R.H.F.M. Schonbrodt. Susteren. Graduation date

Transcription

1 Capital Structure and Firm Performance in Germany and the United States Master thesis Department Accountancy, Faculty of Economics and Business Administration, Tilburg University. By R.H.F.M. Schonbrodt Susteren Graduation date

2 Capital Structure and Firm Performance in Germany and the United States Master thesis Accounting Tilburg School of Economics and Management Author: R.H.F.M. Schonbrodt ANR: Supervisor: Prof.dr.mr. P.M. van der Zanden RA Second Reader: Dr. B.R.C.J. van den Brand

3 Preface This master thesis is the written result of my research to complete the master study Accounting at Tilburg University. In this research I investigated the relation between financial leverage (capital structure) and firm performance in different law families. By reading relating, prior research you get the suggestion that the common law system is a superior system over the civil law system with respect to financial development and economic growth. Since all Western European countries, except the UK, do have a civil law system, the superiority assumption surprised me. After discussing this observation with my supervisor, dr.mr. P.M. van der Zanden RA, we came to the conclusion that it could be an interesting study to investigate the superiority of a superior law system over another. Besides it should be interesting to investigate whether there could be other reasons that explain differences in capital structure between countries. I want to thank my supervisor who helped me to complete this thesis. By sharing his knowledge and supporting me during the entire project, I was able to accomplish this thesis resulting in remarkable and interesting results. Finally I want to thank my family, special thanks to my father and mother; my employer, who made it possible for me to work on my thesis whenever I wanted to, and friends for supporting me during my whole study and made it an enjoyable time. Roel Schonbrodt Susteren, December 2011 Master Thesis Accounting R.H.F.M. Schonbrodt 3

4 Table of Contents Preface... 3 Abstract... 5 Introduction... 6 Discussion of literature and development of the research questions... 8 Data...12 Results...18 Methodology Results: running the regression models Sub-sample test of the liability side of German and US firms Conclusions...26 References...28 Appendix A ANOVA table German firms: Influence of Leverage over Firm Performance ANOVA table US firms: Influence of Leverage over Firm Performance Appendix A ANOVA table German firms: Influence of Adjusted Leverage over Firm Performance ANOVA table US firms: Influence of Adjusted Leverage over Firm Performance Appendix B Sub-Sample information: including US and German firms Master Thesis Accounting R.H.F.M. Schonbrodt 4

5 Abstract This study investigates the influence of firm specific characteristics over firm performance, by analyzing a sample of US and German firms. The study especially focuses on the way how German firms are financed with respect to their US peers. Since prior research assumes better creditor rights in common law countries over civil law countries, they assume that firms in common law countries are more attracted to finance investments with external debt instead of equity. Overall, the such called theory of law and finance even assumes that the common law system is superior over the civil law system with respect to financial development and economic growth. This study shows that there is no reason to assume that creditor rights are better protected in common law countries and that there is no reason to assume that the common law system is superior with respect to civil law system. This study even suggests that external debt is more positively related with firm performance in German firms than it is with firm performance in US firms. However those data are not significant. The suggestion, made in this study, that the differences in capital structure between countries is due to different financial availabilities can not be proven either, unless the results still indicate that this could be a possible reason. The study emphasizes that the wide respected theoretical underpinnings of the observed relation between firm performance and capital structure is still largely unresolved. Master Thesis Accounting R.H.F.M. Schonbrodt 5

6 Introduction Since the late fifties of the last century, several theories have developed which aim to explain the capital structure decisions of financial managers (Brounen, Jong, de, & Koedijk, 2005). The first papers mainly focused on what is called Trade-off problems. Problems discussed in those papers focused on the influence of tax reductions over capital structure, agency-cost problems and influences of bankruptcy costs over capital structure. Indirectly, those theories are based on the assumption that different principal-agent objectives, inadequate motivation and incomplete contracts are sources of extra costs, which lead to a decrease in firm performance (Leibenstein, 1978). In this way financial leverage can be seen as a disciplinary tool for firm managers, to protect the firm owners, to reduce managerial cash flow waste through the threat of liquidation (Grossman & Hart, 1982) or through pressure to generate cash flows to service debt (Jensen M., 1986). In other words; more leverage overall lowers agency costs, which will lead to better firm performance. Later on the such called pecking-order theory developed. This theory suggests that managers have a picking order in which retained earnings represent the first choice, followed by debt financing and finally equity to meet financial objectives (Myers, 1984). In the last decade researchers mainly focused on variables which could have some influence on the financial leverage position of firms all over the world. A leading study in examining variables, is done by La Porte et al. This study suggests that the legal system of common law countries (countries such as the United States and the United Kingdom) is superior over the system of civil law countries (continental Europe as France and Germany) with respect to creditor and shareholder protection (La Porte, Lopez-de-Silanes, Shleifer, & Vishny, 1998). In other words the study by La Porte et al. argues that the common law system provides a better framework for financial developments and economic growth. This paper will not about the fact that different legal systems will, or can have some influence on the capital structure of private firms. However, it will try to disapprove the explanation that common law is by fact superior over civil law. By using linear regressions of firm data from Germany (civil law) and the US (common law) the paper wants to prove that the differences in relation of leverage on firm performance between common law countries and civil law countries are not as huge as suggested by the shareholder- and creditor-protection theories. The study proves that the explanation by creditorprotection theory does not hold for this study, where the results still show different current- and noncurrent liability ratios. However, unless the results indicate that due to another way of financing there are different financial possibilities between German firms and US firms, those results are not significant. As already mentioned, a lot of researchers used the creditor-protection variable by La Porte s theory of law and finance without discussing the truth of the study. As exception of all those researchers, Michael Graff (2008) did investigate the work of La Porte et al.. Although he found some supporting evidence for the assumption that financial investors are threaded differently across legal families, there is not much evidence that common law countries protect financial investors better than civil law Master Thesis Accounting R.H.F.M. Schonbrodt 6

