMU on UK SME f inanc ing

5.20 SMEs tend to raise finance in different ways to large firms, often relying on particularbanks. Venture capital is a small proportion of SME financing, but is important to high growthfirms. EMU entry would reduce the costs of raising cross-border funds within the euro areafor SMEs. But information and monitoring costs are likely to remain an important constrainton SMEs accessing the euro area wholesale capital market. Smaller SMEs in particular arelikely to remain reliant on local retail finance. While EMU entry could increase competition inthe market for bank lending to SMEs and increase the amount of venture capital available toUK firms, both are likely to be limited by home bias.

46

5

The implicationsof EMU for

venture capitalin the UK

6 TH E ST R U C T U R E O F UK CO R P O R AT E

F I N A N C I N G A N D EMU

47

6.1 The previous sections focused on the implications for UK firms of access to a wider euroarea financial market. But the structure of financial markets and of corporate ownership canalso affect the cost of capital. UK financial structures are different to those of much of the euroarea. UK large firms rely more on equity to raise capital, while in the euro area bank lending ismore important. UK firms also have less concentrated ownership than euro area firms.

6.2 Modigliani and Miller (1958) look at how a firm chooses its capital structure, and showthat in a world with perfect capital markets, without corporation tax or bankruptcy costs, afirm’s capital structure (i.e. its debt-to-equity ratio) has no effect on the value of the firm.However, tax and bankruptcy costs do affect the optimal capital structure of firms. Forexample, as interest payments on debt are tax deductible, introducing corporation tax tofirms’ decisions creates an incentive for 100 per cent debt financing. The higher the debtgearing, the less tax a firm has to pay, and the higher in theory the value of the company. Thisis in turn mitigated by the increased risk of bankruptcy associated with rising debt-to-equityratios, so that the optimal proportion of debt finance for any individual firm is somewherebetween 0 per cent and 100 per cent.

6.3 In addition, the structure of financial markets and corporate financing may affect thecost of capital through informational asymmetries. All financing arrangements will besubject to some degree of asymmetric information, for example, it is likely that the borrowerwill know more about the risk of a project than the lender. This may lead to adverse selection,where the lender requires a premium to compensate for the chance of risk being higher thananticipated. The lender’s lack of control over a project also introduces problems of moralhazard. Unless the lender can formulate a contract which completely aligns their interestswith that of the borrower, the borrower has an incentive to opt for excessively risky projects;anticipating this, lenders will demand a premium to compensate for the extra risk.

6.4 Because of these factors, the type of finance a company uses can affect the cost ofcapital. For example, the provision of bank lending often involves a close relationshipbetween the bank and borrower which can potentially reduce asymmetric information.External finance tends to be more costly than internal finance, unless loans are fullycollaterised, due to asymmetry of information. This means the availability of internal financemay be an important determinant of investment. When cash flow is low, perhaps due to aneconomic downturn, investment may fall. In the absence of complete information, afinancer will use a firm’s net worth to provide an indication of the risk of a project. Ifcorporate profitability and net worth decline, the cost of capital may rise as lenders demanda greater risk premium, leading to a fall in aggregate investment. These issues are notcaptured in standard models of investment, which assume the cost of capital is determinedby forward-looking estimates of project risk, not by current net worth.1

UK firms are typically characterised as having a different capital structure from those inthe euro area: ownership is equity orientated and highly diversified.

Some analysts suggest this structure leads to capital market imperfections which raise thecost of capital, though evidence on this is far from clear. It’s also unclear whether EMU willhave a significant impact on these structures. Many indicators suggest the euro area ismoving more toward a UK-style equity-orientated structure.

1 See for example, Ashworth and Davis (2001). These issues are also discussed in the EMU study by HM Treasury EMUand the monetary transmission mechanism.

TH E ST R U C T U R E O F UK CO R P O R AT E F I N A N C I N G A N D EMU

6.5 Evidence suggests these effects are particularly strong in the UK, suggesting problemssuch as adverse selection and moral hazard may be particularly acute here. Ashworth et al.(2001), Schiantarelli (1996) and Bond (1999) survey the literature and find evidence for cashflow being an important driver of investment in the UK, and find that the effect is moresignificant in the UK than in other European countries. One reason why these inefficienciesmight be more prevalent in the UK is the structure of the financial market and corporateownership.

Structure o f corporate f inance in the UK

6.6 Why should the UK structure of corporate financing lead to greater inefficiencies thanin the rest of Europe? It is typically assumed that the UK has a quite different structure ofcorporate financing than some other countries in Europe. Chart 6.1, based on dataconstructed by Byrne and Davis (2002),2 indicates that the UK has high levels of equity, whilein Germany and Italy bank lending is more important.

