The Impact of the Financial Crisis on Firms’ Capital Structure and Cash Holdings in Germany and the United States

(1)

The Impact of the Financial Crisis on Firms’ Capital Structure and Cash

Holdings in Germany and the United States

Master Thesis

University of Groningen – Faculty of Economics and Business Master of Science - International Financial Management

Supervisor: Dr. H. Vrolijk

Johannes Maximilian Duwenhorst Student number – 2197278

Aweg 5c 9718CS Groningen

(2)

1

Abstract

The financial crisis of 2007/2008 had severe consequences for the financial management developments of corporations. This study investigates the impact of the financial crisis on leverage and cash holdings of German (bank based economy) and US firms (market based economy). By using a panel data regression, trends of cash and leverage during the pre-crisis- (2006 to 2007), crisis- (2008 to 2009) and post-crisis period (2010 to 2014) were investigated in both countries. The results indicate that the financial crisis had a significant impact on the leverage of German firms, while the American counterparts seem to be less affected. During the financial crisis, cash holdings of German and US firms increased, however the effect is insignificant. Further, the interplay of cash holdings and leverage shows that firms from the market-based economy (US) increase cash more heavily during the crisis. The insights from this research may enhance the understanding of institutional environments during times of financial distress.

Keywords: Leverage Ratio, Cash Holdings, Bank-Based and Market-Based Economies, Financial Crisis, Financial Markets

JEL: G11, G15, G21, G30

1. Introduction

The global financial crisis of 2007/2008 has left its mark and the economic aftershocks continue to impact firms; especially patterns of finance are strongly affected (Fosberg, 2012). In this regard, different economic structures of countries shape the way in which firms adapt to the changing conditions (Alves & Francisco, 2015). For instance, companies located in Germany and the US are commonly known to rely on different means of finance (Vitols, 2005). More precisely, firms in the United States make large use of equity finance and issue bonds on the capital market. Meanwhile, German companies traditionally finance their operations via banks. This study is interested in distinguishing the effects of the crisis on financial management in countries with contrasting institutional environments. Therefore, a differentiation between German and US firms will be made. As capital structures and cash holdings give good insights into financing decisions of corporations, these concepts will be investigated.

(4)

3 particular, the strength of banks or capital markets has the capacity to directly affect corporate leverage (de Jong et al., 2008). While firms located in countries characterized by strong banks have commonly higher leverage ratios, companies in countries with developed stock markets facilitate lower leverage ratios (Alves & Francisco, 2015). Yazdanfar & Öhmann (2015) outline that debt is generally cheaper than equity. However, debt burdens beyond satisfactory levels decrease profitability through higher financial risk, agency costs and higher interest rates (Arindam & Anupam 2014). Hence, managing capital structures in a balanced way is crucial and is a determinant for profitability. However, the management of capital structures is not always straightforward, as market conditions vary. This became obvious during the financial crisis, which came along with severe changes in the financial system. The accessibility to bank financing fluctuated tremendously, posing challenges to corporations around the world to manage their capital structure (Iqbal & Kume, 2014). Especially for highly levered firms, it became very costly to raise additional debt or equity (Kahle & Stulz, 2013). In the years before the crisis, firms relied on cheap debt and increased their leverage tremendously (Alves & Francisco, 2015). Thereupon, the outbreak of the financial crisis sparked a change in the accessibility of capital, which differed in intensity across countries (Alves & Francisco, 2015). The diverging developments in regard to capital accessibility have been discussed in a debate in July 2015. Carl Icahn, well known investor and owner of hedge funds, and Larry Fink, CEO of BlackRock, highlighted some of the differences between Europe and the United States. Larry Fink explained that the retreat of banks in both countries has hampered the European economy more than its American counterpart. He points out that due to the larger size of the capital market in the United States, private and institutional investors can compensate the lacking funds from banks. As a consequence, the availability of capital accelerated the recovery of firms in the US economy, while companies in European countries, such as Germany, remained in a state of depression for a longer period (Pramuk, 2015, July 15). Based on this argumentation, institutional differences between the United States and Germany seem to have cause diverging developments.

(5)

4 are associated with raising cash holdings during financial turmoil. Hence, further research surrounding the interplay between leverage and cash holding during the financial crisis is needed.

This study will investigate the following research question:

Did the impact of the financial crisis on firms’ capital structure and cash holdings differ between German firms (bank-based economy) and the US firms (market-based economy)? The remainder of this study is structured as follows. In order to refine the goal of this research, a literature review will outline the current knowledge on the relationship between the crisis and capital structure as well as cash holdings. In the third section, I will describe the methodology and the variables used during the empirical analyses. Furthermore, the sample selection will be clarified and preliminary observations will be drawn from the descriptive statistics. Subsequently, I am going to report the empirical results and continue with the corresponding discussion. Ultimately, the conclusion will summarize the main findings and show suggestions for future research.

2. Related Literature

The major contribution of this analysis is the indication of differences of the financial crisis’ effects on German firms (bank-based economy) and US firms (market-based economy). Therefore, it is useful to understand the different institutional environments. Besides, an understanding of the definitions and relationships between the main concepts helps to comprehend the effects of the financial crisis.

2.1. Institutional Environment

Next to firm-specific characteristics (Rajan & Zingales, 1995), the institutional environment is affecting the financial management of firms (de Jong et al., 2008). One key feature of the institutional environment is the setup of the financial market within one country. While companies in some countries rely on strong banks, firms in other economies benefit from a highly developed stock markets (de Jong et al., 2008).

(6)

5 value of firms listed on the stock exchange as proportion to GDP was less than one quarter of that of the USA or the UK (Vitols, 2005). The benefits for firms in bank-based economies can be manifold. Through closer relationships to their clients, banks are supposed to be better equipped than investors to identify profitable projects and monitor them more effectively. Moreover, banks’ commitment to long-term projects is higher than the commitment of fast moving capital market investors (Lee, 2012). Germany has been described as an economy with clear characteristics of a bank-based economy (Vitols, 2005; Iqbal & Kume, 2014). Therefore, Germany is a good example to detect the impact of the financial crisis on capital structures and cash holdings of firms that are more closely associated with bank funding.

The United States on the other hand are often described as the role model of a market-based economy. In 1996, only 25% of the financial system assets were bank based so that the importance of banks is minor in the United States in comparison to the capital market (Appendix 2). The market capitalization of US firms relative to the nation’s GDP was four times as high as in Germany. Additionally, the number of Initial Public Offerings between 1986 and 1996 was 663 in the US, while only 19 firms were newly listed in Germany (Vitols, 2005). Another key element of a market-based economy is the legal framework. Rather than supporting creditors (Germany), US law is very much concerned with protecting investors. La Porta, Lopez-de-Silanes, Shleifer and Vishny (2000) have established an index that represents the level of shareholder protection on which the US scores 5 out of 5, while Germany scores 1. The United States are a clear case of a market-based economy, because firms are less dependent on banks (Lee, 2012). Resultantly, a sample of German and American firms allows me to distinguish between the impacts of the financial crisis on capital structures as well as cash holdings in a bank- and market-based economy.

