An International Comparison of Capital Structure and

An International Comparison of Capital Structure and

Debt Maturity Choices*

Joseph P.H. Fan

Faculty of Business Administration Chinese University of

Shatin, N.T. pjfan@cuhk.edu.hk

and

Sheridan Titman†

McCombs School of Business University of Texas at Austin

Austin, TX 78712

USA

titman@mail.utexas.edu

Garry Twite

School of Finance and Applied Statistics

Australian National University

Canberra, ACT 0200

Australia

garry.twite@anu.edu.au

October 2008

*

This paper has benefited from the useful comments and suggestions provided by Andres Almazan, Heitor Almeida, Lawrence Booth, Stijn Claessens, Joshua Coval, Sudipto Dasgupta, Jay Hartzell, Jiang Luo, Vojislav Maksimovic, Enrico Perotti, Tom Smith, participants at the 2003 European Finance Association Conference, the 2003 Financial Management Association Conference and the 2005 American Finance Association Meeting, seminar participants at the Australian National University, Australian Graduate School of Management, Chinese University of , University of Science and Technology, Shanghai University of Finance & Economics, University of Melbourne, University of Queensland, University of Sydney, and University of Texas at Austin, and an anonymous referee. Joseph Fan thanks the financial support by the Research Grants Council of the Special Administrative Region, China (Project No. CUHK6230/03H) and the University of Queensland for research support during his visit when part of the research was carried out. Garry Twite thanks the financial support by the Australian Research Council Discovery Project (Project ID. DP06505). †

Corresponding author. Finance Department, McCombs School of Business, The University of Texas at Austin, 1 University Station B6600, Austin, TX, 78712; email: titman.@mail.utexas.edu; phone:(512)232-2787; fax: (512)471-5073.

Abstract

This study examines the influence of institutions on the capital structure and debt maturity choices in a cross-section of firms in 39 developed and developing countries. Our evidence indicate that firms operating within legal systems that provide better protection for financial claimants tend to have capital structures with less total debt, and more long-term debt as a proportion of total debt. In addition, we find that firms that choose to cross-list tend to use more equity and longer-term debt. We also find that taxes and the characteristics of the financial institutions that supply capital have an influence on how firms are financed. Finally, we find that the cross-sectional determinants of leverage differ across countries. In particular, the relationship between profitability and leverage tends to be stronger in countries with weaker shareholder protection.

Keywords: Capital structure, Debt maturity JEL classification: G30, G32

1. Introduction

Corporate financing choices are determined by a combination of factors that are related to the characteristics of the firm as well as to their institutional environment. Although most studies focus on the importance of firm characteristics by examining corporate financing choices within individual countries,1 there is a growing literature that considers how institutional differences affect these choices. These more recent papers examine capital structure choices across countries (Booth, Aivazian, Demirguc-Kunt, and Maksimovic, 2001; Claessens, Djankov and Nenova, 2001; Demirguc-Kunt and Maksimovic, 1996, 1998, 1999; Giannetti, 2003).

This study builds on this recent literature in three important ways. First, because we consider these issues within a panel that includes industry dummies, together with firm-level variables, we identify the variation in capital structure across countries that cannot be explained by cross-country differences in the industrial mix and firm-level characteristics. Second, we consider a larger number of countries and a number of important institutional characteristics not previously explored in this literature. Finally, we include interactions between country-wide institutional variables and firm-level characteristics, which allow us to estimate how institutional differences across countries affect the cross-sectional variations in capital structures within the countries.

To understand our motivation, it is useful to illustrate the importance of country factors relative to industry factors in determining capital structure. A regression of firm leverage, measured as the book value of debt over the market value of the firm, on firm- 1

Examples of empirical studies examining the association between firm characteristics and capital structure within specific countries include Titman and Wessels (1988) – U.S., Campbell and Hamao (1995) – Japan and Gatward and Sharpe (1996) – Australia. Barclay and Smith (1995), Stohs and Mauer (1996) and Guedes and Opler (1996) examine the association between firm characteristics and debt maturity in the U.S. Gatward and Sharpe (1996) undertake a similar study of debt maturity in Australia.

specific variables, industry dummy variables and country dummy variables, has an adjusted R-square of 0.19. When the regression is run with all variables except for country dummies, the adjusted R-square is reduced to 0.15.2 However, in a regression that includes all variables except for industry dummies, the adjusted R-square is reduced only half as much, to 0.17. When the full regression is run with debt maturity, measured as the book value of long-term to total debt, as the dependent variable, the R-square is 0.25. When the regression is run with all variables except for country dummies, the R-square is substantially reduced to 0.09. However, in the regression that includes all variables except for industry dummies, the R-square is only slightly reduced to 0.23.

These regressions indicate that the country in which the firm resides is a more important determinant of how it is financed than its industry affiliation, which in turn suggests that cross-country differences in institutional factors are likely to have a first order effect on capital structure choices. To examine this possibility in more detail we estimate a regression on a large panel of firms in 39 different countries that examines the extent to which cross-country differences in capital structures can be explained by differences in taxation policies, legal and financial institutions as well as differences in economic development and inflation.

Our evidence suggests that institutional differences have a significant influence on capital structure choices. Specifically, firms tend to use less debt in countries where dividends are preferentially taxed. This evidence contrasts with Booth, Aivazian, Demirguc-Kunt, and Maksimovic (2001) who do not find a significant relation between debt ratios and tax policy. In addition, we find that the strength of a country’s legal 2

This result is similar in character to a regression reported by Booth, Aivazian, Demirguc-Kunt, and Maksimovic (2001).

2

system and public governance importantly affect firm capital structure. Weaker laws and more government corruption induce higher corporate debt ratios and shorter debt maturity3. We also find that cross-listing matters: firms that choose to cross-list tend to use more equity and longer-term debt and this effect is stronger for firms in weakly governed countries.

We also examine how the preferences of the suppliers of capital influence capital structure choices. The mix of capital suppliers differs across countries and our evidence indicates that that their preferences do influence capital structure choices.4 For example, we find that firms in countries with larger banking sectors have shorter maturity debt and that debt ratios tend to be lower in countries with a larger life insurance industry.

Our analysis also extends the analysis of Rajan and Zingales (1995) and Booth, Aivazian, Demirguc-Kunt, and Maksimovic (2001) who examine whether firm-level characteristics affect leverage differently in different countries. Consistent with these earlier studies we find that the cross-sectional determinants of leverage are roughly consistent across countries; however, we do find some notable cross-country differences. For example, although past profitability has a negative influence on leverage in all countries, the effect tends to be slightly less in countries where there is a lower tax incentive to retain earnings, and is much stronger in countries that are more corrupt. The latter observation is consistent with the hypothesis that firms in more corrupt countries are less willing to pay out retained earnings, reflecting the greater difficulty associated with raising equity capital in these countries. 34

This result is consistent with Demirguc-Kunt, and Maksimovic (1999).

One should interpret these results with some caution, because an analysis of capital suppliers does raise endogeneity concerns. In particular, we expect financial institutions to develop in ways that satisfy the financing needs of firms. However, as discussed in Section 2.3, we have selected variables that are less likely to be influenced by the capital structure preferences of corporations.

3

The paper is organized as follows: Section 2 discusses the association between country level institutional factors and financial choices. Section 3 introduces the set of firm level variables that influence capital structure choice and discusses how the influences of these firm level variables may vary from country to country. Section 4 describes the sample. Section 5 presents our results and Section 6 draws some conclusions.

2. Institutional factors and cross-country determinants of capital structure

This section discusses how institutional differences between countries can potentially affect how firms within these countries are financed. Specifically, we consider institutional variables that can potentially affect (1) the ability to enforce legal contracts (2) the tax system, and (3) the suppliers of capital.

We expect that weaker legal systems and hence weaker public enforcement

should be associated with less external equity and less complicated debt contracts. We also expect that firms in countries that tax cash flow to equity less will be less levered. Finally, we examine whether the suppliers of capital matter. Although most of the capital structure literature focuses on the financing preferences of firms, at the aggregate level, capital structures are determined by the preferences of the suppliers of capital as well as firms. Therefore, exogenous factors that lead the suppliers of capital to want to hold more or less equity relative to debt will also influence capital structures of firms.

The following sub-sections introduce the variables that we consider, and discuss

how these variables are likely to influence typical debt ratios within our sample of countries.

4

2.1. The legal system

Incentive problems - conflicts of interest between corporate insiders (managers, employees and/or majority shareholders) and external investors - are important factors that shape corporate policy and productivity. As pointed out by La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1998), the extent to which contracts can be used to mitigate these problems depends on the legal system, which consists of both the content of the laws and the quality of their enforcement. In the following we will discuss how these legal system factors influence financing choices.

In countries with weak laws and enforcement, financial instruments that allow

insiders less discretion, and are contractually easier to interpret, are likely to dominate. La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1998) find significant variation in the extent of legal protection of external investors across both developed and developing countries, and argue that legal systems based on common law offer outside investors (debt and equity) better protection than those based on civil law. All else equal, this suggests that common law countries will use more outside equity and longer-term debt. To test whether this is the case, we define a dummy variable that takes a value of one if the country’s legal system is based on common law and zero otherwise.

In addition to the content of the law, the integrity and enforceability of the law is also important, which we measure as the perceived corruption level of a country. Corruption has been identified as a key factor shaping a country’s legal system (Djankov, La Porta, Lopez-de-Silanes and Shleifer, 2003), resource allocation and firm behavior (La

5

Porta, Lopez-De-Silanes, Shleifer, and Vishny, 1999; Fisman, 2001; Johnson and Mitton, 2003).

We are not the first to examine the roles of legal factors in corporate financing choices. Demirguc-Kunt and Maksimovic (1999) find that firms have longer duration debt in countries where the legal system has more “integrity”. Integrity, which is measured by a law and order index prepared by the International Country Risk Guide, reflects the extent to which individuals are willing to rely on the legal system to make and implement laws, mediate disputes and enforce contracts. In contrast, we focus on corruption, defined as the abuse of public office for private gain, measured as the Corruption Perception Index (Transparency International), which reflects the extent to which corruption is perceived to exist among public officials and politicians. An advantage of this index is that it provides both time-series and cross-sectional variation; most other measures of corruption, such as the law and order index, do not have comparable historical data.

We reverse the index, which ranges from 0 to 10, with larger values indicating more severe corruption. In the context of the firm’s capital structure choices, the index proxies for the threat of all or part of investor rights being expropriated by managers or public officials. Debt is expected to be used relatively more than equity when the public sector is more corrupt, since it is easier to expropriate outside equity holders than debt holders. Similarly, one can argue that since short-term debt is more difficult to expropriate, it will be used relatively more than long-term debt in more corrupt countries.

We also consider a possible interaction effect between corruption and the legal system. Specifically, common law may provide shareholder protection only when the

6

legal system has sufficient integrity. If this is the case, then we expect that common law will have less influence on capital structure in more corrupt counties. To examine this possibility we interact the common law dummy with a high corruption dummy variable that is equal to one when the value of the corruption index of a given economy-year is greater or equal to the top quartile of the sample, and zero otherwise.