7 countries by fact. As showed by Graff the aggregated data, especially with respect to the civil law countries, are inconsistent. Those inconsistent data, as he argues, make a deviation in just two investor protectiveness not very useful for revealing insights into the law and finance nexus. This raises the question that there could be something else than just the legal family explanation of La Porte et al., which leads to a different capital structure between common law and civil law firms. Saying so, the main question is whether economic growth and financial development is a product of common law and civil law, in which common law is suggested to be superior as argued in the law and finance theory by La Porte et al. (1998) on the one hand; or that common as well as civil law is shaped in such a way to support financial development and economic growth in the best way, taking account of all kind of specific national traditions and beliefs founded in the past on the other hand. In this instance it is particular relevant to investigate the availability of financial resources across different countries and financial law families. Instead of a shocking theory that common law is superior over civil law and assuming that economic growth and financial development is a product of those laws, it is imaginable that it are just those different financial resources which lead to a different capital structure which also leads to other laws and institutions over different countries. The remainder of this thesis is organized as follows. In section 2 previous literature will be outlined and research questions are established. Section 3 describes the data. Empirical results are discussed in section 4, including a robustness test to investigate different financial availabilities. Finally section 5 presents concluding remarks. The thesis is completed with appendixes showing detailed results of the linear regression models and detailed sample information. Master Thesis Accounting R.H.F.M. Schonbrodt 7

8 Discussion of literature and development of the research questions For over fifty years now, researchers are studying about the financial structure of individual firms. Ever since, they tried to develop a technical framework which aims to explain the capital structure decisions of financial managers (Brounen, Jong, de, & Koedijk, 2005). One of the first studies in this kind was done by Miller & Modigliani (1958, 1963). Miller & Modigliani first considered the simple definition of cost of capital in a way of profit maximization and market value maximization (Modigliani & Miller, The cost of capital, corporation finance and the theory of investment, 1958). A few years later they corrected their first study by adding a chapter of tax influences over the optimal capital structure of individual firms (Modiliani & Miller, 1963). This is a typical example of development of traditional such called Trade-off theory studies, done in those early years. The Trade-off theory refers to the theoretical problem of an optimal capital structure. In those studies the savings of tax deductions, due to the use of cheaper but more riskier external debt over equity, will be compared with bankruptcy costs and agency costs, for the use of that external debt. In other words, Those studies predict that firms maintain an optimum capital structure where the marginal benefit of debt equals the marginal cost (Cotei & Farhat, 2009). Studies about bankruptcy costs consider issues surrounding the role of bankruptcy costs, risk of losing values, in models of capital structure (Warner, 1977) and models to predict bankruptcies of which the Z-score model is probably the most famous (Altman, 1968). The agency cost theory is based on the idea that the interest of company s managers and company s owners are not perfectly aligned (Jensen & Meckling, 1976). In the 1980 s the such called pecking-order theory developed. The pecking order theory is based on the assumption of asymmetric information between managers and investors, which in fact is an extension of the agency-cost theory. The pecking-order theory assumes that managers know more about the true value of the firm than investors do. When managers need new money to finance new projects, they will try to use securities which are not undervalued by the market such as internal funds or riskless debt. In this way the choice between internal and external finance is affected. Saying so, the pecking-order theory suggests that firm s leverage is not driven by the Trade-off theory, but is just a result for trying to mitigate information asymmetry (Myers, 1984). In recent years several studies are done, comparing the Trade-off theory with the pecking-order theory. However, until now there is not a compelling answer about which theory is superior over the other. Fama & French (2002) for example found evidence in favor and against the pecking-order theory as well as for the Trade-off theory. On the other hand Frank & Goyal (2003) found evidence against the pecking-order theory with regard to small firms, while Lemon & Zender (2002) found no supporting evidence for the Trade-off theory. In the mid-1990 s a study by Rajan & Zingales (1995) about capital structure concluded that firm leverage is fairly similar across the G-7 countries. However, they also mentioned that a deeper examination of the US and other G-7 countries showed some evidence which suggests that the theoretical underpinnings of the observed conclusion is still largely unresolved (Rajan & Zingales, Master Thesis Accounting R.H.F.M. Schonbrodt 8