2.2 Main Concepts

2.2.1 Capital Structure (Leverage Ratio)

(7)

6 from interest payments that are deducted from taxable income. It is interesting to note that Morellec, Nicolov and Schürhoff (2012) show in their study that the choice of capital structure goes beyond tax issues, bankruptcy- and refinancing costs. Based on the agency theory, empirical evidence has been found that capital structures also depend on the severity of conflicts between shareholders and management (Morellec et al., 2012).

In general terms, capital structure is commonly defined as the proportion of debt to equity a company uses to finance its long-term growth (Alves & Francisco, 2015). More specifically, the short-term debt is comprised of liabilities that is due within 12 months and is usually characterized by short-term bank loans and accounts payable. The long-term debt of firms includes bonds, individual notes payable and leases that mature beyond the 12 months (DeAngelo & Roll, 2015). Equity on the other hand consists mainly of common stocks, but also of preferred stocks and retained earnings (DeAngelo & Roll, 2015). One of the major theoretical issues that concerns research surrounding capital structures is its stability. Lemmon, Zender & Roberts (2008) believe that corporate leverage ratios are stable, since they found that highly leveraged firms are likely to remain so for at least two decades. The recent study of DeAngelo and Roll (2015) objects this claim. They detect that very few firms keep debt to asset ratios constant over time, due to diverging firm- as well as institutional characteristics.

In order to gain insights into capital structures, I will analyse leverage ratios of firms. Leverage ratio is most commonly characterized by the ratio of total debt to total capital (Iqbal & Kume, 2014). As this study distinguishes between bank and market-based financing, leverage ratio will be defined as the proportion of loans to total capital and acts as dependent variable.

2.2.2 Cash Holdings

(8)

7 for firms with short debt maturities and high debt levels, saving cash from its cash flows is reducing refinancing risk.

In times of tight credit, the incremental value of one dollar of cash holdings for firms with shorter maturity debt may increase (Hugonnier, Malamud & Morellec, 2014). While additional cash reduces refinancing risks, holding cash beyond rational levels can be a waste of investing opportunities. In fact, large cash reserves can become very costly, as firms’ returns from liquid assets are relatively low. Therefore, Hugonnier et al. (2014) show in an empirical study “that there exists a target level for cash holdings” (Hugonnier et al., 2014, p. 393). Thus, generally excess cash above this level should be invested in projects, while a shortage of cash should result in decreasing investments.

During the last decade, the cash holdings of major corporations have reached a peak as a result of the financial crisis (Appendix 1). Given that these deposits are returning very little interest, the question of why corporations hold on to a lot of cash arises. Song and Lee (2012) show that a shock such as the financial crisis initiates a decrease of firms’ investments and change the cash demand. Similarly, Yujun et al. (2011) found evidence that companies become cautious in times of financial turmoil and establish cash reserves. In accordance with Song and Lee (2012), this study defines cash ratio as the cash and short-term investments to book-value of total assets. Consequently, cash holdings will act as second dependent variable. During the empirical analyses, I will determine how cash ratios developed from the pre-crisis to the post-crisis period. Moreover, I will analyse how leverage influences the relationship between the financial crisis and cash holdings.

2.2.3 Financial Crisis

(9)

8

2.3 Relationships between Concepts

2.3.1 Relationship between Capital Structure and Financial Crisis

Financial crises have been frequently used to explain the changes in capital structures of firms around the globe (Fosberg, 2012; Kahle & Stulz, 2013). The research, however, was concentrated on developments within single countries. Moreover, the institutional setting as determinant for diverging capital structures and cash holdings is often overlooked. Hence, evidence for differences of firms within bank-based or market-based economies is scarce. One of the related studies (Voutsinas & Werner, 2011) found however that the Asian crisis of 1997 sparked a sudden decrease of Japanese firms’ debt ratios. Similarly, Lim (2003) reported that the role of banks and the general leverage level decreased during the Asian crisis. For example, Korean companies decreased the reliance on banks and used the capital markets so as to compensate for the lacking funds (Lim, 2003).

(10)

9 and France are used as bank-based economies. They find a significant impact of the financial crisis on capital structures in the UK as well as Germany. Further it has been shown that leverage ratios for British and German firms increase from pre-crisis to crisis period, but decrease from crisis to post-crisis period. Based on this literature review, I hypothesize that: Hypothesis 1: Leverage ratios of firms in Germany (bank based economy) decreased more from crisis to post-crisis period than leverage ratios of firms in the United States (market based economy).

Figure 1 – Conceptual Framework Hypothesis 1

2.3.2 Relationship between Cash Holdings and Financial Crisis and the Role of Leverage Song and Lee (2012) have studied the long-term effects of the Asian crisis on cash holdings of corporations in eight countries. They found empirical evidence that “the financial crisis has systematically changed the cash holding policies of the firms and has a long-term effect” (Song & Lee, 2012, p. 639). In fact, the crisis changed firms’ demand function and triggered an increase of cash holdings. In line with these findings, other researchers (Sun & Wang, 2015) have found an initial decrease of cash holdings in response to limited access to external funding during the financial crisis. However, once the demand of the real economy contracted, corporations began to build precautionary savings and increased cash holdings (Sun & Wang, 2015). Such a development has been observed for firms within bank-based economies (Iqbal & Kume, 2014) and market-based economies (Akbar, Rehman & Ormrod, 2013) alike.

(11)

10 higher interest rates. In order to mitigate these refinancing risks, corporations tend to accumulate cash holdings (Harford et al., 2014). I will investigate the development of cash holdings during the financial crisis and hypothesize:

Hypothesis 2a: The outbreak of the financial crisis triggered an increase in cash ratio for firms located in Germany (bank-based economy) and the United States (market-based economy) from crisis to post-crisis period.

Harford et al. (2014) investigate the role of cash holdings in regards to refinancing risks. It has been found that cash decreases the risk of short-term maturing debt and high debt levels. Especially during the financial crisis, firms with high debt levels benefit from high cash ratios (Harford et al., 2014). However, the authors find a negative relationship between leverage ratio and cash holdings, which results from higher interest payments. Similarly, Song and Lee (2012) find empirical evidence that leverage has a negative relationship with cash ratios; especially after the outbreak of the Asian crisis. Hence, low-leveraged firms were found to have higher investment ratios and cash ratios, while highly indebted firms were found to have low cash ratios and decreased investments heavily during the crisis. Additionally, bank dependent firms hoard more cash during the financial crisis (Kahle & Stulz, 2013). As firms located in bank-based economies are characterized by closer relationships with banks, they are more likely to have high leverage ratios (Vitols, 2005). Thus, the following hypothesis has been developed: Hypothesis 2b: The impact of the financial crisis on cash holdings is more severe for firms located in Germany (bank-based economy) than for firms located in the US (market-based economy).

(12)

11 To sum up, this paper attempts to make three distinct contributions. Firstly, I will investigate whether the development of capital structure during differed between a bank- and market-based economy during the financial crisis. As mentioned before, in this study Germany and the United States are included as examples for bank- and market-based countries. The second contribution is to analyse, whether the financial crisis had an effect on cash holdings. Thirdly, I will combine the two concepts and investigate the role of leverage ratio on the development of cash holdings in the financial crisis.