The above arguments suggest that the legal environment of the country in which a firm is located has an influence on how it is financed. However, firms may be able to offset the effect of a weak legal environment by listing their stock in a country with a stronger legal system. If this is the case, we should observe that cross-listed firms have lower leverage and longer debt maturity than non-cross listed firms. Since this effect is likely to be more pronounced for firms in weakly governed countries, the predicted effects of cross-listing on capital structure and debt maturity is likely to be stronger in more corrupt countries.5

To estimate the effect of cross-listing, we create a dummy variable equal to one if a firm has common stocks traded on both its domestic market as well as on the London Stock Exchange or in the U.S. as an American Deposit Receipts (ADRs). This dummy is included by itself as well as interacted with the high corruption dummy variable specified earlier.

When interpreting this coefficient it should be noted that the direction of causation between cross-listing and leverage could go either way. The above argument suggests that firms that cross-list choose to have more equity in their capital structures 5

This argument was first made by Reese and Weisbach (2002) who find that firms from weakly governed countries cross-list in the U.S. to increase protection of their minority shareholders and that equity offerings increase both in the U.S. and in the firm’s domestic market subsequent to cross-listing. Stulz (2008) generalizes this arguing that securities laws are an important determinant of whether and where securities are issued, how they are valued, who owns them, and where they trade.

7

because cross-listing mitigates some of the incentive problems associated with outside equity. However, it is also possible that firms that desire more equity in their capital structure have an incentive to market their equity more broadly, and are more likely to cross-list. It might also be the case that cross-listing makes it easier for the firms in less developed markets to issue long-term debt. However, since it is unlikely that firms cross-list because of a desire for longer maturity debt, we would argue that evidence that cross-listed firms choose longer maturity structures is due to the better legal environment that arises when they cross-list.

2.2. Tax system

The tax system in general, and specifically the tax treatment of interest and dividend payments, has been recognized as an important factor influencing capital structure choices since the seminal work of Modigliani and Miller (1963).6 We observe three main categories of tax regimes:

The first is the classical tax system in which dividend payments are taxed at both the corporate and personal levels and interest payments are tax-deductible corporate expenses. The classical tax system exists in Brazil, Chile, China, , India, Indonesia, Israel, Japan, Korea, Malaysia, Netherlands, Pakistan, Peru, Philippines, Singapore, South Africa, Switzerland, United Kingdom (post 2000)7 and USA.

The second is the dividend relief tax system, where dividend payments are taxed at a reduced rate at the personal level. A dividend relief tax system exists in Austria,

67

See Graham (2003) for a review of the literature on the influence of taxes on capital structure choice. The United Kingdom reverted to a classical tax system in 2001.

8

Belgium, Denmark, Greece, Portugal, Sweden, Thailand and Turkey.8 In Greece and Turkey dividend payments are not taxed at the personal level, that is, a full dividend relief system.

Third is the dividend imputation tax system, where corporations can deduct interest payments, but where the domestic shareholders of a corporation receive a tax credit for the taxes paid by the corporation. The goal of the system is to tax corporate profits only once. Dividend imputation systems are in place in Australia, Canada, Finland, France, Germany, Ireland, Italy, Mexico, New Zealand, Norway, Spain, Taiwan and United Kingdom (pre 2001). The proportion of corporate tax available as a tax credit under these imputation systems varies from country to country. In Australia, Germany, Italy, New Zealand and Norway the full amount of the corporate tax paid is distributed as a tax credit. In other countries only part of the corporate tax credits are distributed.

All else equal, we expect that debt will be used less in countries with dividend imputation or tax relief systems than in countries with classical tax systems that double tax corporate profits. To test for this relationship we define a dividend tax dummy that takes a value of one for countries with either a full dividend relief tax system or a full dividend imputation tax system and zero otherwise.

2.3. Preference of the suppliers of capital

Financial economists have typically viewed the capital structure problem from the perspective of firms that face competitive and complete financial markets that offer debt and equity capital at competitive risk-adjusted rates. However, when this is not the case, the preferences of investors to hold debt versus equity instruments will have an influence 8

The United States currently provides preferential tax treatment for dividend, but not in our sample period.

9

on how firms are financed. For example, in the Miller (1977) model, where individual capital structure choices by firms are irrelevant, the aggregate debt ratio in the economy is determined by investor preferences for holding debt versus equity securities. While these preferences are determined by taxes in Miller’s model, one can more generally consider how investor preferences for holding various debt and equity instruments affect the capital structure choice of firms.9

We will specifically be considering the preferences of insurance companies and banks. Insurance companies have a comparative advantage in holding longer-term securities, since they have long-term obligations. In contrast, banks tend to have short-term liabilities and may thus have a comparative advantage in holding short-term debt. Hence, one might expect that firms in countries with a larger banking sector to have more short-term debt and that in countries with larger life insurance industries we might expect to see firms with more equity and/or long-term debt in their capital structures.

Further, the analysis of supply effects raise endogeneity concerns, since we expect financial intermediaries to develop in ways that satisfy the financing needs of firms. Existing studies (for example, Dermirguc-Kunt and Maksimovic, 1999) examine the effects of stock market size, turnover and bank total assets on capital structure choices. These variables, however, are likely to be influenced by the capital structure preferences of corporations. For example, in countries with industries (like high tech) that require considerable amounts of external capital, the stock market is likely to be larger.10 With 9

See Titman (2002) for a discussion of the effect of investor preferences on capital structure choices. Demirguc-Kunt and Maksimovic (1999) recognise this endogeneity issue and address it by using a two-stage instrumental variable regression. They chose as instruments measures of the size of the economy and the flow of funds, plus proxies for the content, strength and integrity of the legal system. However, one can argue that these variables either directly influence the capital structure choice or are potentially influenced by the types of firms in the economy, and are thus indirectly related to the capital structure choice.

10

10

this in mind, we depart from the existing literature and select proxies that are not likely to be directly influenced by the capital structure preferences of corporations. In particular, we select measures of the supply of funds available to these financial intermediaries.

To proxy for the supply of funds in the banking and insurance sectors we include the following variables in our regressions. We use deposits (or liquid liabilities)/GDP to measure the amount of short-term funds that are available to the banking sector,11 and life insurance penetration (value of life insurance premiums/GDP) to measure the amount of long-term funds that are available to insurance companies.

3. Firm Level Characteristics, Interactions and Capital Structure Choice 3.1 Firm Level Variables

Consistent with the existing literature (Titman and Wessels, 1988; Guedes and Opler, 1996; Rajan and Zingales, 1995) we include a set of firm level variables that capture factors that are known to affect leverage and maturity structure. These variables include asset tangibility (fixed assets over total assets), profitability (net income over total assets), firm size (natural logarithm of total assets) and the market-to-book ratio (market value of equity over book value of equity). In addition, asset maturity (gross property, plant and equipment over total assets times gross property, plant and equipment over depreciation) is included as a determinant of maturity structure at the firm level. Due to data 11

It is possible that there are unobserved factors that affect both the willingness of investors to deposit funds with banks and the willingness of banks to provide long-term funding to firms, creating a spurious relation between deposits and capital structure. For example, one can argue that the financing needs of corporations affect the funds that are available to the different investor sectors. Suppose, for example, that the need for monitoring declines, making bank loans somewhat less attractive to long-term bonds. On the margin, this would increase the interest rate on long-term bonds, making it more attractive for households to invest in fixed income mutual funds rather than bank deposits. While this creates a potential endogeniety problem, it is mitigated by the inclusion of our institutional variables and probably has a minor influence on our estimates.

11

limitations in some of the countries included in our sample, we do not include variables that measure the effective tax rate, operating risk, research and development expenditure, capital expenditure and selling expenses as per Titman and Wessels (1988). In place of these variables we include the market-to-book ratio, which can proxy for growth as well as the collateral value of assets, and industry dummy variables based on two-digit SIC codes.

3.2. Interactive effects of institutional and firm factors

In addition to including these firm-specific variables on their own, we also interact the variables with country variables. These interactions allow us to examine how the cross-sectional determinants of capital structure vary from country to country. Our main interest here is on taxes and the legal system, which are likely to affect the financing choices of different firms differently within an economy. Specifically, we examine the extent to which collateral and past profits influence capital structures differently in different countries.

3.2.1 Past profits

One of the strongest determinants of capital structure found in previous studies is past profitability; firms that were more profitable in the past tend to have lower debt ratios. As we will discuss in Section 5.1.1., this tendency is present in all of the countries that we examine.

The literature provides two explanations for this phenomenon. The first is a tax argument; profitable firms have an incentive to retain earnings to avoid the personal taxes

12

associated with distributions (Auerbach, 1979). The second is based on asymmetric information; issuing equity is costly because of asymmetric information problems, and as a result, firms tend to accumulate financial slack when they are profitable (Myers and Majluf, 1984).

The tax effect is straightforward to test. Our hypothesis is that in those countries that do not double tax distributions, e.g., imputation countries, past profit will have less of an effect on capital structure. To test for this relationship we interact the dividend tax dummy variable with profitability.

To examine the effect of asymmetric information we conjecture that information asymmetries are likely to be more of an issue in countries with weaker legal systems. As a result, the tendency to distribute profits should be higher in common law countries and less corrupt countries. Consistent with this idea, La Porta, Lopez-de-Silanes, Shleifer, and Vishny (2000) find that dividend payouts tend to be higher in common law countries.12 To test whether this behavior has an effect on observed capital structures we interact the common law and low corruption dummy variables with profitability.

3.2.2 Collateral Value

In general, firms with assets that can be collateralized can more easily raise debt financing, particularly long-term debt (Guedes and Opler, 1996). As we will discuss in Section 5.1.1., the tendency of firms with greater collateral value to have higher debt ratios is present in almost all of the countries that we examine. We conjecture that 12

La Porta, Lopez-de-Silanes, Shleifer, and Vishny (2000) argue that investors in common law countries use their stronger legal powers to extract dividends from firms. Alternatively, we conjecture that stronger investor protection reduces the impact of information asymmetry providing better access to equity financing in common law countries. Firms aware of this easier access are willing to pay higher dividends.

13

collateral value will be more important in countries with weaker legal systems since higher levels of collateral restrict the opportunities for insiders to expropriate the wealth of outside investors. To test whether or not this is the case, we interact the common law and low corruption dummy variables with two proxies for the collateral value of assets, assets tangibility and the market-to-book ratio.

4. Data and sample

This section describes the sample and presents the country and industry patterns of capital and debt maturity structures. It then introduces the empirical procedure employed in this study.

4.1. Sample selection

The primary source of our firm-level data is Worldscope, which contains financial data on companies from a wide range of industries in over 50 countries. We restrict the sample to those firms listed on the stock market of the country in which it is domiciled.

Our analysis covers the period of 1991 through 2006. We exclude firm-year observations with missing financial data that is required for the firm-level analysis. The final sample consists of 36,767 firms from 39 countries, totalling 272,092 firm-years. Table 1 provides a description of the sample. The sample covers a broad cross section of developed and developing countries with every continent represented. Most of the countries have observations in each of the 16 years.

As can be seen from the last two columns of Table 1, the coverage of the sample firms varies across countries in terms of number and/or market capitalization, reflecting

14

that Worldscope has uneven coverage of firms across the countries.13 For most of the economies we have more than 60 percent of sample coverage in terms of market capitalization and 50 percent in terms of number of listed firms. The economies with lower data coverage tend to be developing economies.