6.7 Aggregate data on equity and bank lending does not present the whole picture. Factorssuch as the concentration and institutional structure of corporate financing are alsoimportant. Bond (1999) and Mayer (1999a) identify two main factors that differentiate the UKsystem from that of Germany and other continental European countries:

• institutional shareholders, such as pension funds, own a higher proportion ofthe shares of UK companies than elsewhere. Because these institutions spreadrisk by diversifying their assets, UK firms generally have a much lessconcentrated ownership structure. In other European countries institutionalinvestors are less important, and so firms have more concentrated ownershipstructures. For example, around 16 per cent of the 20 largest firms in the UK in1999 had a single shareholder owning 25 per cent or more of the equity. InFrance and Germany, 80 percent of the largest firms had a large singleshareholder (Mayer, 1999a); and

48

6

BondsLoans Equity

ItalyFranceGermanyUK0

10

20

30

40

50

60

70

80

90

100

Source: Byrne and Davis, 2002.

Chart 6.1: Sources of finance in Germany, Italy, France and UK, 2000

Money market instruments

2 HM Treasury is grateful to the authors for kindly making their dataset available for use.

49

• banks own more corporate equity in Europe. Banks owned 6 per cent ofcorporate equity in France, and 13 per cent in Germany,3 over the period 1980-1990; the equivalent figure in the UK was 2 per cent (Mayer, 1999a).

6.8 Is there a relationship between the UK’s equity-orientated, diversified ownershipstructures and UK investment’s sensitivity to cash flow? A diversified structure of ownershipmay increase problems of asymmetric information between lenders and firms, leading toadverse selection. Similarly, less concentrated ownership may be associated with weakinvestor control that could increase moral hazard.

6.9 There is a related debate on the more general respective merits of the UK system versusthe German system.4 One strand of literature claims the German model, based onconcentrated ownership and bank finance, encourages investment and better facilitatesmonitoring and control of managers. For example, McCauley and Zimmer (1989) argue thatclose relations between banks and firms in Germany allow firms to borrow more withoutincreasing risk, as bank ownership reduces the risk that a firm will go bankrupt.

6.10 More recent work has emphasised the advantages of the UK model’s robust market forcorporate control. It is argued that the risk of takeover creates incentives for goodmanagement; and that the system may better facilitate the channelling of funds to efficientindustries. In the German system, funds may become tied up in inefficient firms at the expenseof more dynamic industries. There may be different moral hazard problems associated withthe German model, for example, it can be argued that the significant voting power of Germanbanks in many large firms allows them to distort decision making in order to maximise debt(Morck and Nakamura, 1999), or may lead to the rejection of good investment projects(Shleifer and Vishny, 1996). A heavy reliance on bank financing may make economy-widecredit crunches more likely if banking sector profitability falls and may make the bankingsystem more fragile.

6.11 Mayer (1999a) argues there is little empirical evidence to support the view that any onesystem is inherently superior to the other. He suggests that rather than any one systemleading to more superior overall outcomes, the type of system may instead influence thestructure of economic activity. For example, diverse ownership patterns may allow for theaggregation of a wide range of risk perceptions and so may be better suited for investment inrisky new technologies. Less investor control may also allow managers of such projects moreflexibility to change and implement a variety of different approaches. Concentratedownership involving more monitoring and control by investors may be better suited to large-scale traditional and long-term investments, such as in heavy manufacturing.

Impact o f EMU on the structure o f corporate f inance

6.12 Continued financial market integration in the EU does not necessarily lead toconvergence of financial systems (Mayer, 1999a). Alternatively, it could increase competitionfor incorporation, as companies will increasingly choose to be based wherever the financialmarket structure and regulatory environment is best suited to their needs. Differentiationmay increase, as different systems specialise in different types of activity and attract firmsfrom other countries. However, firms will also consider the cost of different sources offinance, and as discussed in Section 4, EMU entry could result in lower costs for bond andequity financing. Overall, Mayer expects a mix of some convergence where pressure for this isoverwhelming; combined with continued differentiation in other areas.

TH E ST R U C T U R E O F UK CO R P O R AT E F I N A N C I N G A N D EMU6

The merits of theUK and German

systems

3 Banks in Germany control a higher percentage of the voting rights of firms (McCauley and Zimmer, 1989).4 For example see Mayer (1999b).

TH E ST R U C T U R E O F UK CO R P O R AT E F I N A N C I N G A N D EMU

6.13 Davis (1999) argues that EMU may promote the development of UK-style capitalmarkets in the euro area. He points to the rise in institutional investment in the euro area,which is being driven by an increasing need for private pension provision. This trend hasbeen occurring for several years: compared to 1980, assets held by institutional investors inGermany had almost doubled by 1997 and had more than trebled in France. Institutionalinvestors tend to hold a greater proportion of their assets in equities and other securitisedassets.

6.14 This should increase the demand for equity capital and so lower its cost for firms. Butfirms will have to accept increased institutional ownership and the importance ofrelationship banking will be diminished. Davis suggests this may have an adverse impact onfinancing for small firms, for whom there is some evidence that institutional investmentfunds are less willing to provide finance.