3. Data and Methodology

The data for the following empirical analyses stems from Orbis. The database contains information on over 160 million companies with an emphasis on private company information. Thus, Orbis is a suitable tool to generate data for the subsequent financial analyses. The data covers the period from 2006 to 2014 so that the pre-crisis-, crisis as well as post-crisis periods are captured and a longitudinal analysis can be conducted. Orbis solely offers annual data, which renders a more detailed analysis of the crisis period impossible. Firms selected for the analyses are public companies listed in either Germany or the United States. Thus, the access of firms to bank loans as well as capital markets in their home country is ensured. As the study is concerned with firms operating in the non-financial sector; banks, other financial institutions and state owned companies are excluded from the dataset. The data set offers information on micro level, which enables an identification of country specific developments.

3.1 Operationalization of Variables

3.1.1 Dependent and Independent Variables

(13)

12 3.1.2 Control Variables

In order to ensure the accuracy and quality of the results, the analysis controls for certain influences that are not supposed to be analysed. In order to test Hypothesis 1, the following variables are controlled for. Firstly, firms’ size is taken into account, as it heavily influences the accessibility to financial resources (Voutsinas & Werner, 2011). In times of an economic downturn small sized enterprises have been found to face more constraints to access capital than their larger counterparts. Moreover, Frank and Goyal (2003) stress that firms with declining growth opportunities increase their use of debt. Hence, growth of a company must be controlled for in the study of leverage ratios. Another control variable that is crucial to be considered is the proportion of firms’ tangible assets. Companies with more tangible assets have commonly more collateral at their disposal and accumulate more debt than companies with few tangible assets (Frank & Goyal, 2003). Moreover, Frank and Goyal (2003) show that more profitable firms are expected to have lower debt ratios. Thus, this study will include the return on assets as control variable. Despite lacking evidence for the impact of asset uniqueness on leverage ratio, the control variable has been included in order to ensure comparability to the study of Iqbal and Kume (2014).

(14)

13

Table 1 – Overview of Variables

3.2 Method

Analysing effects that cover a long-term period and differ across entities requires a careful application of statistical methods. The most suitable approach to capture time as well as cross-sectional effects is using panel data. The benefit of this method is that unmeasurable variables can be included and controlled during the analysis (Lemmon et al., 2008). Further, panel data regression analysis allows to observe concepts in the research setup that change over time, but not across entities - such as the financial crisis. Especially when testing Hypothesis 2b, the panel data approach enables to differentiate between firms with high and low leverage ratios. Hence, one can make inferences about the impact of the crisis for firms located in different economies. Iqbal and Kume (2014) have shown that by using this fixed-effect panel data regression, valuable insights can be gained from time varying concepts.

Variable Description Calculation

Leverage Dependent Variable, Moderator & Control Variable

Total Loans from Credit Institutions / Total Capital Loan to Debt Alternative Dependent Variable Total Loans from Credit

Institutions / Total Liabilities

Cash Dependent Variable Cash + Short-Term

Investments / Total Assets Pre-Crisis Independent Variable (Dummy) 1 for years 2006 & 2007 Crisis Independent Variable (Dummy) 1 for years 2008 & 2009 Post-Crisis Independent Variable (Dummy) 1 for years 2010 to 2014

Size Control Variable Natural Log of Total Assets

Growth Control Variable Total Asset(t) - Total

Assets(t-1) / Total Assets(t)

Tangibility Control Variable Fixed Assets / Total Assets

Uniqueness Control Variable Research and Development

Expenses / Total Assets

ROA Control Variable Net Income / Total Assets

Dividends Control Variable (Dummy) 1 if Ordinary Dividends have been paid

NWC Control Variable Net Working Capital / Total

(15)

14

3.3 Sample

Initially, 8,013 German and American companies could be retrieved from the Orbis database. After eliminating observations with missing data and other disturbances related to the data, a final sample of 474 companies has been established (for a detailed outline of the steps during the sample selection see Appendix 3). The representative sample contains 244 US companies and 230 German companies so that the observations are evenly spread among the two countries. The sample does also contain a large variety of industries, which allows to draw conclusions on broader economical levels. The main industries are firstly, ‘machinery, equipment, furniture, recycling’ with 138 observations; secondly, ‘wholesale and retail trade’ with 69 observations and thirdly, ‘other services’ with 61 firms. Almost all industries are represented in both countries, except for ‘Hotels & restaurants’, which is only present in the US subsample (Table 2). Hence, inferences from the analysis are applicable to both countries in a similar manner. Only listed firms are included in the sample. Further, the firms’ total assets in the sample range from $1 million to $426 billion with an average of $8.5 billion. Ultimately, the sample includes an array of companies of different industries. With minor exceptions, the observations are fairly even spread among the two countries to a satisfactory degree.

Table 2 – Sample: Country and Industry distribution

Industry Germany United States Total

Chemicals, rubber, plastics, non-metallic products 29 19 48

Construction 4 1 5

Education, Health 6 4 10

Food, beverages, tobacco 10 13 23

Gas, Water, Electricity 10 9 19

Hotels & restaurants 0 6 6

Machinery, equipment, furniture, recycling 70 68 138

Metals & metal products 8 7 15

Other services 34 29 63

Post & telecommunications 7 5 12

Primary sector 3 5 8

Publishing, printing 9 9 18

Textiles, wearing apparel, leather 9 7 16

Transport 11 4 15

Wholesale & retail trade 17 52 69

Wood, cork, paper 3 6 9

(16)

15

3.4 Empirical Models

Since the goal of my analysis is manifold, three regression models are necessary in order to capture the effects of the financial crisis. The first model is concerned with the financial crisis’ impact on firms’ capital structure. The formula will include crisis as well as post-crisis dummies, so as to detect developments throughout the financial crisis. Moreover, Frank and Goyal (2003) suggest the inclusion of various firm-specific control variables (see equation 1). As a result, the following regression model has been formulated:

Leverageit = β0 + β1Crisisit +β2PostCrisisit + β3Sizeit + β4Growthit + β5Uniquenessit

+ β6Tangibilityit + β7ROAit+ ui + eit

As depicted in Table 1, the dependent variable leverage is calculated as the total loans divided by total capital. For the robustness analysis, the leverage ratio will be calculated as total loans to total debt, so as to exclude the impact of equity developments. Total capital is comprised of the total liabilities and total shareholders’ equity. Consequently, the variable captures the importance of bank financing in the capital mix of companies. The crisis variable, is constructed as dummy variable that takes on the value 1 for the period of the years 2008 and 2009 and 0 for the remaining years. The post-crisis variable is another dummy variable, which will be 1 for the period between 2010 and 2014 and the value 0 for the years preceding this period. Moreover, size is calculated as the natural logarithm of a company’s total assets. The formula also includes growth as a variable, which embodies the development of a firm by using the change in total assets. Tangibility is the proportion of fixed assets to total assets. Hence, the variable shows how much physical assets the company possesses and indicates whether the firm is product- or service-oriented. Furthermore, uniqueness refers to how individual a firm can be characterized. Therefore, the variable is calculated as the research and development expenses as a proportion of total assets. Lastly, the return on asset is calculated as the net income divided by total assets and captures the firm’s profitability and efficiency (Table 1).