[Table 1 about here]

4.2. Country financing patterns Our measures of capital structure are: (i)

leverage, measured as the proportion of total debt to market value of the firm (total debt/market value). Total debt is defined to be the book value of short-term and long-term interest bearing debt. Market value of the firm is defined as the market value of common equity plus book value of preferred stock plus total debt, or (ii)

debt maturity, measured as the proportion of long-term debt to total debt (long-term debt/total debt).14

To gain a basic idea about how capital and maturity structures differ across countries, we compute the median leverage and maturity structure by country for the period 1991 to 2006. As can be seen in Figure 1, developing economies occupy both ends of the leverage spectrum, measured as the book value of debt over the market value of the firm. The highest five leverage ratios are observed in South Korea, Indonesia, Brazil, Portugal, 13

Worldscope biases the sample toward large listed companies in each country. This can be seen by comparing the number and value coverage of the sample in the last two columns of Table 1: value coverage is generally higher than number coverage across the countries. 14

Trade credit is an important source of financing in economies with underdeveloped financial institutions (Demirguc-Kunt and Maksimovic, 2001; Fisman and Love, 2003). Our results are robust to including trade credit (measured as accounts payable) in our measure of short-term debt.

15

and Pakistan, while the lowest five are observed in Australia, South Africa, Canada, the United States, and Turkey. Developing economies seem to dominate the higher range, while developed economies tend to be at the lower range. The median leverage ratio for the developing economies in the sample is 0.26,15 while for the developed economies the median leverage ratio is 0.20. The middle range of the leverage spectrum is mixed with both developing and developed economies.

[Figure 1 about here]

Figure 2 presents the median maturity structure by country, measured as the book value of long-term to total debt. It is clear from the figure that debt obligations have longer maturities in more developed economies. The five countries with the highest long-term debt ratios are New Zealand, Norway, Sweden, USA, and Canada. The lowest five median long-term debt ratios are observed in China, Greece, Turkey, Taiwan, and Thailand.16 The median long-term debt ratio for the developing economies in the sample is 0.36, while for the developed economies the median long-term debt ratio is 0.61.

[Figure 2 about here]

In addition to the set of firm and country-level variables described in Section 2, we include inflation and a developed economy dummy variable that takes a value of one if the country is classified as a developed economy according to the World Bank classification that is based on the countries’ gross national income levels.17 Inflation is included because debt contracts are generally nominal contracts and high inflation, which 15

Economies within the sample classified as developing, according to the World Bank, are Brazil, Chile, China, India, Indonesia, Malaysia, Mexico, Pakistan, Peru, Philippines, Taiwan, Thailand, Turkey and South Africa. 16

This parallels the findings of Demirguc-Kunt and Maksimovic (1999) for an early sample period, 1980-1991. 17

The set of country level independent variables are defined in Appendix 1, along with their data sources.

16

is generally associated with high uncertainty about future inflation, may tilt lenders away from long-term debt. A developed economy dummy variable is included because it may pick up an element of economic development that is not already captured by our other variables. Both firm and country level variables are lagged one period to allow for the non-contemporaneous nature of the interaction between firm/country level characteristics and financing choices.

Table 2, which presents the summary statistics, shows cross-sectional variation in the country-level variables. The country-level variables are defined in Appendix 1, along with their data sources. Except for the common law, developed economy, and tax system variables that are constant across time, all of the variables exhibit time-series variation.18 Appendix 2 reports the country-by-country median values of the country-level explanatory variables.

[Table 2 about here]

To gain a basic understanding of how capital and maturity structures are

influenced by these variables, we compute the Pearson correlation coefficients for pairs of the dependent and independent variables. The results, reported in Table 3, suggest that the legal system, the tax system, and the suppliers of funds potentially influence the capital structure choice. In particular, common law is associated with lower leverage and more long-term debt; low levels of corruption are associated with lower leverage ratios and a greater use of long-term rather than short-term debt. In addition, firms in countries that tax dividends less tend to have lower leverage ratios. Firms in countries with more bank deposits tend to have higher leverage and more short-term debt. However, life 18

The corruption index prior to 1995 is taken as the 1988-1992 composite level, because compatible annual data is not available prior to 1995.

17

insurance penetration is insignificantly related to leverage and negatively related to debt maturity, inconsistent with our predictions. We will return to this issue later in regression analysis. As expected firms in more developed economies have lower leverage ratios and more long-term debt. Inflation rate is only weakly correlated with both leverage and maturity structures.

[Table 3 about here]

To investigate whether our selected country factors are likely to be subject to collinearity problems in our later regression analysis, we examine the correlations between the independent variables that are used in our analysis. We find that these variables are generally not highly correlated with each other. An exception is the negative 76 percent correlation between the corruption index and the economic development dummy variable, suggesting that we should properly control for spurious effects of economic development in subsequent analysis.

5. Regression analysis

This section presents regressions that estimate the influence of country-level explanatory variables on capital structure choices controlling for firm- and industry-level characteristics. Our regressions are estimated with a General Methods of Moments (GMM) approach that accounts for the fact that the regression residuals are heteroskedastic and correlated across both firm and country level observations.19 19

The regressions are performed on panel data where the residuals may be correlated across firms and/or

across country, and OLS standard errors can be biased. We use the ordinary least square (OLS) method with heteroscedastic / autocorrelation corrected (HAC) errors (Andrew, 1991) and clustered at the country level (Petersen, 2008). The HAC procedure accounts for the potential heteroscedasticity and auto-correlation at the firm level by deriving the t-statistics of estimated OLS coefficients from Generalized Methods of Moments (GMM) standard errors corrected for heteroscedasticity and auto-correlation.

18

5.1. The determinants of leverage

Table 4 presents the results of the leverage regressions.20 Column one reports the regression for the full sample, column two provides evidence for the sub-sample of developed economies only and column three the sub-sample of developing economies only. Columns four and five provide evidence for the sub-periods, 1991-1998 and 1999-2006, respectively.

[Table 4 about here]

5.1.1 Firm effects

The top half of Table 4 reports the coefficient estimates of our firm-specific variables. These coefficient estimates indicate that leverage is positively related to asset tangibility and firm size and negatively related to profitability, and the market-to-book ratio. These results, which hold in the full sample as well as the sub-samples, are consistent with evidence on U.S. firms (Bradley, Jarrell, and Kim, 1984; Titman and Wessels, 1988) and more recent international evidence (Rajan and Zingales, 1995 and Booth, Aivazian, Demirguc-Kunt, and Maksimovic, 2001). These results are also generally consistent with individual country leverage regressions that we report in Appendix 3. With the exception of asset tangibility and size, the coefficients have the same sign in all country regressions. Asset tangibility and size are positively related to leverage in 37 out of 39 countries. However, there is cross-country variation in the significance of these coefficients that we will be addressing later when we discuss the interaction between firm variables and country variables.

20

The results are robust to the use of alternative proxies for the country’s legal system, corruption, taxation and financial market development. Alternative proxies leave unaffected other estimated coefficients.

19

5.1.2. Country effects

The lower half of Table 4 reports coefficient estimates for country variables. These estimates indicate that leverage is positively related to economic development but unrelated to inflation, which is consistent with Booth, Aivazian, Demirguc-Kunt, and Maksimovic (2001).21 Consistent with better investor protection leading to a greater use of equity financing we find that corruption is associated with higher debt ratios and common law systems are associated with lower debt ratios. Consistent with the hypothesis that the better protection of shareholders offered by common law has little effect on shareholder protection in highly corrupt countries, we find a significant positive sign on the corruption/common law interaction term.

We find that leverage is lower in countries that tax dividends less.22 However, we do not find a significant relation between leverage and either the size of the life insurance industry or the size of the banking system, measured as deposit over GDP.

In general the results in the subsamples are similar. However, we find that common law is associated with lower debt ratios in developed economies, but is associated with higher debt ratios in developing economies. Since corruption is higher in developing economies, this result is consistent with the advantages of common law having less efficacy in more corrupt countries.

We find that life insurance penetration is negatively related to leverage in the developing economy sub-sample. One interpretation of this evidence is that insurance 21

The positive relation between inflation and leverage in developing economies disappears when Brazil is dropped from the sample. Brazil had very high inflation and its companies were highly levered during our sample period. 22

For example, Twite (2001) finds that the introduction of a dividend imputation tax system into Australia led to a reduction in leverage.

20

companies hold more equity in developing countries. This could reflect the lack of a developed bond market and/or tighter restrictions on their holdings of bonds relative to equities in the less developed countries.23

Likewise, the results in the sub-periods are similar. However, we find that the influence of common law on leverage is weaker in the later sub-period.

5.1.3. Firm and country interaction effects

Table 5 extends the leverage regressions shown in Table 4 to include various interaction variables. In addition, we include a dummy variable indicating whether the firm’s equity is either cross-listed or has an ADR. We find that leverage is negatively related to this dummy.

Specifically, we interact asset tangibility, profitability and the market-to-book ratio with legal and tax variables. In addition, to investigate the role of economic development we interact asset tangibility, profitability and the market-to-book ratio with the developed economy dummy variable. We also interact the ADR/cross-listing dummy with the high corruption dummy variable.

23

While the set of countries for which we can observe data is limited (19 developed and 5 developing countries), we find in developing economies that bond market activity is lower and regulatory restrictions are tighter than in developed economies. We measure the level of bond market activity in a country as bond issued (both government and corporate) over GDP as taken from the International Financial Statistics, International Monetary Fund, defining a dummy variable that takes a value of one if the country’s bond market activity in a given year is greater or equal to the top quartile of the sample and zero otherwise. To investigate the possibility that cross country differences in insurance regulations influence the holdings of insurance companies we focus on an index of relative restrictions on debt and equity holdings measured as the ratio of the proportional limit on equity holdings over the proportional limit on debt holdings taken from the Survey of Investment Regulation of Pension Funds, OECD, with larger values indicating tighter restriction on bond holdings. Specifically, the mean value of the high bond market activity dummy is 0.24 and the investments restriction index is 0.99 for developing economies. In comparison, the mean value of the high bond market activity dummy is 0.55 and the investments restriction index is 0.72 for developed economies.

21

As a consequence of multicollinearity between the groups of interaction effects we cannot include all variables in a single regression. Rather, we estimate separate regressions for each interaction effect: dividend tax, common law and corruption. In column one, we include a variable that interacts profitability with the dividend tax dummy variable. Columns two and three include variables that interact profitability, asset tangibility and the market-to-book ratio with the common law and low corruption dummy variables, respectively. In column four, we include the cross-listing dummy variable and interact the variable with the high corruption dummy variable.

[Table 5 about here]

Consistent with the Auerbach (1979) hypothesis that personal taxes induce firms to retain earnings, we find that the negative relation between profitability and leverage is weaker in countries where dividends are preferentially taxed. However, the legal system and the corruption of a country have a stronger influence on the profitability coefficient. Consistent with our earlier conjecture, in legal systems that protect shareholders more, past profitability has less of an influence on the debt ratio. Our evidence on the interaction between collateral and institutional structures is mixed. We find that collateral, as measured by the market-to-book ratio, is more important in more corrupt countries and weaker legal systems. However, the interaction terms that include asset tangibility, another measure of collateral, are not statistically significant. Finally, we fail to find a significant relationship between the country legal system and the magnitude of the ADR/cross-listing effect.