Conclus ion on the impact o f EMU entry on UK corporatestructure

6.15 If EMU and other financial developments promote the development of a more equity-orientated finance structure in the euro area, then EMU entry is less likely to alter thestructure of UK finance, as the UK already has an equity culture.

6.16 It is also possible that other ownership structures will continue to prosper in EMU, asMayer (1999a) points out. Insofar as EMU promotes country-level specialisation in financialservices and reduces the barriers to incorporating in other euro area countries, UK firms maybenefit from being better able to utilise different financing structures for different types ofinvestment. However, joining EMU is unlikely to have an immediate beneficial impact on UKfirms which rely on bank financing, due to the need for geographical proximity.

50

6

Credit risk-freerate

7 CO N C LU S I O N S : TH E IM PAC T O F EMUO N T H E CO S T O F CA P I TA L

51

7.1 This study sets out to examine the potential impact of EMU entry on the UK cost ofcapital. For the purposes of exposition, the study divides up the cost of capital into the creditrisk-free rate and the market risk premium, and looks for evidence that these elements couldfall in the euro area. The particular characteristics of small and medium-sized enterprises(SMEs) are examined in order to see if euro membership could reduce the cost of capital forthis group. The study also looks at what the implications for the cost of capital might be ofchanges in the structure of UK corporate finance prompted by EMU entry.

7.2 Although some euro area countries, such as Spain and Italy, saw a significant decline incredit risk-free rates as a result of joining EMU, the study finds little scope for UK credit risk-free rates to fall significantly, as markets already expect continued low and stable inflation inthe UK. Entry would lead to some shift in the yields on UK government bonds towards thoseprevailing in the euro area, although the overall impact could be muted by other cyclicalimpacts at the short end and by credit risk factors at the long end.

7.3 The study looks at what the impact of joining the larger euro area financial marketscould be on the market premium for UK firms. In particular it examines whether they couldbenefit from lower liquidity and credit risk premia than in the UK market. While there is notyet clear evidence of a fall in the market premium in the euro area, the conditions for such areduction in costs are developing. Euro-denominated markets are growing in size,investment funds are diversifying across borders, there have been changes in the financialinfrastructure and some transactions costs have been falling. For the full gains from thesechanges to be realised there needs to be significant progress on lowering the remaining legal,regulatory and cultural barriers necessary to complete the EU market in financial services.UK firms can access the euro-denominated market from outside EMU at relatively low costat present, but entry would increase access at the margin, and have potentially greaterbenefits for specific groups such as SMEs.

7.4 For this reason, the study examines the scope for SMEs to gain from access to eurofinancial markets. It finds that, as they tend to raise funds locally and rely more heavily onbank lending, there would be little immediate prospect of a significant fall in the cost ofcapital for most SMEs.

7.5 UK firms are typically characterised as having a different, more equity-orientated,capital structure than those in the euro area, which are characterised by more of a relianceon bank finance. It is unclear whether EMU will have a significant impact on these structures.Many indicators suggest the euro area is actually moving more toward an equity-orientatedstructure. If EMU and other financial developments promote the development of a moreequity-orientated finance structure in the euro area, then EMU entry is unlikely to alter thestructure of UK finance.

Market riskpremium

SME financing

Structure ofcorporatefinancing

RE F E R E N C E S

53

Adjaoute, K., Bottazzi, L., Danthine, J., Fischer, A., Hamaui, R., Portes, R. and Wickens, M.(2000) ‘EMU and Portfolio Adjustment’, CEPR Policy Paper No. 5.

Ashworth, P., Hubert, F., Pain, N., and Riley, R. (2001) ‘UK Fixed Capital Formation:Determinants and Constraints’, NIESR paper prepared for the Department of Trade andIndustry and CBI/TUC Working Party on Investment.

Ashworth, P. and Davis, E.P. (2001) ‘Some Evidence on Financial Factors in theDetermination of Aggregate Business Investment for the G7 Countries’, paper presented atMoney, Macro and Finance Research Group Annual Conference, Queen’s University, Belfast.

Bank of England (1999) Practical Issues Arising from the Euro, December 1999.

Bank of England (2000) Practical Issues Arising from the Euro, November 2000.

Bank of England (2001) Finance for Small Firms - 8th Report, March 2001.

Bank of England (2002) Practical Issues Arising from the Euro, November 2002.

Bank for International Settlements (2000) 70th Annual Report.<http://www.bis.org/publ/ar2000e.htm>

Bank for International Settlements (2001) Quarterly Review: International Banking andFinancial Market Developments, March 2001. <http://www.bis.org/publ/r_qt0103.htm>

Bank for International Settlements (2002) Quarterly Review: International Banking andFinancial Market Developments, December 2002. <http://www.bis.org/publ/r_qt0212.htm>

Bond, S. (1999) ‘UK Investment and the Capital Market’, paper presented at HM Treasuryseminar held at 11 Downing Street on 12th October 2000.