(17)

16 of 1 in case a dividend has been paid in the year; 0 if no dividend has been paid. Lastly, following the example of Song and Lee (2012), the formula includes net working capital proportion to total assets.

CashHoldingsit = β0 + β1Crisisit +β2PostCrisisit + β3Leverageit + β4Growthit +

β5Dividendsit + β6itSize+ β7ROAit + β8NWCit + ui + eit

The third regression model attempts to clarify the interplay between cash holdings and leverage for German and US firms during the financial crisis. In order to analyse this, the sample will be split into two subsamples. I will divide the sample into German firms (bank-based economy) US firms (market-based economy). Alternatively, I will include a leverage dummy in the full sample analysis that takes on the value of 1 if the mean leverage ratio is above the mean and 0 otherwise (Equation 3). Therefore, the following formula has been established:

CashHoldingsit = β0 + β1Crisisit +β2PostCrisisit + β3Leverageit + β4Growthit +

β5Dividendsit + β6itSize+ β7ROAit + β8NWCit + β9LeverageDummyit + ui + eit

4. Results

4.1 Empirical Findings – Capital Structure (Leverage Ratio)

4.1.1 Descriptive Statistics

(18)

17 The data regarding the control variables support the notion of a fairly balanced sample. In terms of size US and German firms have comparable amounts of total assets. The mean as well as the standard deviation between the two subsamples differ only marginally. On average, German companies grew more (2.37%) than US companies (1.40%). However, US firms report higher return on assets than German firms. Moreover, the mean asset tangibility moves in similar lines so that the firms in Germany and the US are characterized by a similar proportion of fixed assets. The correlation matrix (Table 4) indicates no correlations between independent and control variables that might distort the results of the analysis. The high correlation between the two dependent variables as well as leverage can be neglected, as they are not regressed in the same model.

Table 3 – Descriptive Statistics of Subsamples

Table 4 – Correlation Matrix Full Sample

Mean Median Maximum Minimum Std. Dev. Mean Median Maximum Minimum Std. Dev.

Leverage 0.0791 0.0489 0.7750 0.0000 0.0918 0.0275 0.0103 0.8560 0.0000 0.0601 Loan to Debt 0.1247 0.0842 0.7700 0.0000 0.1252 0.0443 0.0185 0.7606 0.0000 0.0764 Size 5.7812 5.6038 8.6506 3.5888 0.9813 5.9719 6.0508 8.3112 2.8692 0.9937 Growth 0.0237 0.0301 0.9773 -23.1008 0.5940 0.0140 0.0363 0.9700 -11.8336 0.5121 Tangibility 0.5144 0.5029 0.9775 0.0269 0.1822 0.5426 0.5460 0.9913 0.0443 0.2105 Uniqueness -0.0190 -0.0010 0.0000 -0.2326 0.0319 -0.0722 -0.0429 0.0000 -1.1478 0.0958 ROA 1.7207 3.2680 67.9310 -98.4180 10.7033 2.0861 4.3105 65.0230 -95.3980 13.7579

German Subsample United States Subsample

Leverage Loan to Debt Cash Size Growth Tangibility Uniqueness ROA NWC Dividends Bank-Dep. Crisis Post Crisis Leverage 1.00 Loan to Debt 0.92 1.00 Cash -0.24 -0.22 1.00 Size -0.15 -0.19 -0.12 1.00 Growth -0.01 -0.01 0.01 0.04 1.00 Tangibility 0.07 0.06 -0.44 0.21 -0.04 1.00 Uniqueness 0.18 0.17 -0.40 0.14 0.00 0.25 1.00 ROA -0.15 -0.10 -0.10 0.25 0.04 -0.03 0.26 1.00 NWC 0.06 0.12 -0.20 -0.21 0.04 -0.55 -0.01 0.14 1.00 Dividends -0.10 -0.08 -0.14 0.48 0.06 0.05 0.17 0.33 0.05 1.00 Leverage Dummy 0.67 0.72 -0.26 -0.15 0.00 0.12 0.16 -0.09 0.06 -0.04 1.00 Crisis 0.06 0.06 -0.01 -0.01 -0.04 0.01 0.00 -0.09 0.01 -0.07 0.03 1.00 Post Crisis -0.05 -0.04 -0.01 0.03 -0.04 0.02 -0.01 0.08 -0.02 0.08 -0.04 -0.60 1.00 *italic values indicate correlations beyond the threshold value of 0.70

(19)

18 The descriptive statistics regarding the leverage ratio before, throughout and after the crisis lend preliminary support for the aforementioned assumptions. Figure 3 shows that German firms have a noticeable increase of leverage ratio prior to the crisis, followed by a sudden decrease between crisis and post crisis period. Throughout the post-crisis period (2010 to 2014) the leverage ratio for the German subsample shows a development at a constant level around 8%. In contrast, Figure 3 also shows that firms located in a market-based economy (United States) experience no fundamental decrease of leverage after the outbreak of the financial crisis. In order to check robustness, the annual loan to total debt ratio has been included as alternative dependent variable. This ratio develops in a similar manner as the leverage ratio (Figure 4).

Figure 3 – Leverage Ratio

(20)

19 4.1.2 Univariate Results

The univariate results are obtained by a paired sample t-test. Even though the statistics do not offer insights into the relationships between the central concepts, a paired sample t-test allows to make some general inferences. Especially a distinction of three different time periods enables a comparison of the situation at different time stages. The focus of the t-test lies on the dependent variables and their components. In order to get an idea of varying leverage ratios, I incorporated firms’ debt and equity values. These values enable an understanding of underlying reasons for fluctuations of leverage. The results of Table 5 include means and t-statistics of the German and American subsamples across three stages: pre-crisis (2006 to 2007), crisis (2008 to 2009) and post-crisis (2010 to 2014). The first two columns depict the development between pre-crisis and crisis period, the third and fourth column include crisis to post-crisis period and the last two columns outline pre-crisis to post-crisis developments for both subsamples. The data shown in Table 5 indicate an increase of leverage between pre-crisis and crisis period as well as a decrease from crisis to post-crisis period for German firms (significant at 1% level). However, the leverage means of pre-crisis and post-crisis period are not significantly different from each other. The leverage data of US firms outline a similar development throughout the crisis. In contrast to German firms, the means of US firms are in all three time periods not statistically different from each other. As expected, the leverage ratios of German firms decreased more heavily after the crisis than the leverage ratios of US firms.