5.2 Determinants of maturity structure

22

5.2.1 Firm effects

Table 6 reports regressions that examine the determinants of maturity structure.24 Column one provides evidence for the full sample, column two the sub-sample of developed economies only and column three the sub-sample of developing economies only. Columns four and five provide evidence for the sub-periods, 1991-1998 and 1999-2006, respectively.

[Table 6 about here]

The coefficients of the firm-specific variables are largely consistent with prior research (Barclay and Smith, 1995; Stohs and Mauer, 1996; Guedes and Opler, 1996; Demirguc-Kunt and Maksimovic, 1999) in the full sample and both sub-samples. Long-term debt is used more by firms with greater asset tangibility, larger size, higher profits and higher market-to-book ratios. However, we find that asset maturity is unrelated to debt maturity.

Appendix 4 reports the results of the country-by-country debt maturity regressions. The most robust cross-sectional determinants of debt maturity are firm size and asset tangibility. With only few exceptions, size and asset tangibility are positively related to debt maturity structure. On the other hand, we find cross-country variation in the sign of the estimated coefficients for profitability, the market-to-book ratio and asset maturity. Profitability is positively related to debt maturity structure in 25 out of 39 countries. The market-to-book ratio is positively related to debt maturity structure in 26 out of 39 countries and asset maturity is positively related to debt maturity structure in only 12 out of 39 countries. 24

The results are robust to the use of alternative proxies for the country’s legal system, corruption, taxation and financial market development.

23

5.2.2 Country effects

The estimates of the country level coefficients reveal that debt maturity is negatively related to the level of corruption and positively related to the common law dummy variable, consistent with lower corruption and stronger investor protection encouraging the use of long-term debt financing. In addition, we find that the positive influence of investor protection on the use of long-term debt financing is stronger in more corrupt countries.

Consistent with the preferences of the suppliers of capital having an influence on the firms’ maturity structures, we find that debt maturity is strongly negatively related to the amount of deposits in the country’s banking sector. This is in contrast to the negative but insignificant banking sector result reported by Demirguc-Kunt and Maksimovic (1999). This difference may in part be explained by differences in the variables used to proxy for the size of the banking sector, as discussed in section 2.3. We find no reliable relation between maturity structure and the degree of life insurance penetration.

We find that debt maturity is positively related to the level of economic development. However, inconsistent with expectations we find a significant positive relation between debt maturity and inflation, which is inconsistent with the findings in Demirguc-Kunt and Maksimovic (1999).25 Finally, the results in the developed and developing country subsamples and the two sub-periods are similar.

5.2.3. Firm and country interaction effects 25

The result that inflation being positively related to leverage in developing economies is driven by both the low inflation/short-term maturity characteristic of China and the high inflation/long-term maturity characteristic of Brazil. After dropping Brazil and China, inflation is insignificantly related to maturity.

24

Table 7 reports the maturity structure regressions that include interactions between firm and country variables. In general, the results for the firm and country-level variables are consistent with those reported in Table 6. Columns one and two of this table include variables interacting asset tangibility and the market-to-book ratio with the common law, low corruption and developed economy dummy variables, respectively. Columns three reports results that include the cross-listing dummy variable and its interactive effect with the high corruption dummy variable.

[Table 7 about here]

Our evidence on the interaction between collateral and institutional structures is mixed. We find that collateral, as measured by tangible assets to total assets, is less important in stronger legal systems. However, we do not find the coefficient of the market-to-book ratio to be reliably different in common law countries. Finally, we find that cross-listing is associated with the use of longer-term debt and the effect is stronger in weakly governed countries. This result is consistent with cross-listing enhancing access to long-term debt.

5.3. Fixed-Effects and Cross-Sectional Estimates

This section examines the extent to which the cross-sectional and time-series variation in our explanatory variables drives our results. Up to this point our emphasis has been on the cross-sectional variation in capital structures. However, the debt ratios in individual countries also vary from year to year, and some of that year to year variation may be explained by the year to year changes in our explanatory.

To estimate the extent to which our results are generated from the cross section versus the time series we estimate both fixed-effects and cross-sectional regressions.

25

Specifically, we estimate fixed-effects leverage and maturity structure regressions in columns one and two of Table 8, respectively, and the cross-sectional leverage and maturity structure regressions in columns three and four, respectively. By sweeping out individual firm and country-effects, the fixed-effects regression estimates the extent to which the time-series variation of our independent variables explains the time-series of capital structure choices. In contrast, the cross-sectional regression uses firm and country-level averages for all dependent and independent variables, thereby isolating the cross-sectional determinants of capital structure.

[Table 8 about here]

The regression estimates reported in Table 8 indicate that the relationships between financing choices and firm characteristics are significant in both the time-series and the cross-section, and are consistent with our earlier estimates. The relationships between financing choices and country variables are also significant in the cross-sectional regressions, and the estimates are consistent with our earlier regressions. However, in most cases the time-series relation between the country variables and financing choices are insignificant, which reflects the fact that there is limited time-series variation in most of our country-level variables.

6. Summary and Conclusion

At the outset, we described regression results that indicate that a corporation’s capital structure is determined more by the country in which it is located than by its industry affiliation, suggesting that public policy and institutional differences between countries can have a profound effect on how firms are financed. Specifically, we find that a

26

country’s legal and taxation system, corruption and the preferences of capital suppliers – banks and life insurance – explain a significant portion of the cross-country variation in leverage and debt maturity ratios.

The effects of taxes on capital structure choices are very much in line with our expectations. When dividends are more highly taxed, firms tilt their capital structures towards more debt. The legal environment also has an important influence on capital structure choices. Our strongest finding is that firms in countries that are viewed as more corrupt tend to be more levered and use more short-term debt. We also find that common law countries have lower leverage and use more long-term debt. Further, we find that firms that choose to cross-list tend to use more equity and longer-term debt and the influence of cross-listing on maturity is stronger in weakly governed countries.

Our evidence also indicates that financial institutions, which provide capital to corporations, have an important influence on the type of capital that is used. Although our results regarding life insurance companies are somewhat mixed and difficult to interpret, our results that relate to the size of the banking sector are both strong and intuitive. Specifically, corporations in countries with large amounts of bank deposits tend to have shorter maturity debt, reflecting the preferences of banks to lend short-term.

Our results also indicate that the well-known relationship between profitability and leverage varies across countries. While we find that more profitable firms have lower leverage, the result is weaker in countries where dividends are preferentially taxed, which is consistent with the idea that personal taxes can influence payout policies and that these can in turn influence observed capital structures. However, the tax/profitability result is relatively weak. We find that corruption and the legal system has a stronger

27

influence on the extent to which past profits effects capital structure. Specifically, past profits have a greater influence on capital structure in more corrupt countries with legal systems that offer weaker shareholder protection. Finally, we find mixed support for collateral being more important in more corrupt countries with legal systems that offer weaker shareholder protection.

Although our emphasis has been on the effect of cross-country differences in institutions on corporate financial choices, our analysis may have implications that relate to the literature on how institutions can promote economic growth.26 Specifically, the fact that institutions influence how firms are financed may provide an indirect channel through which a country’s institutions affect economic growth. For example, there is reason to believe that if firms can raise more of their capital with equity and long-term debt they will be better able to make longer-term investments which may better promote economic growth. This suggests that an analysis of the relation between investment horizons and institutional structure will offer an interesting avenue for future research.

26

Demirguc-Kunt and Maksimovic (1998), Levine and Zervos (1998) and Rajan and Zingales (1998) find that, for a sample of developing and developed countries, the development of stock markets, bond markets and banks facilitate economic growth.

28

References:

Andrew, D.W.K., 1991. Heteroskedasticity and Autocorrelation Consistent Covariance Matrix Estimation. Econometrica 59, 817-858.

Antoniou, A., Guney, Y., Paudyal, K., 2002. Determinants of Corporate Capital Structure: Evidence from European Countries. Unpublished working paper. University of Durham. Auerbach, A. J., 1979. Share Valuation and Corporate Equity Policy. Journal of Public Economics 11, 291-305.

Barclay, M.J., Smith, C.W., 1995. The Maturity Structure of Corporate Debt. Journal of Finance 50, 609-631.

Booth, L., Aivazian, V., Demirguc-Kunt, A., Maksimovic, V., 2001. Capital Structures in Developing Countries. Journal of Finance 56, 87-130.

Botero, J., Djankov, S., La Porta, R. Shleifer, A., 2004. The Regulation of Labor. Quarterly Journal of Economics 119, 1339-1382.

Bradley, M., Jarrell, G.A., Kim, H., 1984. On the Existence of an Optimal Capital Structure. Journal of Finance 39, 857-878.

Bronars, S. Deere, D., 1991. The Threat of Unionization, the Use of Debt and the Preservation of Shareholder Wealth. Quarterly Journal of Economics 106, 231-2. Campbell, J., Hamao, Y., 1995. Changing Corporate Financing Structure and the Main Bank System in Japan. In: Aoki, M., Patrick, H. (Eds.), Japanese Main Bank System. Oxford University Press.

Caprio, G., Laeven, L., Levine, R., 2005. Governance and Bank Valuation. Unpublished working paper. World Bank and University of Minnesota.

Claessens, S., Djankov, S., Nenova, T., 2001. Corporate Risk around the World. Unpublished working paper. World Bank, CEPR, and Harvard University.

Dasgupta, S. Sengupta, K., 1993. Sunk Investment, Bargaining and Choice of Capital Structure. International Economic Review 34, 203-220.

Demirguc-Kunt, A., Maksimovic, V., 1996. Stock Market Development and Firm Financing Choices. Word Bank Economic Review 10, 341-369.

Demirguc-Kunt, A., Maksimovic, V., 1998. Law, Finance and Firm Growth. Journal of Finance 53, 2107-2137.

Demirguc-Kunt, A., Maksimovic, V., 1999. Institutions, Financial Markets, and Firm Debt Maturity. Journal of Financial Economics , 295-336.

29

Demirguc-Kunt, A., Maksimovic, V., 2001. Firms as Financial Intermediaries: Evidence from Trade Credit Data. Unpublished working paper. World Bank and the University of Maryland.

Diamond, D., 1993. Seniority and Maturity of Debt Contracts. Journal of Financial Economics 33, 341-368.

Dinc, I.S., 2005. Politicians and Banks: Political Influences on Government-owned Banks in Emerging Markets. Journal of Financial Economics 77, 453-479.

Djankov, S., La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 2003. Courts. Quarterly Journal of Economics 118, 453-517.

Fisman, R., 2001. Estimating the Value of Political Connections. American Economic Review 91, 1095-1102.

Fisman, R. Love, I, 2003. Trade Credit, Financial Intermediary Development and Industry Growth. Journal of Finance 58, 353-374.

Flannery, M., 1986. Asymmetric Information and Risky Debt Maturity Choice. Journal of Finance 41, pp. 19-37.

Gatward, P. Sharpe, I., 1996. Capital Structure Dynamics with Interrelated Adjustment: Australian Evidence. Australian Journal of Management 21, -112.

Giannetti, M., 2003. Do Better Institutions Mitigate Agency Problems? Evidence from Corporate Finance Choices. Journal of Financial and Quantitative Analysis 38, 185-212. Graham, J.R., 1995. Debt and the Marginal Tax Rate. Journal of Financial Economics 41, 41-73.