Brealey, R. A., Cooper, I. and Kaplanis, E. (1999) ‘What is the International Dimension ofInternational Finance?’ European Finance Review 3(1), pp. 103-119.

Brealey, R. A. and Myers, S. C. (2000) Principles of Corporate Finance, London: McGraw-Hill.

Breedon, F., Henry, B., and Williams, G. (1999) ‘Long-Term Real Interest Rates: Evidence onthe Global Capital Market’, Oxford Review of Economic Policy 15(2), pp. 128-142.

British Venture Capital Association (2002) Report on Investment Activity, 2001.

Brooke, M., Clare A., and Lekkos. I. (2000) ‘A Comparison of Long Bond Yields in the UnitedKingdom, the United States, and Germany’, Bank of England Quarterly Bulletin, May 2000.

Byrne, J.P. and Davis, E.P. (2002), ‘A Comparison of Balance Sheet Structures in Major EUCountries’, National Institute Economic Review 180(April), pp. 83-95.

Competition Commission (2002) ‘The Supply of Banking Services by Clearing Banks toSmall and Medium-Sized Enterprises: A Report on the Supply of Banking Services byClearing Banks to Small and Medium-Sized Enterprises within the UK’, paper presented toParliament by the Secretary of State for Trade and Industry and the Chancellor of theExchequer by Command of Her Majesty, March 2002.

Cooper, N. and Scholtes, C. (2001) ‘Government Bond Valuations in an Era of DwindlingSupply’, BIS Discussion Paper No. 5. <http://www.bis.org/publ/bispap05e.pdf>

RE F E R E N C E S

54

Cooper, N., Hillman, R. and Lynch, D. (2001) ‘Interpreting Movement in High-YieldCorporate Bond Market Spreads’, Bank of England Quarterly Bulletin, Spring 2001.

Corvoisier S. and Groop, R. (2001) ‘Bank Concentration and Retail Interest Rates’, ECBWorking Paper No. 72.

Credit Suisse First Boston (2000) The Credit Suisse First Boston Equity-Gilt Study.

Credit Suisse First Boston (2001) The Credit Suisse First Boston Equity-Gilt Study.

Davis, P. (1999) ‘Institutionalisation and EMU: Implications for European Financial Markets’,International Finance 2(1), pp. 33-61.

Detken, C. and Hartmann, P. (2002) ‘Features of the Euro’s Role in International FinancialMarkets’, Economic Policy 35 (October), pp. 553-570.

Dimson, E., Marsh, P. and Staunton, M. (2002) ‘Global Evidence on the Equity RiskPremium’, Journal of Applied Corporate Finance, forthcoming.

European Central Bank (2000) EU Banks’ Margin and Credit Standards.

European Central Bank (2001) The Euro Equity Markets, August 2001.

European Commission, DG Economic and Financial Affairs (2001) ‘The EU Economy 2001Review’, European Economy 73.

European Commission, DG Economic and Financial Affairs (2002) ‘Cross-border Clearingand Settlement Arrangements in the EU’, Economic Paper No. 163.

European Venture Capital Association (2002) ENN Supplement No. 6, August 2002.<http://www.evca.com/admin/attachments/tmpl_1_art_9_att_170.pdf>

Fratzscher, M. (2001) ‘Financial Market Integration in Europe: On the Effects of EMU on theStock Markets’, ECB Working Paper No. 48.

Hardouvelis, G., Malliaropoulos, D. and Priestley, R. (1999) ‘EMU and European StockMarket Integration’, CEPR Discussion Paper No. 2124.

Ibbotson R. and Chen P. (2001) The Supply of Stock Market Returns, Yale School ofManagement Research Paper.

Jorgensen, D. W. (1963) ‘Capital Theory and Investment Behavior’, American EconomicReview 53, pp. 247-257.

Kempa, B and Nelles, M., (2001) ‘International Correlations and Excess Returns in EuropeanStock Markets: Does EMU Matter?’, Applied Financial Economics 11(1), p. 69-73.

Kraus, T. (2001) ‘The Impact of EMU on the Structure of European Equity Returns: AnEmpirical Analysis of the First 21 Months’, IMF Working Paper No. 01/84.

London Economics (2002) Quantification of the Macro-Economic Impact of Integration ofEU Financial Markets: Final Report to The European Commission, DG Internal Market.<http://europa.eu.int/comm/internal_market/en/finances/mobil/overview.htm>

London Stock Exchange (2002) Clearing and Settlement in Europe – Response to the FirstReport of the Giovannini Group.

Lund, M. and Wright, J. (2001) ‘The Financing of Technology-Based Small Firms: A Review ofthe Literature’, Bank of England Quarterly Bulletin, Spring 2001.