9 1995). Another remarkable, recent study with regard to the optimal capital structure of private firms was done by La Porte et al. (1998). Without focusing on the question whether the Trade-off theory is superior over the pecking-order theory or visa verse, La Porte et al. concentrated on legal rules and institutions in different countries. In his study he concluded that in a sample of individual firms out of 49 different countries, common-law countries have the strongest and French-law countries have the weakest legal protections of investors (La Porte, Lopez-de-Silanes, Shleifer, & Vishny, 1998). It is quite surprising that many researchers after La Porte et al. just used those outcomes as facts, without discussing the meaning of those results. In other words the studies by La Porte et al. argue that the common law system provides a better framework for financial developments and economic growth and without questioning, people all over the world take this assumption as a fact. In a study by Demirgüç-Kunt & Maksimovic a country s efficiency index is used, which is based on the such called law-and-order determinant by La Porte et al.. As conclusion of their study, Demirgüç- Kunt & Maksimovic proved that firms stated in countries with better scores on the efficiency index, as have better investor protection according to La Porte et al., do use more long-term external financing (Demirgüç-Kunt & Maksimovic, 1998). Another study by Qian & Strahan (2007) also used the country factors proved by La Porte et al., investigating the relation between laws and institutions with respect to durability and interest rates of bank loans. As explanation they say that a bank must assess not only the credit quality of the borrower, but also assess risks caused due to weak laws and institutions. They finally concluded that under better creditor rights, common law countries as suggested by La Porte et al., loans have longer maturities and lower interest rates (Qian & Strahan, 2007). Giannetti (2003) used the law-and-order variable to show how legal rights and institutions have effect on corporate finance decisions for unlisted firms of eight European countries. As already suggested by the study of La Porte et al., Giannetti found evidence that in countries with good creditor protection rights, as defined by the law and finance theory, it is easier for firms to invest in intangible assets to obtain loans. The protection of creditor rights is also important for ensuring access to long-term debt for firms operating in sectors with highly volatile returns. Creditor protection as defined by the law and finance theory thus may help to avoid distortions due to the excessive liquidation of firms in temporary difficulty, which is often associated with frequent short-term debt renewal decisions (Giannetti, 2003). Althoug I have some doubts about the explanation by La Porte et al., most of those studies prove that firms in common law countries seem to have access to other and or more long-term debt resources compared to their civil law peers. This is the same conclusion Graff came up with in his study. However, he has some critical remarks about the study and conclusions of La Porte et al.. First of all he mentioned the fact that it is not reliable to compare common law countries (mainly the US and the UK), with civil law countries (all other continental European countries). His study does not exclude, it even supports, the idea that legal tradition could be a major factor in shaping corporate law, although not necessarily along the lines suggested by the theory of law and finance by La Porte et al.. In particular he proved that there is not much evidence that common law countries protect financial investors better than civil law countries, although he found support for the assumption that financial investors are treated differently across legal families (Graff, 2008). This conclusion makes it relevant to Master Thesis Accounting R.H.F.M. Schonbrodt 9