Graham, J.R., 2003. Taxes and Corporate Finance: A Review. Review of Financial Studies 16, 1075-1129.

Guedes, J., Opler, T., 1996. The Determinants of the Maturity of Corporate Debt Issues. Journal of Finance 51, 1809-1833.

Himmelberg, C., Hubbard, R.G., Palia, D., 1999. Understanding the Determinants of Managerial Ownership and the Link Between Ownership and Performance. Journal of Financial Economics 53, 353-384.

Johnson, S., Mitton, T., 2003. Cronyism and Capital Controls: Evidence from Malaysia. Journal of Financial Economics 67, 351-382.

Khawaja, A.L., Mian, A., 2005. Do Lenders Favor Politically Connected Firms? Rent Provision in An Emerging Financial Market. Quarterly Journal of Economics 120, 1371-1411.

30

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1997. Legal Determinants of External Finance. Journal of Finance 52, 1131-1150.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1998. Law and Finance. Journal of Political Economy 106, 1113-1155.

La Porta, R., Lopez-De-Silanes, F., Shleifer, A., 1999. Corporate ownership around the world. Journal of Finance , 471-518.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 2000. Agency Problems and Dividend Policies around the World, Journal of Finance 55, 1-33.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 2002. Government Ownership of Banks. Journal of Finance 57, 265-301.

Levine, R., Zervos, S., 1998. Stock Markets, Banks and Economic Growth. American Economic Review 88, 537-558.

MacKay, P., Phillips, G.M., 2005. How does industry affect firm financial structure? Review of Financial Studies 18, 2005, 1433-1466.

Miguel, A., Pindado, J., 2001. Determinants of capital structure: new evidence from Spanish panel data. Journal of Corporate Finance 7, 77–99.

Miller, M.H., 1977. Debt and Taxes. Journal of Finance, 32, 261-275.

Modigliani, F., Miller, M.H., 1958. The Cost of Capital, Corporate Finance, and the Theory of Investment. American Economic Review 48, 261-297.

Modigliani, F., Miller, M.H., 1963. Corporate Income Taxes and the cost of Capital: A Correction. American Economic Review 53, 433-443.

Myers, S., 2001. Capital Structure. Journal of Economic Perspective 15, 81-102. Myers, S., Majluf, N., 1984. Corporate Financing and Investment Decisions When Firms Have Information That Investors Do Not Have. Journal of Financial Economics 13, 187-221.

Ozkan, A., 2001. Determinants of capital structure and adjustment to long run targets: evidence from UK company panel data. Journal of Business Finance and Accounting 28, 175-198.

Perotti, E., Spier, K., 1993. Capital Structure as a Bargaining Tool: The Role of Leverage in Contract Renegotiation. American Economic Review 83, 1131-1141.

Petersen, M.A., 2008. Estimating Standard Errors in Finance Panel Data Sets: Comparing Approaches. Review of Financial Studies (forthcoming).

31

Rajan, R., Zingales, L., 1995. What Do We Know about Capital Structure? Some Evidence from International Data. Journal of Finance 50, 1421-1460.

Rajan R., Zingales, L., 1998. Financial Development and Growth. American Economic Review 88, 559-586.

Reese, W., Weisbach, M., 2002. Protection of minority shareholder interests, cross-listings in the United States, and subsequent equity offerings. Journal of Financial Economics 66, 65-104.

Sapienza, P., 2004. The Effect of Government Ownership on Bank Lending. Journal of Financial Economics 72, 357-384.

Schmukler, S., Vesperoni, E., 2001. Globalization and Firms’ Financing Choices: Evidence from Emerging Economies. Unpublished working paper. World Bank and International Monetary Fund.

Shleifer, A., Vishny, R., 1998, The Grabbing Hand: Government Pathologies and their Cures. Cambridge, MA: Harvard University Press.

Smith, C.W., Warner, J.B., 1979. On Financial Contracting: An Analysis of Bond Covenants. Journal of Financial Economics 7, 117-161.

Stohs, M., and Mauer, D., 1996. The Determinants of Corporate Debt Maturity Structure. Journal of Business 69, 279-312.

Stulz, R., 2008. Securities Laws, Disclosure, and National Capital Markets in the Age of Financial Globalization. ECGI - Finance Working Paper 2008-13.

Titman, S., Wessels, R., 1988. The Determinants of Capital Structure Choice. Journal of Finance 43, 1-19.

Titman, S., 2002. The Modigliani and Miller Theorem and Market Efficiency. Financial Management 31, 101-115.

Treisman, D., 2000. The Causes of Corruption: A Cross National Study. Journal of Public Economics 76, 399-457.

Twite, G., 2001. Capital Structure Choices and Taxes: Evidence from the Australian Dividend Imputation Tax System. International Review of Finance 2, 217-234.

32

Figure 1

Median leverage ratio of sample firms (1991-2006)

This figure plots the median leverage ratio across 39 different countries. The leverage ratio is measured as total debt over the market value of the firm. Total debt is defined to be the book value of current and long-term interest bearing debt. Market value of the firm is defined to be the market value of common equity plus book value of preferred stock plus total debt.

33

Figure 2

Median long-term debt ratio of sample firms (1991-2006)

This figure plots the median debt maturity ratio across 39 different countries. The debt maturity ratio is measured as long-term interest bearing debt over total debt. Total debt is defined to be the book value of current and long-term interest bearing debt.

34

Table 1 The sample

The table provides a description of the sample. The number of years that data is available for each country. The mean number of firms per year for each country. The median value of the proportion of firms represented in the sample for each country, by number of firms and market capitalization.

Country Australia Austria Belgium Brazil Canada Switzerland Chile China Germany Denmark Spain Finland France UK Greece Indonesia India Ireland Israel Italy Japan Korea Mexico Malaysia Netherlands Norway

New Zealand Pakistan Peru

Philippine Portugal Singapore Sweden Thailand Turkey Taiwan USA

South Africa

Number of years of data

used 16 16 16 16 16 16 16 13 16 16 16 16 16 16 16 16 16 15 16 13 16 16 16 16 16 16 16 16 16 14 16 16 16 16 16 15 16 16 16

Number of firms in the sample 15 139 169 351 1865 274 158 1530 1011 208 223 175 1205 2861 321 939 295 637 109 181 343 4088 970 151 1011 280 266 134 114 74 188 110 628 447 481 201 1399 11119 558

Firm-years 8308 1144 1485 2591 10988 2656 1424 6827 9209 2123 2315 1684 96 21785 2511 7108 2573 4388 880 949 2810 42611 6741 1230 7586 2612 1826 9 1061 491 18 867 4111 3394 3457 1422 7051 77909 3699

Time series median value

Number of firms in the Market capitalization of firms in sample/Total number of the sample/Stock market

listed firms capitalization 0.50 0.79 0.75 0.61 0.59 0.49 0.46 0.87 0.34 0.79 0.72 0.77 0.41 0.58 0.57 0.44 0.90 0.69 0. 0.88 0.17 0.48 0.95 0.83 0.87 0.73 0.67 0.62 0. 0.53 0.62 0.85 0.62 0.70 0.07 0.39 0. 0.49 0.15 0.36 0.73 0.55 0.97 0.86 0.40 0.72 0.51 0.92 0.71 0.86 0.73 0.88 0.77 0.92 0.49 0.94 0.11 0.42 0.20 0.55 0.53 0.81 0.67 0.61 0.76 0.82 0.86 0.91 0.60 0.73 0.40 0.74 0.68 0.74 0.81 0.81 0.53 0.78

35

Table 2

Summary statistics

The table provides the mean, standard deviation, median, minimum and maximum values of each variable. Leverage ratio is the ratio of total debt to market value of the firm. Total debt is defined to be the book value of short-term and long-term interest bearing debt. Market value of the firm is defined to be the market value of common equity plus book value of preferred stock plus total debt. Maturity structure ratio is the ratio of long-term debt to total debt. Tangible assets/total assets is the ratio of fixed assets to total assets, operating risk is measured as the absolute value of the annual change in ROA, ROA is the ratio of net income to total assets, firm size is measured as the natural logarithm of total assets, the market-to-book ratio is the ratio of market value of equity plus book value of total debt over total assets and asset maturity is measured as gross property, plant and equipment over total assets times gross property, plant and equipment over depreciation. Cross-listing is a dummy variable equal to one when the firm is cross-listed on either the LSE and/or trades as an ADR. Country characteristic variables are Development economy is a dummy variable equal to one when the country is classified as developed according to the World Bank

classification based on countries’ gross national income levels. Inflation rate is the annual rate of change in a country’s CPI. Corruption index is an index ranging from 0 to 10, with larger value indicating more severe corruption. Common law is a dummy variable equal to one when a country adopts the common law system. Dividend tax is a dummy variable equal to one when a country adopts a full dividend relief tax

system or full dividend imputation tax system. Deposits / GDP is the ratio of a country’s bank deposits to GDP. Life insurance penetration is the value of a country’s life insurance premiums to GDP.

Variables

Leverage ratio

Maturity structure ratio Tangible assets/Total assets ROA

Log total assets

Market-to-book ratio Asset maturity Cross-listing

Developed economy Inflation rate Corruption index Common law Dividend tax Deposits/GDP

Life insurance penetration N 271745 205175 271745 271745 271745 271745 205175 271745 624 624 624 624 624 624 624 Mean 0.29 0. 0.34 -0.08 19.68 2.40 33.50 0.51 0.84 0.03 3.01 0.58 0.13 0.93 0.05 Std Dev 0.27 0.34 0.24 0. 4.32 4.44 125.44 0.00 - 0.05 1.73 - - 0.57 0.03 Median Minimum Maximum 0.23 0.00 1.00 0.58 0.00 1.00 0.29 0.00 0.95 0.02 -4.58 0.30 19.78 -10.93 31.94 1.49 -12.68 29.31 8.79 0.01 1042.55 0.22 0.00 1.00 - 0.00 1.00 0.02 -0.04 0. 2.50 0.00 9.43 - 0.00 1.00 - 0.00 1.00 0.67 0.13 2.46 0.04 0.00 0.31

36

Table 3

Correlation matrix

The table provides correlation matrix for our sample. Pearson correlation coefficients for all independent variables, leverage and debt maturity, together with each pairing of independent variables are presented. Variables are as defined in Table 2.