55

RE F E R E N C E S

Mann, C. and Meade, E. (2002) ‘Home Bias, Transaction Costs, and Prospects for the Euro: AMore Detailed Analysis’, International Financial Stability Programme, Centre for EconomicPerformance, Working Paper No. 02-3.

Mayer, C. (1999a) ‘European Capital Markets: Competition between Systems’, EuropeanInvestment Bank Economic and Financial Reports.

Mayer, C. (1999b) ‘Investment and Growth: The Role of Corporate Governance’, paperpresented at HM Treasury seminar held at 11 Downing Street on 12th October 2000.

McCauley, R. and Zimmer, S. (1989) ‘Explaining International Differences in the Cost ofCapital’, Federal Reserve of New York Quarterly Review.

Merrill Lynch ‘Fund Manager Survey’ September 2002.<http://online.tnsofres.com/Multimedia/Merrill_lynch/september2002/september2002reports_wj.htm>

Modigliani, F., and Miller, M. (1958) ‘The Cost of Capital, Corporation Finance, and theTheory of Investment’, American Economic Review 48 (June), pp. 261-297.

Morck, R. and Nakomura, M. (1999) ‘Banks and Corporate Control in Japan’ Journal ofFinance 54, pp. 319-339.

Murray, A. (2001) The Future of European Stock Markets, London: Centre for EuropeanReform.

OECD (2002) ‘Increases in Investment in the 1990s: The Role of Output, Cost of Capital andFinance’ Financial Market Trends 83, November 2002, pp. 121-142.

Santos, J. and Tsatsaronis, K. (2002) ‘The Cost of Barriers to Entry: Evidence from the Marketfor Corporate Euro Bond Underwriting’, paper for conference on The Euro: Valuation,Hedging and Capital Market Issues, April 5 2002, New York University Salomon Center.

Schiantarelli, F. (1996) ‘Financial Constraints and Investment: Methological Issues andInternational Evidence’, Oxford Review of Economic Policy 12(2), pp. 70-89.

Shleifer, A. and Vishney, R.W. (1996) ‘A Survey of Corporate Governance’, NBER WorkingPaper No. 5554.

Small Business Service (2002) ‘Small and Medium Sized Enterprise Statistics for the UK’,press release.

Stulz, R. (1999) ‘Globalization of Equity Markets and the Cost of Capital’, NBER WorkingPaper No. 7201.

Tobin, J. (1969) ‘A General Equilibrium Approach to Monetary Theory’, Journal of Money,Credit and Banking 1(1), pp. 15-29.

A AN N E X A : EC O N O M E T R I C ST U D I E S O F

EQ U I T Y MA R K E T IN T E G R AT I O N

57

A1 Several recent studies have used statistical techniques to examine whether EMU equitymarkets have converged due to economic and financial market integration. The results ofthese studies are summarised in Tables A1 and A2 and are discussed in more detail below.These studies use two basic approaches:

• analysing the factors which drive equity returns; and

• analysing equity index correlations.

A2 If the euro area is becoming more economically integrated, then individual equityreturns will be driven less by country-specific factors and more by euro area-wide and/orsector-wide factors. The first approach analyses the factors which drive equity returns bydecomposing the variance of equity returns and attributing it to different sources of risk:world risk, country-specific risk, sector-specific risk, EU-wide risk and idiosyncratic risk.Table A1 shows that since convergence to EMU began, the contribution to returns fromsector and EU-wide factors has increased in importance, and country and world factors arestable or have decreased.

Table A1: Summary results of equity market studies, changes tocontributions to returns since the start of convergence to EMU

Contribution to returns from:

World EU Sector Country Idiosyncratic

Tsatsaronis in BIS, 2001 Increased StableFratzscher, 2001 Stable IncreasedKraus, 2001 Decreased Increased Decreased IncreasedHardouvelis et al.,1999 Increased Decreased

A3 The second approach examines the correlation structure of equity indices; for example,Table A2 indicates that Hardouvelis et al. (1999) find that the correlation of national indicesto the euro area index has increased. A number of the other results from this approach haveless clear-cut implications. For example, Kraus (2001) finds that sector-to-sector correlationshave decreased since EMU. Kraus argues that this suggests increased economic integration,as sector returns are now being driven more by diverse sector-specific factors rather thanshared country factors.

Table A2: Summary results of equity market studies, changes incorrelations of returns since the start of convergence to EMU

Correlations of returns:

Cross Cross Country to Cross countrysector country euro area by sector

Adjaoute et al., 2000 Increased IncreasedKraus, 2001 Decreased DecreasedHardouvelis et al., 1999 IncreasedKempa and Nelles1, 2001 Increased Increased1Kempa and Nelles (2001) simulate pre- and post-EMU periods by modelling returns with and without exchange rate volatility.