10 investigate whether it are not the law families which lead to a different capital structure, but solely the available financial resources in those countries. The idea of having just other financial resources in common law countries compared to civil law countries, is supported by studies about the difference of long-term debt resources in Europe and the US. In this case, studies of venture capital markets in Europe and the United States are leading. The study by Gordon (2004), about firm performance and productivity, already suggests that there is a huge difference between the common law system in the US and the civil law system of most countries in Europe. In his study he gives differences in the education system, cultural hierarchy differences and lifestyle differences as possible reasons for the gap between the US and Europe. However, his most important argument is the financial flexibility in the US compared to Europe. As explanation he argues that a long tradition of public disclosures and information and access of equity research analysts to internal company information has fostered a large and active market for public offerings. This together with available venture capital provided the financial needs (start-up) firms are looking for (Gordon, 2004). There are not only studies which provide suggestions about the influence of venture capital over the financial market, but also studies that provide numerical evidence. A study by Schertler (2001) gives an overall data analyses in which the venture capital market of Europe is compared with the US. By giving an objective analysis of both markets, her conclusion is that the private-equity investments in Europe are significantly lower than those in the US. Searching for an explanation, she found that the venture capital market in the US is much more developed than the European venture capital market. Despite this is not due to better creditor protection rights in common law countries, regulations are responsible for those different markets. Since the late seventies of last century, the American government has improved the environment for venture capital by reducing the capital gain tax rate and moreover by allowing pension funds to take higher-risk investments. It took Europe until at least the mid-1990 s before venture capital was seen as a accepted way of financing (Schertler, 2001). The main question of this study will be whether the suggested better creditor protection laws, as defined in the law and finance theory, result in different capital structures and whether this will have an effect on firm performance. To do this, the study will be divided. The first part of the study will purely focus on the debt-to-total asset ratio. I will try to prove that unless the differences in creditor protection and shareholder protection rights between civil law countries and common law countries, following La Porte et al., the differences with respect to the relation between leverage and firm performance is not that huge as suggested. Assuming that a superior system, as mentioned by La Porte et al., should lead to a more positive relation between leverage and firm performance, I first want to disapprove the following Hypothesis: H 1 : Leverage will have a more positive effect on firm performance for firms vested in countries in which creditors are better protected as suggested by La Porte et al.. The second part of the study will focus on the way how firms are financed in those different legal families. Due to the potentially different financial resources available in those countries, it could be Master Thesis Accounting R.H.F.M. Schonbrodt 10

11 possible that the relationship between equity, current liabilities and non-current liabilities is not equal. Assuming that it is easier for common law firms to find other sources of long-term external debt next to or as substitute for the traditional bank loans, the use of those cheaper forms of external debts should have a more positive influence on firm performance. This with respect to their civil law peers. Assuming this, I want to prove the following Hypothesis: H 2 : The use of long-term debt by American firms (common law) will have a more positive effect on firm performance than the use of long-term debt by German firms (civil law). Mention that, even if this study proves or disapproves Hypothesis 1 and Hypothesis 2, the purpose of this study is that it will show that it is dangerous to claim one superior law system over another. Assuming that the availability of resources could be a reason of different capital structures and different laws and institutions, those differences could be a result of those different resources (opportunities). This in contrast to theory of law and finance, which assumes that those resources (opportunities) are just a product of superiority by common law over civil law. In that way, Hypothesis 1 is used to test whether the common law system is better than the civil law system, following La Porte et al., without discussing the problem whether it is possible that one system is superior over another. Results with respect to Hypothesis 2 are used to emphasize the financial differences instead of claiming a superior system over another. Finally I will use a random sub-sample of German firms and US firms to study the liability side of both countries in more detail. Doing so the study wants to prove that the structure of financial resources of US firms differ from the financial resources of German firms, due to other market conditions. Mention that I will especially focus on the structure of the financial debts of firms instead of the total liability side, expecting that the partition of equity and external debt will be more or less equal as mentioned by the study of Rajan & Zingales (1995). Master Thesis Accounting R.H.F.M. Schonbrodt 11

12 Data This study aims to prove that common law is not superior over civil law in a way suggested by studies as done by La Porte et al.. To compare common law with civil law, the study will use a sample of US firms and German firms. Since the US market is the biggest common law market and the German market is the biggest civil law market, this choice is straightforward. Because this study wants to prove differences of (external) financing possibilities between German firms and US firms, it will just focus on firms with 5 up to 500 employees. By restricting the number of employees as a variable of firm size, international influences should be mitigated as much as possible. Where studies done by Gordon (2004) and Schertler (2001) mention the differences of financing in the different industry sectors, especially with respect to the new high-tech firms, the sample will also be restricted by industries. This study will only use firms registered in old fashioned industry branches. This means that all firms are registered by the industry code NAICS 11 until 33. The study will be an up to date investigation, using firm data over the financial years 2009 and Data of both years are included to reduce the influence of economic changes. Ensuring that the data will finally contain enough information, the minimum value of the total assets is set to be 0. All data used in this paper are generated from the data base Orbis. Orbis is part of Bureau van Dijk Electronic Publishing (BvDEP), which is a leading provider of customized business information solutions. Orbis itself is a global database which has information approaching up to 55 million companies. The information is sourced by many different local information providers, all experts in their regions or disciplines. BvDEP standardizes balance sheet information with the stated objective of achieving uniformity and enabling cross-border analysis (BvDEP, 2011). Table 1 shows the total number of selected firms by Orbis, after the full selection restrictions are implemented. Although there are far more firms from the US selected in the total sample, it is quite remarkable that after the selection criteria the total number of firms of the US selected by Orbis are less than the total number of firms selected in Germany. Reason for the enormous drop of US firms is especially caused by the last restriction step; selecting for financial information. Apparently it is easier for BvDEP to collect complete financial information of selected German firms compared to selected US firms. This is a bit surprising since Gordon (2004) argues that the US market is characterized by the long tradition of public disclosures and information and access of equity research analysts to internal company information compared to foreign markets. On the other hand, those results could be expected, realizing that US firms are not required to fill 10-K financial reports when their Total Asset value does not exceed $75 million (The University of Cincinnati College of Law, 2009). Overall, this already emphasizes the huge influence a data base as Orbis could have over studies like this. Master Thesis Accounting R.H.F.M. Schonbrodt 12