[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]

Leverage ratio [1] 1.000 Long-term to total debt ratio [2] 0.067 1.000 Tangible assets/Total assets [3] 0.248 0.227 1.000 ROA [4] 0.033 0.110 0.091 1.000 Log total assets [5] 0.167 -0.038 0.035 0.269 1.000 Market-to-book ratio [6] -0.200 0.018 -0.110 0.065 -0.053 1.000 Asset maturity [7] 0.067 0.079 0.315 0.012 -0.107 -0.042 1.000 Cross-listing [8] -0.014 0.049 0.045 0.033 0.117 0.012 0.010 1.000 Developed economy [9] -0.069 0.163 -0.121 -0.081 -0.143 0.046 0.023 -0.037 1.000 Inflation rate [10] 0.020 0.001 0.056 0.018 -0.056 -0.002 -0.004 0.059 -0.334 1.000 Corruption index [11] 0.156 -0.212 0.063 0.086 0.303 -0.058 -0.050 -0.003 -0.756 0.329 1.000 Common law [12] -0.1 0.170 0.005 -0.139 -0.392 0.047 0.024 -0.019 -0.004 -0.011 -0.288 1.000 Dividend tax [13] -0.042 -0.026 0.008 0.020 -0.114 0.007 0.047 0.0 0.125 0.082 -0.100 -0.080 1.000 Deposits/GDP [14] 0.071 -0.135 -0.023 0.053 0.327 -0.042 0.074 0.024 0.285 -0.326 -0.144 -0.305 0.019 1.000 Life insurance penetration [15] 0.006 -0.063 -0.045 0.022 0.176 -0.017 0.062 0.001 0.217 -0.256 -0.206 -0.060 -0.211 0.461

37

Table 4

Leverage, firm and country level determinants

This table presents regressions of leverage on both firm and country level variables, as defined in Table 2. (Common law)*(High

corruption) interacts Common law dummy variable with High corruption dummy variable. High corruption is a dummy variable equal to one when the value of the corruption index in a given economy-year is greater or equal to the top quartile of the sample, and zero otherwise. All regressions include dummy variables for industry (two digit SIC codes). The sample is divided between developed and developing economies as defined by the developed economy dummy. The sample is split into two sub-samples, 1991-1998 and 1999-2006. This table also reports the adjusted R-squared and number of firm-year observations. Standard errors are robust to clustering within country over time. T-statistics are given in parentheses.

Dependent variable: Total debt/Market value of the firm Sample:

Independent variable:

Firm factors:

Tangible assets/Total assets

ROA

Log total assets

Market-to-book ratio

Country factors: Developed economy

Inflation rate

Corruption index

Common law

Dividend tax

Deposits/GDP

Life insurance penetration

Full Sample

Developed Economies

Developing Economies

1991-1998 (4) 0.1694 (18.50)*** -0.1771 (-3.53)*** 0.0076 (5.61)*** -0.0118 (-12.92)***

0.1004 (3.00)*** 0.0593 (0.69) 0.0163 (2.99)*** -0.05 (-6.10)*** -0.0421 (-3.)*** -0.0169 (-0.) -0.07 (-0.30) 0.1369 (3.42)*** 93408 0.1813

1999-2006 (5) 0.2444 (26.)*** -0.0159 (-3.05)*** 0.0071 (8.10)*** -0.0074 (-15.38)***

0.0838 (4.16)*** 0.0709 (1.04) 0.0205 (5.12)*** -0.0057 (-0.50) -0.0267 (-2.70)*** 0.0027 (0.27) -0.0246 (-0.20) 0.0276 (1.10) 178337 0.1913

(1) (2) (3) 0.2203 0.2242 0.1627 (29.24)*** (25.43)*** (14.41)*** -0.0228 -0.0172 -0.2799 (-3.88)*** (-3.55)*** (-8.)*** 0.0072 0.0067 0.0098 (9.)*** (8.70)*** (6.50)*** -0.0085 -0.0077 -0.0143 (-15.65)*** (-15.99)*** (-11.42)***

0.0809 (4.86)*** 0.02 -0.0993 0.2245 (0.55) (-0.42) (4.13)*** 0.0190 0.0212 0.0136 (5.79)*** (5.76)*** (2.53)*** -0.0223 -0.0301 0.0752 (-2.83)*** (-3.55)*** (3.84)*** -0.0287 -0.0270 -0.1157 (-4.05)*** (-3.74)*** (-6.08)*** -0.0016 -0.0069 -0.0196 (-0.19) (-0.70) (-1.00) -0.0012 -0.0201 -0.7981 (-0.01) (-0.17) (-3.47)*** 0.0615 (Common law)*(High corruption)

(2.82)*** Number of observations 271475 2276 44056 Adjusted R-square 0.1795 0.1914 0.1819

*,**, and ***, significant at the 10, 5 and 1 percent level, respectively.

38

Table 5

Leverage, firm and country level determinants with interaction effects

This table presents regressions of leverage on both firm and country level variables, as defined in Tables 2 and 4. (Common law)*(High corruption) interacts Common law dummy variable with High corruption dummy variable. (ROA)*(Dividend tax) interacts ROA with Dividend tax.

(ROA)*(Common law) interacts ROA with Common law dummy variable. (Tangible assets)*(Common law) interacts Tangible assets/Total assets with Common law. (Market-to-book)*(Common law) interacts Market-to-book ratio with Common law. (ROA)*(Low corruption) interacts ROA with Low corruption. Low corruption is 1 minus High corruption. (Tangible assets)*(Low corruption) interacts Tangible assets/Total assets with Low corruption. (Market-to-book)*(Low corruption) interacts Market-to-book ratio with Low corruption. (Cross-listing)*(High corruption) interacts Cross-listing dummy variable with High corruption dummy variable. All regressions include dummy variables for industry (two digit SIC codes). This table also reports the adjusted R-squared and number of firm-year observations. Standard errors are robust to clustering within country over time. T-statistics are given in parentheses.

Dependent variable: Total debt/Market value of the firm Independent variable: (1) (2) (3) (4) Firm factors: Tangible assets/Total assets 0.2203 0.2253 0.1939 0.2205 (29.22)*** (14.59)*** (7.)*** (29.29)*** ROA -0.0224 -0.0740 -0.0720 -0.0229 (-3.67)*** (-5.52)*** (-5.00)*** (-3.90)*** Log total assets 0.0072 0.0070 0.0070 0.0074 (9.86)*** (9.68)*** (9.88)*** (9.93)*** Market-to-book ratio -0.0085 -0.0150 -0.0212 -0.0084 (-15.70)*** (-19.19)*** (-26.81)*** (-15.72)*** Country factors: Developed economy 0.0811 0.0779 0.0730 0.0793 (4.86)*** (4.75)*** (4.61)*** (4.79)*** Inflation rate 0.0292 0.0287 0.0533 0.0371 (0.56) (0.57) (1.10) (0.71) Corruption index 0.0191 0.0181 0.0199 0.0186 (5.79)*** (5.61)*** (6.08)*** (5.68)*** Common law -0.0223 -0.0333 -0.0265 -0.0222 (-2.83)*** (-3.33)*** (-3.45)*** (-2.80)** Dividend tax -0.0290 -0.0275 -0.0295 -0.0271 (-4.07)*** (-3.97)*** (-4.28)*** (-3.88)*** Deposits/GDP -0.0016 -0.0016 -0.0029 -0.0015 (-0.19) (-0.19) (-0.35) (-0.17) Life insurance penetration -0.0017 -0.0094 -0.0504 -0.0003 (-0.02) (-0.09) (-0.52) (-0.01) (Common law)*(High corruption) 0.0615 0.0665 0.0794 0.0611 (2.81)*** (3.08)*** (3.61)*** (2.80)*** Firm/country interactions: Cross-listing -0.0232 (-5.63)*** (Cross-listing)*(High corruption) -0.0001 (-0.02) (ROA)*(Dividend tax) -0.0036 (-0.53) (ROA)*(Common law) 0.0325 (5.47)*** (Tangible assets)*(Common law) -0.0122 (-0.74) (Market-to-book)*(Common law) 0.0056 (11.23)*** (ROA)*(Low corruption) 0.0309 (5.08)*** (Tangible assets)*(Low corruption) 0.0235 (0.84) (Market-to-book)*(Low corruption) 0.0092 (22.95)*** Number of observations 271475 271475 271475 271475 Adjusted R-square 0.1795 0.1883 0.1919 0.1799

*,**, and ***, significant at the 10, 5 and 1 percent level, respectively. 39

Table 6

Debt maturity structure, firm and country level determinants

This table presents regressions of debt maturity on both firm and country level variables, as defined in Tables 2 and 4. All regressions include dummy variables for industry (two digit SIC codes). The sample is dividend between developed and developing economies as

defined by the developed economy dummy. This table also reports the adjusted R-squared and number of firm-year observations. Standard errors are robust to clustering within country over time. T-statistics are given in parentheses. Dependent variable: Long-term debt/Total debt Sample:

Independent variable:

Firm factors:

Tangible assets/Total assets ROA

Log total assets

Market-to-book ratio

Asset maturity

Country factors: Developed economy

Inflation rate

Corruption index

Common law

Deposits/GDP

Life insurance penetration

Full Sample

(1) 0.2510 (30.71)*** 0.0723 (15.70)*** 0.0117 (16.24)*** 0.0007 (2.15)** -0.0001 (-0.34) 0.1823 (7.69)*** 0.3335 (4.51)*** -0.0293 (-6.57)*** 0.0836 (6.37)*** -0.1087 (-11.58)*** -0.1524 (-0.73) 0.1185 (4.88)*** 205175 0.19

Developed Economies

(2) 0.23 (27.12)*** 0.0747 (15.18)*** 0.0111 (15.30)*** 0.0005 (1.60) -0.0001 (-0.10) -0.4743 (-1.17) -0.0297 (-5.63)*** 0.0877 (5.94)*** -0.1214 (-9.47)*** -0.2505 (-0.93) 167108 0.1787

Developing Economies

(3) 0.3441 (24.37)*** 0.1112 (6.26)*** 0.0152 (8.14)*** 0.0016 (1.39) 0.0001 (1.04) 0.3328 (4.20)*** -0.0081 (-1.12) 0.1867 (6.78)*** -0.1660 (-5.29)*** -0.5695 (-1.59) 38067 0.1777

1991-1998

(4) 0.2413 (27.80)*** 0.0563 (4.79)*** 0.0092 (10.79)*** 0.0005 (0.76) -0.0001 (-1.36) 0.0977 (3.11)*** 0.0304 (0.30) -0.0294 (-5.50)*** 0.0848 (3.92)*** -0.0812 (-4.35)*** -0.3247 (-0.74) 0.0483 (1.30) 738 0.1757

1999-2006

(5) 0.2622 (26.00)*** 0.0728 (13.95)*** 0.0121 (12.06)*** 0.0007 (1.77)* 0.0001 (0.21) 0.1926 (6.25)*** 0.4098 (3.60)*** -0.0301 (-4.88)*** 0.0852 (5.28)*** -0.1149 (-10.43)*** -0.1528 (-0.67) 0.1383 (4.74)*** 131286 0.1981

(Common law)*(High corruption)

Number of observations Adjusted R-square

*,**, and ***, significant at the 10, 5 and 1 percent level, respectively.

40

Table 7

Debt maturity structure, firm and country level determinants with interaction effects

This table presents regressions of debt maturity on both firm and country level variables, as defined in Tables 2, 4 and 5. All regressions include dummy variables for industry (two digit SIC codes). This table also reports the adjusted R-squared and number of firm-year observations. Standard errors are robust to clustering within country over time. T-statistics are given in parentheses.