(21)

20

Table 5 – T-tests and means for differences across the three periods

4.1.3 Regression Results

In order to analyse the relationships between the concepts of leverage ratio and the financial crisis, a fixed-effect panel data regression has been conducted. The results related to Hypothesis 1 can be found in Table 6 (column 3 & 4). Two different time periods have been included. The first two columns cover the years 2006 to 2014 for both countries. This period has also been used during the research of Iqbal and Kume (2014) and enables a comparison of the results. The second period covers the years 2008 to 2014. Hence, a detection of potential differing leverage ratios between German and US firms is possible. Based on Hypothesis 1, the leverage ratio of German firms is expected to decrease significantly more than the one of American companies. Table 6 outlines that companies in Germany experience an increase in leverage ratio from the pre-crisis to crisis period, indicated by the statistically significant value of the crisis dummy in the first column. The post-crisis dummy outlines a negative and insignificant coefficient (Column 1). This result shows that the leverage ratio of German firms increased from pre-crisis to crisis period and then returned to pre-crisis level after the crisis. The development for German firms is in accordance with Iqbal and Kume (2014). However, the results of firms located in a market economy stand in contrast to Iqbal and Kume (2014). Table 6 shows that US firms experience decreasing values of leverage ratio across all periods (column 2). Iqbal and Kume (2014) found that the leverage ratios for firms in a market-oriented economy (United Kingdom) increases during the crisis and then decreases to the pre-crisis level.

Periods Subsample

Leverage Ratio Means 0.0776 0.0913 0.0301 0.0267 0.0913 0.0745 0.0267 0.0268 0.0776 0.0745 0.0301 0.0268

t-stat Loan-to-Debt Means 0.1205 0.1405 0.0477 0.0431 0.1405 0.1196 0.0431 0.0434 0.1205 0.1196 0.0477 0.0434 t-stat ln (Debt) Means 12.6347 12.7861 12.9625 13.0880 12.7861 12.7991 13.0880 13.2723 12.6347 12.7991 12.9625 13.2723 t-stat ln (Equity) Means 12.1286 12.2238 12.6452 12.7016 12.2238 12.3099 12.7016 12.8720 12.1286 12.3099 12.6452 12.8720 t-stat

*,** & *** refer to 10%, 5% and 1% significance level, respectively.

Leverage Ratio is calculated as the total loans divided by total capital; Loan to Debt ratio consists of the proportion of total loans to total debt; Debt and Equity are the natural logarithms of total debt and equity, respectively

-4.3648*** -8.5169*** -5.8878*** -5.7373*** -0.5542 -6.6874*** -2.6625*** -1.3097*** -2.9592*** -4.5000*** -4.2005*** -4.4513*** -3.6178*** 1.2348 3.6754*** -0.0706 0.1563 0.9304 -3.2444*** 1.0786 3.8808*** -0.0269 0.7074 0.8924 Pre-Crisis to Post-Crisis

Germany United States Germany United States Germany United States

(22)

21

Table 6 – Regression Results for Leverage Ratio – Fixed Entity Effects

The third and fourth column of Table 6 are related to the analysis of Hypothesis 1 and hence, incorporate the period between 2008 and 2014. The data reflects how the leverage ratio of German and American companies developed after the outbreak of the financial crisis. Column 3 indicates a significant decrease of leverage ratio of German firms. On the other hand, American firms experience an insignificant increase in the post-crisis period (column 4). These findings lend partial support for Hypothesis 1, since German firms undergo a more severe decrease of leverage than US firms. It can only be partially supported, as the analysis has shown that the leverage ratios of German firms normalize in the post-crisis period – this differs from expectations. Their American counterparts experience neither an increase prior to-, nor a decrease after the crisis.

Sample Period

Subsample Germany United States Germany United States

(1) (2) (3) (4) Crisis (2008/9) 0.0122*** -0.0055* Post Crisis (2010-2014) -0.0008 -0.0034 -0.0122*** 0.003302 Size -0.0138 -0.0073 -0.0246 -0.0322*** Growth (in %) 0.0083*** 0.0011 -0.0054 0.0146* Tangibility 0.0048 0.0130 -0.0108 0.0348* Uniqueness 0.0402 0.3268*** 0.0731 0.2769*** ROA (in %) -0.00138*** -0.0006*** -0.0016*** -0.0005*** Constant 0.1570 0.0736 0.2398 0.2031 No of observtion 2070 2184 1610 1699 No of cross-sections 230 244 230 244 R-squared 0.6234 0.3731 0.6528 0.4254 F-Test 12.8559*** 4.6024*** 10.9940*** 4.3078*** *; **; *** refers to significance level of 10%, 5% and 1%, respectively

Pre-Crisis (2006/7) will act as reference value

The dependent variable in this analysis is the Leverage Ratio, calculated as the Total Loans divided by Total Capital. The dummy variables Pre-Crisis, Crisis and Post-Crisis act as independent variables, whereby pre-crisis is the reference value. Size is the natural logarithm of Total Assets and Growth is the percentage increase of Total Assets. Tangibility is calculated as Fixed assets as proportion of Total Assets. Further, Uniqueness takes the R&D Expenses divided by Total Assets. Return on Assets is the Net Income divided by Total Assets.

(23)

22 4.1.4 Robustness Tests

In order to verify the findings of the preceding analysis, a second regression analysis, including the loan to debt ratio has been conducted (Table 7). The results follow a similar pattern as the regression of the leverage ratio. While German firms experience a significant increase of loan to debt ratio from pre-crisis to crisis period, American firms outline a negative and insignificant development. The post-crisis dummy of the German subsample is slightly positive, but insignificant. Thus, the coefficient of the post-crisis variable is indicating a return to pre-crisis levels. The American subsample has a negative coefficient for the post-crisis dummy that is significant at the 10% level. Hence, the only difference between the regression analyses of leverage and loan to debt ratio is the change of significance from crisis to post-crisis dummy of the American subsample.

Table 7 – Regression Results for Loan to Debt Ratio – Fixed Entity Effects

Sample Period

Subsample Germany United States Germany United States

(1) (2) (3) (4) Crisis (2008/9) 0.0195*** -0.0064 Post Crisis (2010-2014) 0.0022 -0.0064* -0.0083** 0.003158 Size -0.0228 0.0162 -0.0090 -0.0323** Growth (in %) 0.0121*** -0.0012 0.0104*** 0.00772 Tangibility 0.0433* 0.0092 0.0456* 0.030807 Uniqueness 0.1149 0.2567*** 0.0905 0.2103* ROA (in %) -0.0012*** -0.0003** -0.0012*** -0.00021 Constant 0.2325 -0.0481 0.1614 0.2219 No of observtion 2070 2184 1610 1699 No of cross-sections 230 244 230 244 R-squared 0.6114 0.3993 0.6089 0.4423 F-Test 12.2215*** 5.1403*** 12.1508*** 4.6148***

*; **; *** refers to significance level of 10%, 5% and 1%, respectively Pre-Crisis (2006/7) will act as reference value

The dependent variable in this analysis is the Loan-to-Debt Ratio, calculated as the Total Loans divided by Total Liabilities. The dummy variables Crisis, Crisis and Post-Crisis act as independent variables, whereby Pre-Crisis is the reference value. Size is the natural logarithm of Total Assets and Growth is the percentage increase of Total Assets. Tangibility is calculated as Fixed assets as proportion of Total Assets. Further, Uniqueness takes the R&D Expenses divided by Total Assets. Return on Assets is the Net Income divided by Total Assets.

