Master Thesis
MSc International Economics and Business
Corporate Governance and Bank Performance:
a study of European banks during the financial crisis of 2007-2009
Maureen Hovens, s1746946, m.l.hovens@student.rug.nl University of Groningen, Faculty of Economics and Business
1
Acknowledgements
Before you lies the final product of my master study International Economics and Business at the University of Groningen. Corporate governance has always been an interest of me, especially since I also study Corporate law. When I had to decide on a topic, it was not that hard to find something relating to both the financial crisis and corporate governance. The question: “Why did some banks perform better during the financial crisis?”, posited by Beltratti and Stulz (2011) got stuck in my head and it was not hard to write on this subject because I was really interested in this topic. Incorporating the possible effects of legal institutions on performance was also very fascinating and it was interesting to see the differences between countries.
Hereby I would like to thank a number of people who helped me with my thesis. First of all, I would like to thank my supervisor, prof. dr. H. Van Ees, for his support, advice, feedback, and empathy for personal circumstances. Although I was not able to meet the deadline for this thesis, I was allowed to hand in my thesis later and I was provided with useful feedback for the final version. Next I would like to thank Miroslava Karadzhova for allowing me to use some of the data she collected. Finally I would like to thank my friends for their continued support.
2
Abstract: This research investigates the role of corporate governance, the specialization of the
bank, and legal environment on bank performance and risk-taking behavior of banks during the financial crisis. Using a sample of European banks, the results indicate that banks with more independent boards had a higher level of risk. Firm size had a negative impact on buy-and-hold stock returns during the crisis. Banks with institutional ownership took more risk during the crisis period. Commercial banks, savings, and cooperative banks performed all poorly. No significant results with regard to anti-director rights were found.
3
Table of contents
1. Introduction ... 4
2. Literature review and hypothesis development ... 7
2.1 Board composition ... 7
2.2 Ownership structure ... 9
2.3 Country characteristics and bank specialization ... 12
3. Data and Methods... 14
3.1 Data collection, sample and measures... 14
3.2 Methods and working data set ... 15
3.3 Theoretical model ... 17
3.4 Methodology ... 17
4. Empirical Results ... 19
4.1 Summary statistics... 19
4.2 Model estimation and discussion ... 20
4.3 Additional robustness tests ... 22
4
“Negligence and profusion,
therefore, must always prevail, more or less, in the management of the affairs of such a company.”
Adam Smith (1776)
1. Introduction
In the economics literature, bank relationships are viewed as a way to overcome problems of credit rationing and adverse selection due to asymmetric information (Stiglitz and Weiss, 1981). Banks can achieve economies of scale in the monitoring process and firms will have an incentive to invest in their own reputation (Diamond, 1984). The corporate governance of banks, however, differs from governance of nonfinancial firms. Banks have more stakeholders compared to other firms, namely depositors, debt holders, and the government. Moreover, the stakeholders are very large since over 90 percent of the balance sheet of banks is debt (Mehran, Morrison, and Shapiro, 2011). Yet it are the shareholders who control the firm and they respond to their own incentives, leading to possible conflicts with other stakeholders. The quality of corporate governance of banks is a much-debated topic and the approach to regulation and supervision is constantly changing. See for example the potentially transformation of Basel 3 into Basel 4, since recent developments in the banking market have suggested that even stricter rules should be applied. The financial sector is exposed to failures on a global scale despite accounting rules and the disclosure of risks to shareholders and the board of directors. Internationalization and financial innovation in banking has increased the possibilities for contagion and systemic risk, as evidenced by the subprime mortgage crisis that started at the beginning of August 2007 in the United States and spread to the European Union and the rest of the world.
5 better performance and lower risk-taking during the crisis because banks could adopt a different governance system to prevent crises from occurring or avoid poor performance during such a crisis. Studying performance differences of European banks during the crisis gives us a better insight in why the crisis evolved as it did and allows us to have a better understanding of the characteristics of banks that make the banking system more sensitive to systemic risk, because we can investigate which determinants of good governance caused some banks to perform better than others. It is also possible to explore which country characteristics and what type of specialization caused banks to perform worse.