Dependent variable: Long-term debt/Total debt Independent variable:

Firm factors:

Tangible assets/Total assets

ROA

Log total assets

Market-to-book ratio

Asset maturity

Country factors: Developed economy

Inflation rate

Corruption index

Common law

Deposits/GDP

Life insurance penetration

(Common law)*(High corruption)

Firm/country interactions: Cross-listing

(Cross-listing)*(High corruption)

(Tangible assets)*(Common law)

(Market-to-book)*(Common law)

(Tangible assets)*(Low corruption)

(Market-to-book)*(Low corruption)

Number of observations Adjusted R-square

(1) 0.2831 (22.96)*** 0.0730 (15.84)*** 0.0117 (16.26)*** 0.0017 (3.04)*** 0.0000 (0.02) 0.1834 (7.72)*** 0.3291 (4.48)*** -0.0292 (-6.)*** 0.1047 (6.51)*** -0.1080 (-11.53)*** -0.1533 (-0.73) 0.1206 (4.99)***

-0.0531 (-3.77)*** -0.0009 (-2.40)**

205175 0.1992 (2) 0.2830 (10.05)*** 0.0729 (15.82)*** 0.0116 (16.38)*** 0.0011 (1.15) 0.0000 (-0.43) 0.1842 (8.00)*** 0.3192 (4.15)*** -0.0306 (-6.63)*** 0.0855 (6.47)*** -0.1075 (-11.49)*** -0.1570 (-0.77) 0.1106 (4.26)***

-0.0361 (-1.14) -0.0002 (-0.41) 205175 0.1994 (3) 0.2497 (30.76)*** 0.0727 (15.70)*** 0.0114 (15.43)*** 0.0007 (2.07)** -0.0001 (-0.40) 0.1886 (8.04)*** 0.2852 (4.11)*** -0.0287 (-6.47)*** 0.0855 (6.40)*** -0.1086 (-11.53)*** -0.1659 (-0.80) 0.1193 (4.90)*** 0.0301 (4.35)*** 0.1118 (7.67)***

205175 0.2007

*,**, and ***, significant at the 10, 5 and 1 percent level, respectively.

41

Table 8

Sources of Variation in Leverage and Debt Maturity

This table presents regressions of both leverage and debt maturity. Column (1) reports a fixed effects model for leverage. Column (3) reports a cross-sectional regression for leverage using the averages within each firm of the dependent and firm level explanatory variables and within country averages of country level explanatory variables. Columns (2) and (4) report the corresponding results for maturity structure. All variables are as defined in Tables 2 and 4. Industry dummy variables (two digit SIC codes) are included in Columns (3) and (4). This table also reports the adjusted R-squared and number of firm-year observations. T-statistics are given in parentheses.

Dependent variable: Independent variable:

Tangible assets/Total assets

ROA

Log total assets

Market-to-book ratio

Asset maturity

Developed economy

Inflation rate

Corruption index

Common law

Dividend tax

Deposits/GDP

Life insurance penetration

(Common law)*(High corruption)

Fixed effects

Total debt/Market value of the firm

(1) 0.1675 (7.85)*** -0.0249 (-4.98)*** 0.0053 (4.08)*** -0.0037 (-4.67)***

-0.0356 (-1.28) -0.0015 (-0.02) 0.0181 (1.60) 0.0524 (2.58)*** -0.0169 (-0.79) 0.0781 (1.62) 0.0214 (0.18) -0.0515 (-1.02) 271745 0.1235

Long-term debt/Total debt

(2) 0.0933 (4.40)*** 0.0287 (7.58)*** 0.0033 (2.76)*** 0.0007 (2.69)*** 0.0000 (1.41) 0.0940 (3.11)*** -0.0135 (-0.26) -0.0192 (-4.01)*** 0.0698 (2.26)**

-0.0350 (-1.51) 0.1134 (1.72)* 0.0666 (2.90)***

205175 0.1145

Number of observations Adjusted R-square

Cross-sectional Total debt/Market Long-term value of the firm debt/Total debt

(3) (4) 0.2221 0.28 (39.42)*** (35.41)*** -0.0095 0.0790 (-4.)*** (25.63)*** 0.0082 0.0147 (20.22)*** (27.38)*** -0.0110 0.0007 (-36.44)*** (1.66)*

0.0000 (-2.56)*** 0.1019 0.1865 (19.99)*** (27.21)*** 0.1417 0.5119 (4.20)*** (11.91)*** 0.0178 -0.0349 (13.91)*** (-20.50)*** -0.0066 0.0678 (-2.14)** (16.16)*** -0.0342 (-9.65)*** -0.0126 -0.1155 (-4.)*** (-32.75)*** -0.2066 -0.3570 (-4.76)*** (-6.51)*** 0.0357 0.1433 (4.85)*** (15.17)***

271745 205175 0.1740 0.1966

*,**, and ***, significant at the 10, 5 and 1 percent level, respectively.

42

Appendix 1

Definitions and data sources of country level variables

Variable

Developed economy

Description

A zero or one dummy variable indicating whether the country is classified as developed according to the World Bank classification based on countries’ gross national income levels

Annual rate of change on Consumer Price Index

Source

World Development Indicators, World Bank

Inflation rate Corruption index Common law Dividend tax Deposits / GDP Life insurance penetration

World Development Indicators, World Bank

An index ranges from 0 to 10, with larger value indicating Corruption Perception more severe corruption Index, Transparency

International

An zero or one dummy variable indicating whether a Treisman [2000] country adopts the common law system

A zero or one dummy variable indicating whether a Price Waterhouse Coopers country adopts a full dividend relief tax system or full dividend imputation tax system

A proxy for the degree of financial intermediation of a International Financial country, measure as the country’s deposits (liquid liability) Statistics, International over GDP Monetary Fund Value of life insurance premiums/GDP International Financial

Statistics, International Monetary Fund

43

Appendix 2

Median values of country level dependent variables

The table provides the median value of country level the dependent variables, classified by country. Variables are as defined in Table 2 and Appendix 1

Country Developed economy

Inflation rate

Corruption index

Common law

Dividend tax

Deposits / GDP

Life insurance penetration

Australia 1.00 0.03 1.40 1.00 1.00 0.60 0.04 Austria 1.00 0.02 2.35 0.00 0.00 0.82 0.02 Belgium 1.00 0.02 2.90 0.00 0.00 0.78 0.04 Brazil 0.00 0.08 6.35 0.00 0.00 0.37 0.00 Canada 1.00 0.02 1.03 1.00 0.00 0.75 0.03 Switzerland 1.00 0.01 1.10 0.00 0.00 1.25 0.07 Chile 0.00 0.05 2.85 0.00 0.00 0.43 0.02 China 0.00 0.02 6.60 0.00 0.00 0.33 0.01 Germany 1.00 0.02 1.94 0.00 1.00 0.70 0.03 Denmark 1.00 0.02 0.50 0.00 0.00 0.53 0.04 Spain 1.00 0.03 3.35 0.00 0.00 0.65 0.02 Finland 1.00 0.01 0.40 0.00 0.00 0.48 0.07 France

1.00 0.02 3.00 0.00 0.00 0.63 0.06 United Kingdom 1.00 0.03 1.52 1.00 0.00 0.90 0.09 Greece

1.00 0.04 5.25 0.00 1.00 0.55 0.01 , 1.00 0.02 2.25 1.00 1.00 1.92 0.03 Indonesia 0.00 0.09 8.10 0.00 0.00 0.39 0.01 India 0.00 0.06 7.20 1.00 0.00 0.38 0.01 Ireland 1.00 0.03 2.50 1.00 0.00 0.70 0.07 Israel 1.00 0.04 2.95 1.00 0.00 0.74 0.03 Italy 1.00 0.03 5.25 0.00 1.00 0.53 0.03 Japan

1.00 0.00 2.90 0.00 0.00 1.94 0.09 Korea, Rep. 1.00 0.04 5.76 0.00 0.00 0.49 0.09 Mexico 0.00 0.10 6.70 0.00 0.00 0.23 0.01 Malaysia 0.00 0.03 4.90 1.00 0.00 1.12 0.02 Netherlands 1.00 0.02 1.05 0.00 0.00 0.99 0.05 Norway

1.00 0.02 1.25 0.00 1.00 0.50 0.02 New Zealand 1.00 0.02 0.60 1.00 1.00 0.82 0.02 Pakistan 0.00 0.08 7.80 1.00 0.00 0.30 0.00 Peru

0.00 0.04 5.90 0.00 0.00 0.21 0.00 Philippines 0.00 0.07 7.40 0.00 0.00 0.47 0.01 Portugal 1.00 0.03 3.65 0.00 0.00 0.88 0.02 Singapore 1.00 0.02 0.82 1.00 0.00 0.98 0.03 Sweden 1.00 0.02 0.80 0.00 0.00 0.40 0.04 Thailand 0.00 0.04 6.80 1.00 0.00 0.94 0.01 Turkey 0.00 0. 6.40 0.00 1.00 0.32 0.00 Taiwan

1.00 0.01 4.45 0.00 0.00 0.24 0.04 United States 1.00 0.03 2.40 1.00 0.00 0.65 0.04 South Africa

0.00 0.07 5.30 1.00 0.00 0.50 0.11

44

Appendix 3

Pooled firm-level regressions of leverage by country

The table presents the regression of leverage on firm level variables as defined in Table 2. The regression equation is estimated for each country using the pooled time-series and cross-sectional sample. All regressions include dummy variables for industry (two digit SIC codes). This table also reports the adjusted R-squared and number of firm-year observations. Standard errors are robust to clustering within firm over time. T-statistics are given in parentheses.

Country Australia Austria Belgium Brazil Canada Switzerland Chile China Germany Denmark Spain Finland France United Kingdom Greece Indonesia India Ireland Israel Italy Japan Korea Mexico Malaysia Netherlands Norway New Zealand Tangible assets/ Total Assets 0.0826 (4.95)*** 0.2587 (2.27)** 0.0749 (1.19) 0.2181 (3.29)*** 0.1608 (8.67)*** 0.2123 (4.17)*** 0.1433 (2.69)*** 0.0659 (2.68)*** 0.2486 (6.59)*** 0.1932 (2.95)*** 0.0592 (1.12) 0.3539 (5.20)*** 0.3043 (7.34)*** 0.1692 (10.58)*** 0.0498 (0.98) 0.1853 (5.82)*** 0.2381 (4.06)*** 0.4287 (10.63)*** 0.1197 (1.48) 0.6199 (8.14)*** 0.0078 (0.08) 0.3949 (16.09)*** 0.1400 (3.)*** 0.0693 (0.94) 0.1344 (4.15)*** 0.2024 (4.27)*** 0.4221 (9.62)*** 0.0778 (0.98) ROA -0.0120 (-2.49)** -0.0495 (-1.10) -0.5686 (-4.93)*** -0.1691 (-3.19)*** -0.0127 (-1.79)* -0.1102 (-2.32)** -0.1430 (-0.87) -0.1416 (-6.13)*** -0.0766 (-3.93)*** -0.1049 (-2.15)** -0.6820 (-6.69)*** -0.2751 (-2.87)*** -0.2121 (-4.00)*** -0.0296 (-4.93)*** -1.1313 (-8.15)*** -0.0694 (-4.98)*** -0.4838 (-5.49)*** -1.2238 (-12.97)*** -0.0197 (-1.40) -0.0375 (-1.24) -0.5614 (-2.93)*** -0.6184 (-6.61)*** -0.1919 (-5.07)*** -0.7191 (-3.88)*** -0.3425 (-6.48)*** -0.1272 (-5.56)*** -0.0353 (-1.53) -0.0136 (-0.91) Market-to-book Log total assets ratio 0.0200 -0.0061 (8.53)*** (-8.08)*** 0.0065 -0.0113 (1.72)* (-4.10)*** 0.0124 -0.0100 (4.46)*** (-3.45)*** 0.0033 -0.0093 (1.04) (-3.33)*** 0.0131 -0.0080 (4.61)*** (-12.76)*** 0.0056 -0.0158 (1.55) (-5.63)*** 0.0308 -0.0085 (2.78)*** (-2.34)*** 0.0404 -0.0072 (4.22)*** (-6.32)*** 0.0081 -0.0111 (6.86)*** (-8.99)*** 0.0023 -0.0177 (0.76) (-5.61)*** 0.0045 -0.0117 (3.20)*** (-3.39)*** 0.0031 -0.0104 (0.98) (-3.46)*** 0.00 -0.0119 (3.78)*** (-11.62)*** 0.0038 -0.0058 (4.38)*** (-14.22)*** 0.0228 -0.0068 (3.34)*** (-6.75)*** 0.0043 -0.0116 (3.16)*** (-10.16)*** 0.0100 -0.0118 (2.)*** (-4.12)*** -0.0003 -0.0171 (-0.07) (-8.34)*** 0.0058 -0.0080 (1.21) (-3.63)*** 0.02 -0.0059 (2.56)** (-2.55)** 0.0138 -0.0168 (3.52)*** (-5.51)*** 0.0115 -0.0046 (6.74)*** (-5.09)*** 0.0578 -0.0172 (7.)*** (-5.88)*** 0.0153 -0.0292 (1.) (-3.27)*** 0.0082 -0.0108 (4.69)*** (-6.50)*** 0.0094 -0.0072 (4.01)*** (-4.95)*** 0.0029 -0.0121 (1.03) (-5.23)*** 0.0121 -0.0141 (2.12)** (-2.78)*** No of observations/