13 Table 1 Number of firms in the final sample Germany United States Firms in the total sample Firms after industry restriction Firms after employment restriction Firms after financial year restriction Firms after TA restriction No. of firms in original sample Table 1 reports the number of firms offered by the data base Orbis, which are used in this paper. All firm data represent the financial years 2009 and 2010; represent the industry code NAICS 11 till 33; have 5 up to 500 employees; and have a minum of 0 Total Assets. Table 2 gives a more detailed look of how the balance sheet ratios are divided per balance sheet item by country. It has to be mentioned that the standardized financial overview by Orbis is used for this table. Studies by Giannetti (2003) and Rajan & Zingales (1995) already mentioned remarkable differences between countries regarding to current and non-current liabilities. This while the equity to total liability ratio is more or less equal to each other. Giannetti argues that the differences in current versus non-current liabilities are due to better creditor protection, following La Porte et al.. On the other hand Rajan & Zingales mention the fact that the overall leverage between different countries is almost the same. Those two different points of view just summarize the problem which will be discussed in this paper. As the paper will focus on the capital structure, differences on the asset side will not be discussed. However, it has to be mentioned that the value of Total Assets of US firms is significantly larger than the total value of German firms, even though the average number of employees is not that different between both countries. As explanation of this gap it is useful to look to the median of both countries total assets. By comparing the median of both countries total assets, it can be concluded that the difference between both countries is already significantly smaller. This means that the enormous gap between the average total assets of both countries is partly caused by large positive outliners in the US-sample. Another and probably more determinative explanation is the fact that US firms are not required to fill 10-K financial reports when they have an aggregate worldwide market value of the voting and non-voting common equity held by its non-affiliates of $75 million or less as of the last business day of the issuer's most recently completed second fiscal quarter. This requirement is defined by Rule 12b-2 of the Securities Exchange Act of 1934 (The University of Cincinnati College of Law, 2009). By using ratios with respect to total assets, this problem will be mitigated as far as possible. Finally Table 2 shows that even with a sample of over German companies and more than US companies, there are significantly less firms available with complete financial information, despite of the minimum value of Total Assets restriction mentioned in Master Thesis Accounting R.H.F.M. Schonbrodt 13

14 Table 2 Balance sheets for firms in Orbis Germany United States Assets Non-current assets 46,4% 74,6% fixed tangible 36,2% 59,8% fixed Intangible 2,1% 10,1% other non-current 8,0% 4,6% Current asstes 53,6% 25,4% current assets; stock 15,4% 4,0% current assets; debtors 12,9% 6,6% cash and equivalents 9,1% 11,6% other current assets 16,2% 3,2% Total assets 100,0% 100,0% Liabilities Shareholder funds Non-current liabilities long-term debt 39,5% 32,0% 17,8% 43,5% 42,5% 29,0% provissions 14,0% 7,8% other non-current liabilities 0,2% 5,7% Current liabilities Loans 28,5% 5,1% 14,1% 1,6% Creditors 7,4% 3,7% Other current-liabilites 16,0% 8,7% Total Liabilities 100,0% 100,0% Average Total Assets $ $ Median total assets $ $ No. observations Average No. of employees The value of each item is calculated as a fraction of the bookvalue of total assets and averaged across all the firms reporting a given balance sheet item per country in 2009 and It have to be mentioned that Orbis makes standardized financial overviews which ables cross-border analysis. Empty cells are deleted to keep the balance sheets clean of bias. Table 1. This time the data are also restricted by the availability of variables which will be used for the linear regression model. To reduce the influence of outliners in the linear regression models, the top and bottom 1 percentage of all variables are excluded. Again the data suggest that it is much easier for BvDEP to collect complete financial information for German firms than it is to collect complete financial information of US firms. The linear regression models will use variables which are also used in studies by Dermirgüç-Kunt & Maksimovic (1998), Rajan & Zingales (1995), Jong, Kabir & Nguyen (2008), Gaud, Hoesli & Bender (2007) and Margaritis & Psillaki (2010). As argued by Rajan & Zingales, those variables have shown up most consistently as being correlated with leverage in previous studies. Due to this specific correlation, most studies used more or less the same variables to investigate the influence of country specific laws and regulations over the capital structure. To measure firm performance (PERF), this paper will use the Return On Assets (ROA), which is measured by total Earnings Before Interest, Taxes, Depreciation and Amortization (EBITDA) over Total Assets (TA). High and stable operating profits decrease the possibility of bankruptcies and decrease the probability of fully exhausting tax Master Thesis Accounting R.H.F.M. Schonbrodt 14