(24)

23 Another robustness test conducted during the course of this study is presented in Table 8. Some researchers argue that the financial crisis started already in 2007 and economic uncertainty impacted firms’ financing decisions (Vyas, 2011). Especially for the American subsample, employing the period between 2008 and 2009 as crisis dummy might be inaccurate. Hence, this robustness test incorporates a crisis dummy that takes on the value 1 between 2007 and 2008 and 0 otherwise. The results of the corresponding regression are highly comparable to earlier outcomes. Except for the post crisis dummy of the American subsample that became insignificant, all other coefficients and significance levels remain stable. Therefore, it can be argued that the chosen periods for the crisis- and post-crisis dummy are appropriate and capture the concept of the financial crisis well.

Table 8 – Regression Results for Leverage Ratio (Crisis Dummy 2007/2008) – Fixed Entity Effects

Sample Period

Subsample Germany United States

(1) (2) Crisis (2007/8) 0.0159*** -0.0055 Post Crisis (2009-2014) 0.0005 -0.0046 Size -0.0189 -0.0076 Growth (in %) 0.0091*** 0.0014 Tangibility 0.0091 0.0132 Uniqueness 0.0559 0.3236*** ROA (in %) -0.0014*** -0.0006*** Constant 0.1797** 0.0758 No of observtion 2070 2184 No of cross-sections 230 244 R-squared 0.6230 0.3728 F-Test 12.8375*** 4.5962***

*; **; *** refers to significance level of 10%, 5% and 1%, respectively Pre-Crisis acts as reference value

2006 to 2014 The dependent variable in this analysis is the Leverage Ratio, calculated as the Total Loans divided by Total Capital. The dummy variables Pre-Crisis, Crisis and Post-Crisis act as independent

(25)

24

4.2 Empirical Findings – Cash Holdings

4.2.1 Descriptive Statistics

The descriptive statistics for the analysis of cash holdings show that the US and German subsamples possess similar average cash ratios of 10.88% and 10.83%, respectively (Table 9). This is in line with Song and Lee (2012) who found that German and American firms tend to report comparable levels of cash and short-term investments. Also the control variable net working capital, which is often used as substitute for cash, outlines similar results between the firms of the two subsamples. However, dividends seem to be paid more frequently by the German firms than by their American counterparts. The descriptive statistics of the remaining control variables in Table 9 are the same as in section 4.1.1.

Table 9 – Descriptive Statistics of Subsamples

The development of cash holdings before and during the financial crisis indicates a decrease prior to the financial meltdown and an increase in the crisis and post-crisis period. Figure 5 shows that this is true for German as well as the American firms. In fact, companies located in a market based economy (United States) increase cash holdings even more than companies in a bank based economy (Germany) during financial turmoil.

Mean Median Maximum Minimum Std. Dev. Mean Median Maximum Minimum Std. Dev.

Cash 0.1083 0.0802 0.7463 0.0001 0.0973 0.1088 0.0675 0.7972 0.0001 0.1178 Size 5.7812 5.6038 8.6506 3.5888 0.9813 5.9719 6.0508 8.3112 2.8692 0.9937 Leverage 0.0791 0.0489 0.7750 0.0000 0.0918 0.0275 0.0103 0.8560 0.0000 0.0601 Growth 0.0237 0.0301 0.9773 -23.1008 0.5940 0.0140 0.0363 0.9700 -11.8336 0.5121 Tangibility 0.5144 0.5029 0.9775 0.0269 0.1822 0.5426 0.5460 0.9913 0.0443 0.2105 Uniqueness -0.0190 -0.0010 0.0000 -0.2326 0.0319 -0.0722 -0.0429 0.0000 -1.1478 0.0958 ROA 1.7207 3.2680 67.9310 -98.4180 10.7033 2.0861 4.3105 65.0230 -95.3980 13.7579 Dividends 0.6075 1.0000 1.0000 0.0000 0.4884 0.4909 0.0000 1.0000 0.0000 0.5000 NWC 0.2036 0.1902 0.8035 -0.3598 0.1543 0.1992 0.1803 0.7560 -0.1845 0.1401

(26)

25

Figure 5 – Cash Ratio

4.2.2 Univariate Results

The literature review outlined that companies are expected to increase cash ratios in response to the financial crisis. Veritably, both subsamples indicate an increase of cash holdings from crisis to post-crisis period (Table 10). However, only the means of American firms are significantly different (10% level). These findings are partially in line with Song and Lee (2012) who have shown that cash holdings increased during the Asian crisis. The data of Table 10 shows that American firms, which are assumed to be relative less bank-dependent, experience a significant increase of the average cash holdings in crisis and post-crisis period. Meanwhile, the German companies do not show any significant differences. Nevertheless, these findings are preliminary and offer mere indications.

Table 10 – T-tests and means for differences across the three periods

4.2.3 Regression Results

The results of the regression analysis of cash holdings can be found in Table 11. Hypothesis 2a tests whether cash holdings increased during the financial crisis for firms in both countries. The

Periods Subsample

Cash Ratio Means 0.1114 0.1072 0.1033 0.1045 0.1072 0.1079 0.1045 0.1127 0.1114 0.1079 0.1033 0.1127

t-stat

*,** & *** refer to 10%, 5% and 1% significance level, respectively.

Cash Ratio is compried of cash and short-term investments divided by total assets

Pre-Crisis to Post-Crisis

Germany United States Germany United States Germany United States

Pre Crisis to Crisis Crisis to Post Crisis

-1.8277*

(27)

26 corresponding results can be found in first column of Table 11. It includes the full sample and is related to the general development of cash ratios during the crisis. Hypothesis 2b investigates the interplay of leverage and cash holdings during the crisis. Column 2 includes the leverage dummy. Alternative measures for the impact of leverage on cash holdings are displayed in column 3 and 4, which are divided into German firms (bank-based economy) and US companies (market-based economy).

Table 11 – Regression Results for Cash Ratio – Fixed Entity Effects

Column 1 of Table 11 does in fact show an increase of cash holdings after the crisis, the effect however is insignificant. Hence, Hypothesis 2a does not receive statistical support. Thus, the findings of this research challenge the expectation of a significant increase of cash holdings during crises. When dividing the sample into German (bank-based economy) and American firms (market-based economy), the results show that American firms increase cash holdings more than their German counterparts. While the subsample of German firms has an insignificant

Germany United Sates

(1) (2) (3) (4) Crisis (2008/9) 0.0026 0.0025 0.0007 0.0050 Post Crisis (2010-2014) 0.0024 0.0023 -0.0044 0.0090*** Size -0.0306*** -0.0302*** -0.0269** -0.0383*** Growth (in %) -0.0002 -0.0002 -0.0009 0.0001 Divdends 0.0046 0.0044 0.0056 0.0020 ROA 0.0010*** 0.000991*** 0.0010*** 0.0001*** NetWorkingCapital -0.2777*** -0.2772*** -0.2432*** -0.3326*** Leverage -0.0810*** -0.0406* -0.1058*** -0.0463* Leverage Dummy -0.0121*** Constant 0.3383*** 0.3380*** 0.3191*** 0.3876*** No of observtion 4251 4251 2070 2184 No of cross-sections 474 474 230 244 R-squared 0.7321 0.7330 0.6737 0.7719 F-Test 21.4171*** 21.4611*** 15.9579*** 26.0449***

*; **; *** refers to significance level of 10%, 5% and 1%, respectively Pre-Crisis will act as reference value

Full Sample

(28)

27 and negative value for the post-crisis dummy, the subsample of firms from the US have a statistically significant and positive relationship between post-crisis dummy and cash holdings. Hence, German firms (bank-based economy), usually characterized by higher leverage ratios, do not increase cash ratios more than US firms (market-based economy). Therefore, Hypothesis 2b does not receive support.