AN N E X A : EC O N O M E T R I C ST U D I E S O F EQ U I T Y MA R K E T IN T E G R AT I O N

58

AA4 Kraus (2001) also finds that correlation between country indices has decreased - incontrast to Adjaoute et al. (2000) who find that this has increased. First intuition mightsuggest that country correlations would rise in EMU as a single monetary policy increaseseconomic convergence. Kraus (2001) argues that decreased correlation may be due toincreased industrial specialisation in EMU, which increases country divergence. However, itseems unlikely that such change would have occurred to any significant degree in therelatively short period of time since EMU began. The difference in the results may be due todifferent time periods. Kraus compares the periods 1997-1999 and 1999-2000; Adjaoute et al.the periods 1990-1994 and 1997-1999.

A5 Adjaoute et al. (2000) also find that country correlations based on sector indices haveincreased in the post-convergence period, though these correlations are less strong than thepure country correlations. The authors argue that as EMU country indices are now morecorrelated, portfolio allocations based on European sector strategies are superior to thosebased on EU country strategies, or on home-based sector strategies. The authors argue thatfinancial market integration is driven by economic convergence, which creates incentives forpan-European investment allocations, rather than the removal of currency risk or otherfactors.

A6 Kempa and Nelles (2001) find that EMU-wide investment strategies dominate home-based investment strategies. They show that both country-to-country and country-to-euroarea correlations are higher in the absence of exchange rate volatility, indicating thatexchange rates are a source of country-specific risk in the equity market. The paper alsoestimates national betas (see Section 2), and finds that these fall when exchange rate volatilityis removed, suggesting the equity cost of capital will be lower in EMU.

A7 The results of all these studies suggest that the equity market has become moreintegrated since EMU. The question which remains is what impact this will have on the costof capital. Kampa and Nelles (2001) show that national betas, and therefore the national costof capital, are lower in the absence of exchange rate volatility. However, as has been discussed,the impact of integration on the cost of capital can work in two directions: increasedopportunity for diversification should lower the cost of capital, while increased economicconvergence can work in the other direction, by reducing the benefits of diversification.

A8 Hardouvelis et al. (1999) make a formal attempt to model implications for the cost ofcapital. They find that integration has lowered the cost of capital, in other words that thediversification effect dominates the convergence effect. They first use a capital asset pricingmodel (CAPM) methodology to estimate equity market returns due to EU-wide, country-specific and currency risk factors. They find the proportion of equity returns due to EU-widerisk rose from 25 per cent in 1991-1995 to 65 per cent in 1996-1998. They also estimateintegration weights for individual EU countries. Except for the UK these all approach unity atthe introduction of the euro in 1999, indicating close correlation between country equityreturns and euro area returns.

A9 The authors argue that theory gives no firm indication as to the impact of integration onthe cost of capital, but their intuition suggests that it should fall, as EU-wide risk could beexpected to be lower than the individual country risk it replaces. Hardouvelis et al. (1999) usea model which decomposes the cost of capital into market, currency and local riskcomponents; as integration increases, the local risk and currency risk components decreaseand the market-wide risk component increases. The impact on the cost of capital depends onthe interaction of these effects.

59

A10 The key results are shown in Table A3. The estimates suggest that the cost of capital fellby close to 2 per cent for a group of eight1 EU countries (which includes the UK) over theperiod examined. This is the combined result of a reduction in local risk of around 3 per cent,a rise in EU-wide risk of 1 per cent and a small increase in currency risk. The counter-intuitiveincrease in currency risk is attributed to the influence of UK currency variability and to theEMU currencies not having locked together by the end of the estimation period in 1998. Theauthors expect the currency risk component to fall considerably after EMU is completed. Fora group of 122 EU countries, the reduction in the cost of capital was around 1.3 per cent.

Table A3: Impact of integration on cost of capital, per cent

Period Total effect Decomposition of total effect

EU wide Currency Local

EU7 1992–98 –1.90 1.00 0.24 –3.14EU12 1996–98 –1.26 0.62 0.15 –2.03Source: Hardouvelis et al., 1999.

A11 Fratzscher (2001) tests for euro area financial integration using an uncovered assetreturn parity model, which relates equity returns to past local, euro area and global marketinformation, and to local, euro area and global shocks. The US acts as a proxy for the globalmarket. Fratzscher (2001) assumes that a more integrated market is one where returns arestrongly correlated to euro area shocks. Table A4 illustrates some of the key results.Increasingly large spillovers from euro area shocks to individual euro area country equityreturns are found, which the author attributes to increased integration. Averaged over thewhole period, a euro area shock of 1 per cent leads to a 0.445 per cent change in returns; inthe 1998-2000 period the spillover has increased to 0.911 per cent. The spillover from the USremains fairly constant over the period.