15 shields. In this way profitability should increase the target debt level which suggests a positive relation between firm performance and leverage, according to the trade-off theory. In this case, the ROA variable is directly linked with the debt-equity choices. However, empirical studies about capital structure from the past showed a negative relation between firm performance variables and leverage. Arguments in favor of this result are based on the pecking-order theory. Pecking order models argue that internal financing is cheaper over external debt, because it avoids underinvestment costs. Lack of pressure on managers or transaction costs may also result in a financing behavior where profitable firms accumulate internal financing without rebalancing (Gaud, Hoesli, & Bender, 2007). According to all those studies about capital structure, a negative relation between ROA and leverage is expected. Although the total number of employees is used as a restriction variable computing the total sample, the natural logarithm of sales is used as a proxy for firm size (SIZE). This is a common measure for size, following studies by Gaud et al. (2007), Rajan & Zingales (1995), Jong et al. (2008) and Magaritis & Psillaki (2010). As larger firms normally have more stable cash flows and lower bankruptcy costs, it should be easier and cheaper for them to attract external debt. According to this, a positive relation is expected between leverage and firm size. Tangibility (TANG) is measured by the ratio of tangible assets over total assets, as done by Rajan & Zingales (1995) and Jong et al. (2008). If a large fraction of a firm s assets are tangible assets, then assets should serve as collateral, diminishing the risk of the lender suffering the agency costs of debt. Therefore, the greater the portion of tangible assets, the more willing lenders should be to supply loans (Rajan & Zingales, 1995). Because of this, a positive relation between tangibility and leverage is expected. The growth variable (GROWTH) is measured by sales growth. This variable can also serve for growth prospects and investment opportunities. As growth opportunities stand for future profitability, it is assumed that external debt suppliers are more prepared to supply financials (Margaritis & Psillaki, 2010). Therefore a positive relation between the growth proxy and leverage is expected. Liquidity (LIQUID) is measured by the ratio total current assets divided by the total value of current liabilities. The more liquid a firm is, the lower the chance of financial distress. In this way, a positive relation between liquidity and leverage is expected. Finally, a proxy is set with regard to the capital structure. Accordingly to measure capital structure, we have to deal with the leverage ratio. However, as already discussed by Rajan & Zingales (1995), there are many ways to measure such called leverage. The broadest definition of leverage is the ratio of total liabilities to total assets. This can be viewed as a proxy for what is left for shareholders in case of liquidation. On the other hand several researchers argue that the broad definition of leverage should be corrected for trade credit and provisions arising from labor market contracts or specific regulations with no importance to financing decisions, before attempting any international comparison. Such a less comprehensive definition of leverage should only depend on short-term and long-term financial liabilities and shareholder funds (Giannetti, 2003). Due to a lack of uniformity, the definition of leverage can have enormous impact over the outcome of a particular study. Regarding to this paper, both two different variables for leverage are used to answer respectively Hypothesis 1 and Hypothesis 2. As the first hypothesis wants to prove that there is not that much difference between common law countries and civil law countries with respect to the relation of leverage and firm performance, the way to measure leverage has to be in line with how creditor- and shareholder protection is measured. By the Master Thesis Accounting R.H.F.M. Schonbrodt 15