5. Discussion

5.1 Interpretation of Findings - Capital Structure (Leverage Ratio)

In this study I tested whether capital structures evolve differently for firms in bank- and market-based countries. Precisely, I hypothesized that with the outbreak of the financial crisis, German firms decreased their leverage ratio more heavily than their US counterparts. The analysis outlined that the financial crisis was associated with a sudden decrease of German companies’ leverage. Their American counterparts were found to experience no significant fluctuations during the crisis period. These results are only partially in line with initial expectations. As expected, German firms did in fact experience a more drastic decrease of leverage. However, after the crisis the debt level of German firms normalized and returned to pre-crisis levels. Further, the US firms’ leverage fluctuations are less distinctive than predicted.

The fact that American firms experience relatively less fluctuations of leverage than German companies could be explained by the following reasons. Firstly, German banks reported very high lending rates prior to the crisis and decreased lending tremendously with the outbreak of the financial crisis (Blaes, 2011). At the same time, American financial institutions reduced the lending rates after the outbreak of the crisis. However, the leverage ratio of US firms remained constant. This could be due to the draw-down of existing credit lines (Ivashina & Scharfstein, 2010). Moreover, leverage levels of firms with strong relationships to banks (such as German firms) have been found to fluctuate more than for firms that rely on the capital market (Voutsinas & Werner, 2011). Additionally, as US firms rely on market financing to a large extent, banks and leverage ratios are rendered less important (Lee, 2012). These additional insights explain the findings in a more nuanced light and help managers to understand developments during financial crises.

(29)

28 in market-based economies, the findings diverge. Iqbal and Kume (2014) report a significant impact of the financial crisis on leverage ratios of British firms. In contrast, my study outlines no significant variation of American firms’ leverage throughout the crisis. One of the underlying reasons could be that even though the UK and US are both characterized by a strong market orientation (Alves & Francisco, 2015), some differences persist. In fact, as the United Kingdom is a member of the European Union, the economy is subject to an alignment to other European countries (Bartram & Wang, 2015). The enhanced trade intensities among member states come along with further convergence. US firms on the other hand have a strong historic relationship with the UK, but there seems to be no further alignment towards European firms, which might explain the diverging results. Moreover, the application of different databases and time periods within the studies might have caused different findings for firms located in the two market-based economies; US and UK (see Table 12).

Table 12 – Comparison between Iqbal & Kume (2014) and current study

5.2 Interpretation of Findings - Cash Holdings

Another major interest of this study was the analysis of cash holding developments for German and US firms throughout the crisis. I hypothesized that firms in both, bank- and market-based economies increase cash with the outbreak of the financial crisis. In line with my initial expectations, the findings of the empirical analysis reveal that the German and US firms increase cash holdings in the crisis and post-crisis period. However, the effect has been found to be insignificant. As firms already increased their cash holdings in the early 2000s (Barnes & Pancost, 2010) there was no “sudden” increase of cash, which might explain the relatively stable developments of cash holdings and the insignificance of results. Nevertheless, the slight

Iqbal & Kume, 2014 Current Study

Time period 2006 to 2011 2006 to 2014

Countries Germany, France & United Kingdom Germany & United States

Database Datastream Orbis

Statistical Analysis Fixed Effect Panel Data Regression Fixed Effect Panel Data Regression

Findings

Firms experience increase of leverage from pre- crisis to crisis period and return to pre-crisis levels during the post-crisis period in the United Kingdom and Germany

(30)

29 increase in cash holdings during the crisis can most likely be attributed to the accumulation of precautionary savings (Sun & Wang, 2015).

The third part of the analysis aimed to investigate the interplay of cash holdings and leverage ratio during the financial crisis. The hypothesis stated that the impact of the financial crisis on cash holdings has been more severe for firms located in Germany (bank-based economy) than for US firms (market-based economy). This is because firms located in bank-based economies are argued to possess more leverage. In contrast with my initial expectations, the findings of the empirical analysis show a different picture. While US companies increase cash holdings during crisis (insignificant) and post-crisis period (significant), their German counterparts increase cash holdings in the crisis period, but experience a decrease of cash in the post-crisis period (both insignificant). Resultantly, firms in a market-based economy increase cash more severely than companies located in bank-based economies. The findings seem to highlight the fact thatfirms with higher bank debt face higher interest payments and use excess cash to repay loans (Harford et al., 2014). Another potential explanation for the results of US firms might be that lower leverage ratios are associated with more intense increases of cash holdings during crises (Yujun et al., 2011).

5.3 Implications

(31)

30

6. Conclusion

The severe consequences of the financial crisis in 2007/2008 for firms also impacted their financial management. My investigation put emphasis on the financial crisis’ impact on firms’ capital structures and cash holdings in a bank-based (Germany) and a market-based economy (United States). The findings reveal that German and American firms experience different developments of their capital structure throughout the crisis. German companies take on excessive debt prior to the crisis and experience a sudden decrease in leverage with the outbreak of the financial crisis. Contrastingly, American firms do not experience any significant fluctuations of leverage throughout the crisis. The analysis regarding cash holdings revealed that even though the findings have been insignificant, firms in Germany as well as the US increase cash holdings during the financial crisis. Additionally, companies in the market-based economy increased their cash ratios more heavily than firms located in a bank-based economy. Despite the interesting insights, this research does not come without limitations. The scope of the analysis is characterized by a limited period. Data prior to 2006 could have offered a better and more extensive understanding of how leverage- and cash ratios developed prior to the financial crisis. Additionally, the study is limited by the lack of information on alternative funding sources in both countries. German firms might have access to global capital markets, while firms from the United States may take advantage of bank funding worldwide. Thirdly, the study did not incorporate small- and medium-sized corporations, which may react differently to a shock such as the financial crisis. Lastly, a statistical comparison between countries might come with some deficiencies, as advanced statistical assessment is necessary. Further research can be advised to take market characteristics, such as political or legal differences, into consideration. This could clarify the effects of the financial crisis on capital structures and cash holdings even more. Besides, incorporating smaller firms in the sample potentially offers greater insights into the transmission of economic shocks. This is because SMEs usually have less access to alternative funding sources and should experience fiercer developments. Moreover, future investigations should consider whether supply or demand for cash and leverage determine the development of the discussed concepts during the financial crisis.