Adjusted R-square 8308 0.2426 1144 0.2031 1485 0.2865 2591 0.1798 10988 0.2979 2656 0.2882 1424 0.2033 6827 0.13 9209 0.1876 2123 0.2318 2315 0.2038 1684 0.3961 96 0.2270 21785 0.2486 2511 0.2699 7108 0.1627 2573 0.2337 4388 0.4806 880 0.3133 949 0.4948 2810 0.14 42611 0.1624 6741 0.2345 1230 0.3663 7586 0.1388 2612

0.3316

1826 0.558 9 0.3148

45

Pakistan 0.3031 -1.4441 0.0061 -0.0112 1061 (2.73)*** (-9.53)*** (0.77) (-2.44)** 0.5333 Peru 0.0470 -1.2808 0.0347 -0.0134 491 (0.39) (-8.05)*** (2.27)** (-1.15) 0.3557 Philippines 0.1009 -0.1348 0.0174 -0.0203 18 (1.53) (-3.)*** (4.74)*** (-4.59)*** 0.26 Portugal -0.1427 -0.6142 0.0019 -0.0268 867 (-1.76)* (-2.31)** (0.30) (-4.18)*** 0.3338 Singapore 0.2465 -0.0920 0.0105 -0.0149 4111 (7.03)*** (-2.82)*** (3.56)*** (-5.65)*** 0.1862 Sweden 0.3220 -0.0003 0.0041 -0.0144 3394 (4.48)*** (-0.03) (2.57)*** (-7.95)*** 0.5070 Thailand 0.10 -0.52 0.0267 -0.0108 3457 (2.34)** (-3.23)*** (3.56)*** (-3.33)*** 0.1925 Turkey -0.0211 -0.4782 0.0144 -0.0049 1422 (-0.35) (-5.08)*** (3.05)*** (-2.31)** 0.2990 Taiwan 0.0482 -0.7784 0.0361 -0.0369 7051 (1.80)* (-10.41)*** (9.70)*** (-9.05)*** 0.3500 United States 0.2016 -0.0019 0.0019 -0.0065 77909 (18.95)*** (-1.00) (3.24)*** (-35.10)*** 0.1600 South Africa 0.09 -0.0625 -0.0042 -0.0039 3699 (2.22)** (-2.66)*** (-1.87)* (-2.08)** 0.1249

*,**, and ***, significant at the 10, 5 and 1 percent level, respectively.

46

Appendix 4

Pooled firm-level regressions of debt maturity structure by country

The table presents the regression of debt maturity on firm level variables as defined in Table 2. The regression equation is estimated for each country using the pooled time-series and cross-sectional sample. All regressions include dummy variables for industry (two digit SIC codes). This table also reports the adjusted R-squared and number of firm-year observations. Standard errors are robust to clustering within firm over time. T-statistics are given in parentheses.

Tangible assets/ Market-to-book No of observations/

Country Total Assets ROA Log total assets ratio Asset Maturity Adjusted R-square Australia Austria Belgium Brazil Canada Switzerland Chile China Germany Denmark Spain Finland France United Kingdom Greece Indonesia India Ireland Israel Italy Japan Korea Mexico Malaysia Netherlands Norway New Zealand 0.20 (6.62)*** 0.1239 (1.05) 0.4787 (5.21)*** 0.2082 (3.90)*** 0.1972 (6.31)*** 0.3256 (5.12)*** 0.2811 (2.62)** 0.3748 (7.11)*** 0.3404 (7.47)*** 0.3728 (4.82)*** 0.2321 (3.99)*** 0.1193 (1.33) 0.2609 (4.84)*** 0.2602 (11.40)*** 0.3940 (5.32)*** 0.2206 (5.52)*** 0.3923 (6.13)*** 0.4123 (8.87)*** 0.2166 (2.35)** 0.2995 (2.57)** 0.1001 (1.20) 0.2983 (14.84)*** 0.1700 (4.69)*** 0.4925 (5.03)*** 0.2186 (5.41)*** 0.3267 (4.57)*** 0.2753 (5.56)*** 0.3410 (3.92)*** 0.0425 0.0358 (2.88)*** (7.04)*** -0.0352 0.0139 (-0.73) (2.79)*** 0.2277 0.0077 (1.62) (1.52) 0.1485 0.0119 (3.22)*** (3.53)*** 0.0309 0.0315 (2.03)** (7.19)*** 0.0316 -0.0029 (0.39) (-0.73) -0.3479 0.0273 (-2.00)** (1.72)* 0.0557 0.0329 (2.31)** (1.95)* 0.0282 0.0024 (0.90) (1.41) -0.0630 -0.0032 (-2.38)** (-0.85) -0.2207 0.0046 (-1.58) (2.11)** -0.1396 0.0014 (-2.73)*** (0.45) 0.0126 0.0093 (0.35) (5.22)*** 0.0303 0.0168 (2.72)*** (11.27)*** 0.26 0.0162 (1.59) (1.46) 0.0239 0.0073 (1.41) (3.74)*** 0.20 0.0111 (5.71)*** (2.36)** 0.1201 0.0206 (1.34) (3.74)*** 0.0223 0.0204 (0.50) (2.40)** -0.0304 0.0222 (-0.33) (2.12)** 0.1734 0.0060 (2.10)** (1.47) 0.10586 0.0133 (3.58)*** (7.91)*** -0.0662 0.0216 (-2.16)** (4.47)*** 0.30 0.0393 (3.)*** (2.24)** 0.1613 0.0048 (6.23)*** (2.26)** -0.0066 0.0253 (-0.09) (4.47)*** 0.1138 0.0066 (4.41)*** (2.30)** 0.0493 0.0077 (1.09) (1.24) 0.0003 -0.0001 (0.21) (-1.16) -0.0073 0.00004 (-1.46) (0.30) -0.0016 -0.0001 (-0.65) (-1.38) 0.0034 -0.0002 (1.49) (-1.90)* -0.0012 -0.0001 (-0.91) (-1.35) -0.0038 -0.0002 (-1.00) (-2.56)** 0.0081 -0.0002 (1.90)* (-1.45) -0.0004 0.0005 (-0.25) (1.53) 0.0029 -0.0001 (1.60) (-0.08) 0.00004 -0.0001 (0.01) (-0.48) 0.0043 -0.0001 (1.39) (-0.53) -0.0050 -0.00001 (-1.38) (-0.18) 0.0002 0.00001 (0.13) (0.25) 0.0002 0.0001 (0.35) (2.24)** -0.0024 -0.0003 (-1.20) (-3.31)*** 0.0027 0.0001 (1.75)* (1.32) 0.0020 -0.0003 (0.77) (-2.94)*** 0.0012 -0.0002 (0.55) (-1.29) 0.0023 0.0003 (0.65) (3.99)*** 0.0011 -0.0004 (0.28) (-1.23) -0.0011 0.0003 (-0.34) (1.12) 0.0031 0.0001 (3.69)*** (0.04) 0.0057 -0.0001 (2.38)** (-1.44) 0.0024 -0.0007 (0.29) (-2.23)** 0.0016 -0.00004 (0.78) (-0.34) 0.0032 0.0001 (1.32) (0.97) -0.0036 -0.00001 (-1.06) (-0.15) 0.0036 -0.0001 (0.41) (-0.92) 6102 0.1432 865 0.2061 1121 0.2421 19 0.2476 8239 0.1566 2065 0.1676 1137 0.2815 6339 0.2443 7021 0.1166 1766 0.1992 1969 0.22 1191 0.0915 8018 0.0916 18588 0.1745 1798 0.1946 6281 0.1947 2353 0.1742 4074 0.1530 766 0.2826 674 0.2156 2270 0.1080 34424 0.09 5349 0.0873 1060 0.2121 6773 0.1219 2102 0.2042 1375 0.2690 763 0.2816

47

Pakistan 0.4011 -0.3071 0.0025 (3.43)*** (-1.88)* (0.32) Peru 0.3417 -0.7551 0.0822 (2.87)*** (-3.52)*** (4.81)*** Philippines 0.1400 0.1056 0.0185 (1.49) (2.06)** (3.27)*** Portugal 0.1117 0.0970 0.0236 (0.94) (0.62) (2.79)*** Singapore 0.4018 0.0819 0.0096 (8.79)*** (2.87)*** (2.49)** Sweden 0.0968 -0.0121 -0.0045 (1.42) (-0.38) (-2.02)** Thailand 0.2927 -0.0431 0.0290 (5.59)*** (-1.28) (3.69)*** Turkey 0.2032 -0.0837 0.0200 (2.37)** (-1.48) (2.51)** Taiwan 0.4055 0.11 0.0487 (11.04)*** (2.87)*** (10.34)*** United States 0.2583 0.0973 0.0155 (19.42)*** (29.38)*** (21.38)*** South Africa 0.3160 -0.0222 0.0046 (4.95)*** (-0.) (1.65)* *,**, and ***, significant at the 10, 5 and 1 percent level, respectively.

0.0069 (2.05)** 0.0206 (4.83)*** 0.0051 (0.94) -0.0102 (-1.37) -0.00001 (-0.00) -0.0033 (-0.79) 0.0068 (2.95)*** 0.0068 (2.12)** 0.0132 (3.12)*** -0.0014 (-3.96)*** 0.0050 (1.82)* 0.00001 (0.15) -0.0001 (-0.65) -0.0002 (-1.45) 0.0019 (1.15) -0.0001 (-1.27) 0.00004 (0.86) -0.00003 (-0.12) -0.00002 (-0.16) -0.0004 (-2.28)** -0.0001 (-2.04)** -0.0001 (-0.09) 907 0.3343 360 0.27 1283 0.1681 705 0.2193 3810 0.2392 2529 0.0687 3245 0.2188 11 0.0929 50 0.1552 46303 0.1977 2279 0.2049

48