16 studies of La Porte et al. funders of financial resources are divided into just two groups, which both have own (control) rights. On the one hand you have shareholders providing equity and protected by such called shareholder rights, on the other hand you have creditors providing all kinds of external debts protected by such called creditor rights. In this case, since you have just a deviation in shareholders and creditors, the broadest definition of leverage is the only appropriate measure as a proxy for capital structure. Saying so, leverage (LEV) is measured by the ratio of total liabilities divided to total assets. According to the second hypothesis, it is not the deviation between creditors and shareholders on which the definition of leverage has to been based. The second hypothesis wants to prove that the better developed external debt markets in the US, as suggested by Gordon (2004) and Schertler (2001), lead to cheaper external debt for US firms compared to German firms with a less developed external debt market. In other words, Hypothesis 2 wants to prove that firms are financed in a way financial resources are available in that particular market. In this way it are not the laws and institutions which are the reason for a different capital structure between different countries, but assume to be a result of the different use of financial resources. The fact that laws and institutions react and interact with the availability of resources and meaning of definitions, often based on decisions of the past, is also concluded by Schee. Van der Schee argues that decisions made in history have lead to different laws and institutions in the present, resulting in different threats and opportunities (Schee, 2011). To measure the availability of financial resources the less comprehensive definition of leverage is used. The adjusted leverage (AdjLEV) is measured as the ratio of financial debt to the book value of shareholders funds plus financial debt, as defined in the study by Giannetti (2003). In this study financial debt is defined by long term debts and (short-term) loans. Table 3 presents the descriptive statistics of firm characteristics by country. From all those variables, LEV, AdjLEV and the PERF-variable will be the most important for this study. In that way it is good to see that the other variables between US firms and German firms are pretty comparable to each other. Without discussing those other variables by every detail, it is remarkable that the average growth rate of German firms is decreasing, even more than 50 percent of the firms showed a decreasing variable, while the performance ratio is highly positive. This while the US ratio just shows the other way round. The negative sign of the average growth rate variable of German firms can be explained by the significant economic downturn in 2009 by almost 5%, by an economic growth in 2010 of just 3,6% (Bhageloe-Datadin, Rita; Jong, Jurriën de, 2011). In the same time period the US economy decreased by just a little bit more than 2,5% in 2009 while it increased by almost 3% in (Boussebaa, Dieperink, & Kneepkens, 2011). Looking to the three main variables of this study, it is more interesting, and regarding to the law and finance theory by La Porte et al. even more surprising, that the average Leverage (LEV) and average Adjusted Leverage (AdjLEV) ratios of German firms are significantly higher than those of comparing US firms. In countries with better creditor protection rights as defined by La Porte, you should assume more, longer and cheaper external debt as proved by Qian & Strahan (2007) and Giannetti (2003). Demigüç-Kunt & Maksimovic (1998) even claim that they have proved that firms stated in countries in Master Thesis Accounting R.H.F.M. Schonbrodt 16

17 Table 3 Descriptive statistics of firm characteristics LEV AdjLEV PERF SIZE TANG GROWTH LIQUID ob s. United States 0,4852 0,2876 0,0143 4,6948 0,2917 0,2829 2, (04580) (0,2046) (0,0560) (4,7402) (0,1772) (0,1148) (2,6263) Germany 0,6561 0,3787 0,1245 4,4611 0,3046-0,026 2, (0,6729) (03403) (0,1180) (4,5126) (0,2585) (-0,039) (1,8066) Table 3 represents mean (and median in parentheses) values of leverage and other firm specific characteristics from both countries. All data are averaged, using the unifrom Orbis Data base over the year Variables are as follows. LEV: ratio of total liabilities divided to total assets; AdjLEV: adjusted ratio of financial debt to shareholder fund plus financial debt, where financial debt is calculated by long-term debt and (short-term) loans; PERF by ROA as total Earnings Before Interest, Taxes, Depreciation and Amortization over Total Assets; SIZE as a natural logarithm of turnover; TANG defined by the ratio tangible assets over total assets; GROWTH is measured by sales growth; and finaly LIQUID as the ratio total current assets divided by the total value of current liabilities. Obs. mentions the number of firms per obeserved per country, which is equal to the number of obersevations in table 2. which creditors are better protected as stated in the law and finance theory, do use more external debt financing. On the other hand, the outcomes of Table 3 are almost equal to the research of Graff (2008). In his study he mentioned that La Porte et al. (1998) already came to the conclusion that, despite of his law and finance theory, the US is one of the most anti-creditor common law countries in the world. Another conclusion they made was the fact that their predictions did not hold for the German origin country group, which La Porte et al. judged as just doing better than suggested. Saying so, it is not hard to agree with Graff, when he claims that the aggregation of data by La Porte et al. into just two dimensions of investor protectiveness does not offer particularly revealing insights into the law and finance relationship. It has to be mentioned that although the higher leverage percentages of the German firms compared to the US firms are not in line with the expectations of the law and finance theory, it does not mean that those descriptive statistics already disapprove the assumed better creditor protection rights of common law countries compared to the creditor protection rights of civil law countries. As will be discussed, this study will measure superiority as an effect over the performance variable (Return On Assets), as far as it can be tested at all. Despite the fact that as well the Leverage as the Adjusted Leverage variables of German firms are significantly more positive than their US peers, it does not mean that their relationship with the performance variable is more positive. Also note the difference between the leverage variable of US firms reported in Table 3 and the average leverage ratio of US firms reported in Table 2. This difference can be explained by the structure of the US sample. Since the US sample contains relative much positive outliners with respect to the value of total asstes compared to the German sample, those outliners have more influence over the average leverage ratio as reported in Table 2. Besides, as argued by Jong et al. (2008), those larger firms are normally financed with more external debt than smaller firms. Other remarkable results might be visible after running the regression models. Master Thesis Accounting R.H.F.M. Schonbrodt 17