(32)

(33)

32

References

Akbar, S., Rehman, S., & Ormrod, P. (2013). The impact of recent financial shocks on the financing and investment policies of UK private firms. International Review of Financial Analysis, 26(4), 59-70.

Alves, P., & Francisco, P. (2015). The impact of institutional environment on the capital structure of firms during recent financial crises. The Quarterly Review of Economics and Finance, 57, 129-146.

Antoniou, A., Guney, Y., & Paudyal, K. (2008). The Determinants of Capital Structure: Capital Market-Oriented versus Bank-Oriented Institutions. Journal of Financial and Quantitative Analysis, 43(1), 59-92.

Arindam, B., & Anupam, D. (2014). Determinants of Corporate Financial Performance Relating to Capital Structure Decisions in Indian Iron and Steel Industry: An Empirical Study. Paradigm, 18(1), 35-50.

Bancel, F., & Mittoo, U. (2004). The Determinants of Capital Structure Choice: A Survey of European Firms. Financial Management, 33(Winter), 103-132.

Barnes, M., & Pancost, N. (2010). Internal Sources of Finance and the Great Recession. Working Paper Series (Federal Reserve Bank of Boston), 10(15), 1-36.

Bartram, S., & Wang, Y. (2015). European financial market dependence: An industry analysis. Journal of Banking & Finance, 59, 146-163.

Bates, T. W., Kahle, K. M., & Stulz, R. M. (2009) Why Do U.S. Firms Hold So Much More Cash than They Used to? Journal of Finance, 64, 1985–2021.

Blaes, B. (2011). Bank-related loan supply factors during the crisis: An analysis based on the German bank lending survey. Deutsche Bundesbank: Economic Studies, (31), 1-48. Calingaert, M. (2010). The Special Relationship - Economic and Business Aspects: American

Perspective. Current Politics & Economics of Europe, 21(2/3), 271-284.

(34)

33 De Jong, A., Kabir, R., & Nguyen, T. (2008). Capital structure around the world: The roles of firm- and country-specific determinants. Journal of Banking & Finance, 32(9), 1954-1969.

Demirgüç-Kunt, A., & Maksimovic, V. (1999). Funding growth in bank-based and market-based financial systems: Evidence from firm-level data. Journal of Financial Economics, 54, 337-363.

Fosberg, R. H. 2012. Capital structure and financial crisis. Journal of Finance and Accountancy, 11, 46-52.

Frank, M., & Goyal, V. (2003). Testing the Pecking Order Theory of Capital Structure. Journal of Financial Economics, 67(2), 217-248.

Harford, J., Klasa, S., & Maxwell, W. (2014). Refinancing Risk and Cash Holdings. The Journal of Finance, 69(3), 975-1012.

Hugonnier, J., Malamud, S., & Morellec, E. (2014). Capital Supply Uncertainty, Cash Holdings, and Investment. Review of Financial Studies, 28(2), 391-445.

Hunstad, M. (2010). How much liquidity to hold? An autoregressive conditional heteroskedastic model. Journal of Corporate Treasury Management, 3(2), 117-126. Iqbal, A., & Kume, O. (2014). Impact of Financial Crisis on Firms’ Capital Structure in UK,

France, and Germany. Multinational Finance Journal, Forthcoming

Ivashina, V., & Scharfstein, D. (2010). Bank lending during the financial crisis of 2008. Journal of Financial Economics, 97(3), 319-338.

Kahle, K., & Stulz, R. (2013). Access to capital, investment, and the financial crisis. Journal of Financial Economics, 110(2), 280-299.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., and Vishny, R (2000). Investor Protection and Corporate Governance, Journal of Financial Economics, 58, 3-28.

Lee, B. (2012). Bank-based and market-based financial systems: Time-series evidence. Pacific-Basin Finance Journal, 20(2), 173-197.

(35)

34 Lim, Y. (2003). Sources of Corporate Financing and Economic Crisis in Korea: A

Micro-evidence. NBER Working Paper No. 9575. Cambridge, Mass: National Bureau of Economic Research.

Masulis, R. (1983). The Impact of Capital Structure Change on Firm Value: Some Estimates. The Journal of Finance, 38(1), 107-126.

Mikkelson, W. H., & Partch, M. M. (2003) Do Persistent Large Cash Reserves Hinder Performance? Journal of Financial and Quantitative Analysis, 38, 275–294.

Modigliani, F., & Miller, M. (1958). The Cost of Capital, Corporation Finance and the Theory of Investment. American Economic Review, 48(3), 261-298.

Morellec, E., Nikolov, B., & Schürhoff, N. (2012). Corporate Governance and Capital Structure Dynamics. The Journal of Finance, 67(3), 803-848.

Pattani, A., Vera, G., and Wackett, J. (2011). Going public: UK companies’ use of capital markets. Bank of England Quarterly Bulletin 51(4): 319-330.

Pramuk, J. (2015, July 15). Icahn blames BlackRock for bond market volatility. CNBC. Retrived from: http://www.cnbc.com/2015/07/15/

Rajan, R., & Zingales, L. (1995). What Do We Know About Capital Structure? Some Evidence from International Data. Journal of Finance, 50(5), 1421-1460.

Song, K., & Lee, Y. (2012). Long-Term Effects of a Financial Crisis: Evidence from Cash Holdings of East Asian Firms. Journal of Financial and Quantitative Analysis, 47(3), 617-641.

Sun, Z., & Wang, Y. (2015). Corporate precautionary savings: Evidence from the recent financial crisis. The Quarterly Review of Economics and Finance, 56, 175-186.

Vitols, S. (2005). Changes in Germany's Bank-Based Financial System: Implications for corporate governance. Corporate Governance: An International Review Corporate Governance, 13(3), 386-396.

(36)

35 Vyas, D. (2011). The Timeliness of Accounting Write-Downs by U.S. Financial Institutions During the Financial Crisis of 2007-2008. Journal of Accounting Research, 49(3), 823-860.

Wehinger, G. (2012). Bank deleveraging, the move from bank to market-based financing, and SME financing. OECD Journal: Financial Market Trends, 2012(1), 65-79.

Yazdanfar, D., & Öhman, P. (2015). Debt financing and firm performance: An empirical study based on Swedish data. The Journal of Risk Finance, 16(1), 102-118.

(37)

36

Appendix

Appendix 1 – Business Cash Holdings (Forbes 25th March,2015)

Appendix 2 – Vitols (2005, page 388), Comparative statistics on the German Japanese and US financial system, mid 1990s

Appendix 3 – Sample Search Strategy

# Description of Search Step Step Result Search Result

1 Listed/Unlisted companies: publicly listed companies only 65,041 64,203

2 World Region/Country/Region in Country: Germany and US 28,618,968 11,408

3 Type of Entities: Industrial Companies 161,086,434 8,013

Sub-total 8,013

4 Data on Loans unavailable -7,522 491

5 Data on Cash Holdings unavailable -6 485

6 Problems with data for Leverage -11 474

The Impact of the Financial Crisis on Firms’ Capital Structure and Cash Holdings in Germany and the United States