Table A4: Results for equity market integration, per cent

Impact on euro area returns of 1 per cent shock…

. . . from euro area . . . from USA

1986–2000 0.445 0.3671986–1992 0.270 0.3211998–2000 0.911 0.345Source: Fratzscher, 2001.

A12 Similar results are found when the methodology is adjusted to allow time-varyingcoefficients. The results indicate relatively low integration in the early 1990s, a very rapidincrease in the period 1996 to 1999, followed by a levelling off over 1999 to 2000. During theperiod of rapid expansion the euro area overtakes the US as the major factor influencingreturns. (Interestingly, the same pattern is evident for the UK).

A13 The paper attempts to uncover the causes of this integration. The hypothesis used is thatexchange rate risk was a significant barrier to financial integration, preventing investors fromdiversifying across the euro area. This hypothesis is tested by regressing an integrationparameter on the exchange rate volatility of each euro area market against the Deutschmarkand the US dollar. The resulting coefficients are all negative, indicating that a reduction involatility leads to increased integration. This result stands when controlled for the correlationbetween exchange rate stability and real and monetary convergence through EMU, whichmay also be a reason for financial market integration.

AN N E X A : EC O N O M E T R I C ST U D I E S O F EQ U I T Y MA R K E T IN T E G R AT I O NA

Fratzscher(2001)

1 Belgium, Luxembourg, France, Italy, Germany, Netherlands, Spain, UK.2 The above plus Austria, Finland, Ireland and Portugal.

AN N E X A : EC O N O M E T R I C ST U D I E S O F EQ U I T Y MA R K E T IN T E G R AT I O N

A14 Adjaoute et al. (2000) analyse developments in the variance-covariance matrices ofcountry and sector equity returns across the euro area, with the objective of uncoveringchanges in optimum portfolio allocation strategies for euro area investors. Unlike Fratzscher(2001) their prior assumption is that currency risk has not been a significant explanation forequity portfolio home bias in the past, so that EMU could only be expected to have a minordirect role in promoting portfolio diversification.

A15 The analysis first compares variance-covariance matrices in the period 1990-1994 tothose from 1995-1998. Then a more narrowly defined convergence period is examined:November 1995 to July 1997 compared to August 1997 to April 1999. In both cases, the paperfinds that cross-country equity return correlations increased strongly in the second period. Totest whether this was simply due to increased capital flows as a consequence of globalisationrather than due to EMU, a similar analysis is undertaken for the rest of the world. Althoughcorrelation also increased between other regions of the world over the same time period, thedegree of correlation was much less strong than in the euro area.

A16 The analysis is repeated using sector indices. Again country correlations based on sectorindices have increased in the post-convergence period. It is noted that these correlations areless strong than the pure country correlations. The authors argue that, as EMU countryindices are now more correlated, investment strategies based on European sector strategiesare superior to those based on EU country, or on home-based sector strategies.

A17 This conclusion is cross-checked through optimal portfolio analysis which confirmsthat diversification by sector across the euro area is superior to diversification by countryacross the euro area and to diversification by sector within a country. The authors believe thissuggests that rather than becoming more specialised, countries in EMU are becoming morediversified.

A18 Kraus (2001) examines the impact of EMU on the sources of equity returns usingvariance decomposition analysis. Five risk factors are used: world risk, European risk, countryrisk, European sector risk and idiosyncratic risk. Initially, stock returns are regressed solely onthe world risk factor. A succession of new models is then regressed, each with one additionalfactor added; the marginal R2 of each new model is attributed to the last factor added. Thedata period is from April 1997 to October 2000.

A19 The results are shown in Table A5. These indicate that the contribution of world risk tostock returns has fallen significantly, suggesting the European market has become moreautonomous since EMU. Country-specific factors have also fallen. Industry-specific factorshave increased and now account for 12 per cent of total return variance. There has also beena significant rise in the contribution of idiosyncratic risk, which means that the importanceof diversifiable risk has increased sharply since EMU.

A20 Kraus (2001) views the rise in the contribution of industry-specific factors as animportant change, which he believes has been driven by two key factors. First, the singlemonetary policy has removed a significant source of country-specific shocks, so increasingthe profile of industry-specific factors. Second, there has been a shift in industry portfolioallocation strategies toward sector strategies, which has had a self-fulfilling role in promotingsector profiles.

60

AAdjaoute et al.

(2000)

Kraus (2001)

61

Table A5: Sources of equity returns, per cent

Before EMU After EMU Change

World 20.5 8.6 –11.91

Europe 6.4 6.7 +0.3Country 9.3 7.1 –2.21

Sector 6.0 12.0 +6.01

Unexplained 57.9 65.7 +7.81

1 Signifies change significant at 1 per cent level.Source: Kraus, 2001.

A21 Kraus (2001) then uses a variety of techniques to examine cross-country and industryequity return correlations before and after EMU. This shows that correlations between sectorshave fallen significantly since the start of EMU, which is attributed to the reduction incountry-specific risk factors making sector risk more pronounced and more diverse. Theresults also indicate that cross-country correlations have decreased. This runs counter toKraus’ initial intuition – given that EMU countries now share a common monetary policy, heexpected to find an increase in cross-country correlations. An explanation may be thatcountry divergence is due to increased industrial specialisation, or that there has been realinterest rate divergence.