18 Results Methodology With Hypothesis 1 the conventional theoretical framework on capital structure choice of firms is tested. As the influence of capital structure over firm performance is tested, the dependent variable in this study is firm performance measured by ROA, where other studies use the leverage variable as dependent variable. This choice is argued by the fact that this study wants to discuss whether common law is superior over civil law or not, as suggested by La Porte et al.. As La Porte et al. argues that the common law system provides a better framework for financial developments and economic growth, you should at least expect that firm performance should be more positively affected by this better framework. Since those differences are measured by the creditor- and shareholder protection variables as suggested by La Porte et al., the influence is directly measured by the leverage variable. So to conclude whether a law system is superior over another system, this study will use data which measure the impact of the independent variable Leverage over the dependent variable ROA. Unlike other researchers who do believe in the theory of law and finance, this study will not use the outcomes of Hypothesis 1 to claim that the one law family is superior over the other family. The outcomes of Hypothesis 1 will be used to emphasize the differences between financing in common law countries compared to financing in civil law countries tested in Hypothesis 2. Where the law and finance theory suggests that financial development and economic growth is a product of the particular law system, this study suggests that the law system could be a product of economic growth and financial development instead of being the reason for financial development and economic growth. To complete the equation and to increase the accuracy of the regression model used to answer hypothesis 1, other frequently used variables with respect to capital structure studies are added. The model used to run the ordinary-least-squares regression trying to disapprove Hypothesis 1 is as follows: PERF = SIZE + 4 TANG Since Hypothesis 2 especially focused on the way financial decisions are made in different countries, especially with respect to long-term debts, the leverage variable used for Hypothesis 1 is adjusted. However the idea is more or less the same. As counterargument for a superior law system over another, this study suggests the availability of financial resources as possible explanation for capital structure differences between countries. Where several studies prove that (external) financial debt markets in the US are more developed than (external) financial debt markets in Europe, this study suggests that it is easier for American firms to find other sources of long-term external debt next to or as substitute for the traditional bank loans. The use of those cheaper forms of external debts should have a more positive influence on firm performance. This again with respect to their German peers. To find evidence for this suggestion, the leverage variable is adjusted in a way that only financial debt is included. To test Hypothesis 2 the following model is used to run the ordinary-least-squares regression: PERF = SIZE + 4 TANG Master Thesis Accounting R.H.F.M. Schonbrodt 18

19 Both models will run separately for German firms and for US firms. Although both tests will focus on the interaction between the leverage variable and performance variable, there are expectations for the other variables. As mentioned before, Size; Tangibility; Growth and Liquidity are expected to have a positive relation with the Leverage variable. Note that it is not the leverage variable which is the dependent variable this time, but the performance variable. However, by expecting a negative relationship between leverage and firm performance, the expected relation between performance and Size; Tangibility; Growth and Liquidity is the same as the relation between those four variables and leverage, by reshuffling the equation. The relation of firm performance and the leverage variables of German firms will be compared to those of US firms. In the first model this means that the relation of the leverage variable (LEV) with respect to firm performance (ROA) will be compared with its US peers. Hypothesis 1 will test whether firm performance is more positive affected by leverage in common law countries over civil law countries. In that way Hypothesis 1 is disapproved when the lower bound of the US leverage variable with a 95% confidence interval is smaller than the upper bound of the German leverage variable with a 95% confidence interval. The test for Hypothesis 2 is comparable to the first test. This time an adjusted leverage variable is used to test whether firm performance is more positive affected by external debt in the US than it is in Germany, by assuming that financial external debt markets in the US are better developed. Results: running the regression models The regressions reported in this section investigate the aspects of how firm performance, measured by Return On Assets, is affected by firm specific characteristics as Size, Tangibility, Growth, Liquidity but overall how firm performance is affected by different leverage ratios. The sample contains all German firms as well as all US firms, as reported in Table 3, which contains enough financial information to run the regression models. Table 4 represents the results of the first regression. The results indicate that the Tangibility variable is indeed positively related to firm performance, for both countries, as already concluded in prior research by Rajan & Zingales (1995) and Jong et al. (2008). The expected positive influence of Size over firm performance, by Gaud et al. (2007), Rajan & Zingales (1995), Jong et al. (2008) and Magaritis & Psillaki (2010), does only hold for US firms, as the expected positive influence of firm growth, Magaritis & Psillaki (2010), does only hold for German firms. The ambiguous results for the German Size variable can be the result of the selection criteria. Those criteria lead to relatively small German firms and much less variance in the German sample with respect to their US peers. The results of the US Growth variable indicate a slightly positive influence of firm growth over firm performance. However, in a 95% confidence interval those results seem uncertain. The same holds for the Liquid variable for German firms as well as for US firms. Looking in more detail to those unexpected results, especially to the Liquid variables, it is noticeable that all those outcomes are not significant (see Appendix A.1). The significant value, called the p-value in statistics, indicates the Master Thesis Accounting R.H.F.M. Schonbrodt 19