A22 Another feature of this analysis is that clusters of highly-correlated countries can beidentified. In particular the large markets, Germany, France, Netherlands, Spain and Italy,have become very closely correlated. Spain and Italy only joined this cluster after EMU,suggesting that, previously, investors did not perceive them as part of the group of large EUmarkets, perhaps due to their individual monetary policies. Further analytical techniquesalso indicate a country size factor in the data, but Kraus (2001) can find no economicrationale for why this should be present.

A23 This study analyses national equity indices’ correlations over the period 1994 to 1997,and then simulates the impact of EMU by removing exchange rate volatility from the data.The analysis focuses solely on the impact of the exchange rate; there is no attempt to modelthe impact of economic convergence.

A24 Country-to-country equity index correlations are reported with and without exchangerate volatility. On average, correlations are higher in the presence of exchange rate volatility,indicating that exchange rate movements are a source of country-specific shocks. As wouldbe expected from this result, correlations between national indices and a euro area index arelower in the presence of exchange rate volatility.

A25 Kempa and Nelles (2001) estimate national betas with and without exchange ratevolatility. Betas tend to be lower once exchange rate effects are removed, suggesting that EMUshould lower the cost of capital. The next step is to look at the excess returns to investors frominvesting in an efficiently diversified euro area portfolio, rather than the national market.These are positive with or without exchange rate volatility, suggesting that regardless of theexchange rate regime, EU portfolios offer higher returns than national portfolios. In mostcases, the gains are higher in the presence of exchange rate volatility, suggesting that theincentives for diversification are lower in EMU. However, this analysis takes no account ofsector returns which may affect this result; for example, the Adjaoute et al. (2000) study foundthat sector portfolio strategies would dominate country strategies in EMU.

AN N E X A : EC O N O M E T R I C ST U D I E S O F EQ U I T Y MA R K E T IN T E G R AT I O NA

Kempa andNelles (2001)

B AN N E X B: CO R P O R AT E BO N D

SP R E A D S

63

Basis points

12.25 per cent 2006 sterling

7.125 per cent 2006 sterling6.125 per cent 2006 euro

Jul-02

May-02

Mar-02

Jan-02

Sep-0

2

Nov-01

Sep-0

1

Chart B1: Spread of British Telecom bonds over swap rates

0

20

40

60

80

100

120

140

160

180

200

Source: Bloomberg.

Basis points


Aug-0

2

Jun-02

Apr-0

2

Feb-0

1

Dec-01

Oct-01

Aug-0

1

Jun-01

Apr-0

1

Feb-0

1

Dec-00

Oct-00

Aug-0

0

Jun-00

Apr-0

0

Feb-0

0

Dec-99

Oct-99

Chart B2: Spread of Gallaher bonds over swap rates

0

20

40

60

80

100

120

140

160

180

Source: Bloomberg.

AN N E X B: CO R P O R AT E BO N D SP R E A D S

64

B

Source: Bloomberg.

0

20

40

60

80

100

120

140

160

180

200Chart B3: Spread of Pearson bonds over swap rates

Oct-99

Dec-99

Feb-0

0

Apr-0

0

Jun-00

Aug-0

0

Oct-00

Dec-00

Feb-0

1

Apr-0

1

Jun-01

Aug-0

1

Oct-01

Dec-01

Feb-0

1

Apr-0

2

Jun-02

Aug-0

2

4.625 per cent 2004 euro 9.5 per cent 2004 sterling

Basis points

Basis points


Aug-0

2

Jun-02

Apr-0

2

Feb-0

1

Dec-01

Oct-01

Aug-0

1

Jun-01

Apr-0

1

Feb-0

1

Dec-00

Oct-00

Aug-0

0

Jun-00

Apr-0

0

Feb-0

0

Dec-99

Oct-99

Chart B4: Spread of HBOS bonds over swap rates

-20

-10

0

10

20

30

40

50

Source: Bloomberg.

65

AN N E X B: CO R P O R AT E BO N D SP R E A D SB

Basis points


Oct-02

Aug-0

2

Jun-02

Apr-0

2

Feb-0

2

Dec-01

Oct-01

Aug-0

1

Jun-01

Apr-0

1

Feb-0

1

Dec-00

Oct-00

Aug-0

0

Chart B5: Spread of Lloyds TSB bonds over swap rates

0

10

20

30

40

50

60

70

80

90

Source: Bloomberg.

emu and the cost of capital - home - bbc...

Documents