The OECD concluded that failures and weaknesses in corporate governance of financial services companies attributed to an important extent to the financial crisis (Kirkpatrick, 2009). In their final document, “Principles for enhancing corporate governance”, the Basel Committee on Banking Supervision has issued a final set of principles to enhance governance practices at banking organizations and considers the role of the board of directors a key area of focus. Also, in the literature it is argued that banks that suffered large losses during the crisis were banks with poor governance that engaged in excessive risk-taking (Diamond and Rajan, 2009; Bebchuck and Spamann, 2010). Quite on the contrary, Beltratti and Stulz (2011) find no evidence for the hypothesis that the financial crisis resulted from excessive risk-taking by banks with poor governance. More specifically, Beltratti and Stulz find that banks with a shareholder-friendly board performed worse during the crisis. Some authors find that governance of banks is relevant in a broad context, because well-functioning banks promote economic growth (Caprio et al., 2007) and the risk-taking behavior of banks affects financial fragility (Laeven and Levine, 2009). Whereas others find no link between the governance of banks and their performance (Beltratti and Stulz, 2011). Research documenting which aspects of governance contributed to the crisis is thus not conclusive and the role bank governance played during the crisis is not yet unraveled. Therefore, a new approach is required to resolve this gap. In this paper I will not only look at variables of good governance, but also at country characteristics of the country the bank operates in and the type of specialization of the bank.
6 hazard implications. Banks do not simply conform to the concept of the firm used in agency theory because banks have a more complex structure of information asymmetry arising from the presence of regulation. Regulation alters the parameters of the agency relationship by introducing the regulator as a third party making the agency problem more complex. Regulation limits the power of markets to discipline the owners and managers of banks and this has consequences for the governance in banks (Ciancanelli and Gonzalez, 2010). Throughout the paper, it should be borne in mind that regulation affects the banking sector and this alters the principal-agent problem in banks. Likewise, because banks play a distinctive role in providing payment and settlement services, and the transformation of liquidity, maturity, and denomination, problems with governance of banks are more severe and come with higher costs than other corporate firms.
This paper wants to add to the emerging body of research that attempts to identify why some European banks performed better than other European banks during the financial meltdown. More specifically, I will focus on risk-taking behavior, the board composition, the ownership structure (i.e. corporate governance structure) of banks, and country characteristics. The main research question is:
“Why did some European banks perform better than other European banks during the financial crisis?”
The main question will be answered by looking at how board composition and ownership structure affect bank performance and risk-taking. By looking at these corporate governance characteristics, a part of the answer to the main question will be given because independent boards and large and institutional owners are thought to have a negative impact on performance via increased risk-taking. Moreover, a good legal environment of a country should benefit banks resulting in higher stock returns and commercial banks are expected to have a higher level of risk as compared to savings and cooperative banks, resulting in lower stock returns. Thus, the main research question will be accompanied by the following sub questions:
1. How does board composition influence bank performance and bank risk-taking? 2. How does ownership structure influence bank performance and bank risk-taking? 3. How do country characteristics and the specialization of the bank affect bank
7 This paper contributes to the literature in several ways. First of all, it takes an institutional organizational perspective by comparing performance differences and risk-taking incentives of banks during the crisis and looks at country differences and specialization differences of banks with bank governance. It is interesting to take this approach because country and specialization differences have not been investigated thus far and they could both have a positive impact on performance during the crisis. If banks in certain countries performed better, other countries could change their rule of law and anti-director rights. Secondly, it will investigate the impact of the board composition and ownership structure of banks. Using a sample of 80 European banks during the period 2007-2009 this study finds that that the board structure of banks is important to bank performance and to the risk-taking behavior of banks. The results show that banks with more independent boards performed worse during the crisis period and engaged in more risk-taking behavior. Also banks with institutional ownership engaged in more risky behavior. Thirdly, the type of specialization of a bank is investigated. There is no significant indication of performance differences between the three types of banks.
The paper proceeds as follows. Section two contains a review of the relevant literature, with a set of testable hypotheses and the theoretical model. Section three covers the data and methodology, with a specification of the data sources and final sample, a specification of the measures for the variables in the theoretical model and an overview of the applied econometric techniques and its requirements. Section four encompasses the empirical results with a description of the data and model, the estimation, several robustness checks, and a discussion of the results. Finally, section five concludes the paper with an answer to the research questions and a discussion of the implications of the results.
2. Literature review and hypothesis development
In this section relevant literature will be discussed leading to the development of the hypotheses. Table 2.1 (see Appendix) offers a non-exclusive overview of previous literature on the link between bank governance, performance, and risk.
2.1 Board composition
8 the CEO. The role of boards takes on special importance the context of banks because of limited competition, regulation, and higher informational asymmetries due to the complexity of the banking industry (Andres and Vallelado, 2008). The role of directors is more significant in banks because of their specific knowledge of the banking business and their ability to monitor managers. The board also takes charge of links with the regulator to avoid conflicts of interest and advises managers on strategy identification and implementation (De Andres and Vallelado, 2008). Bank lending activities are often both opaque as vague and this makes it difficult for stakeholders such as shareholders and debt holders to monitor bank managers (Levine, 2004). The board of directors solves this gap and is thus of crucial importance to the governance of banks.
Firms choose their board composition, i.e. the level of independent boards members relative to dependent board members to balance monitoring and advising needs. The composition of the board should be a reliable proxy of how well the board can process information and advice as well as monitor the bank in the best interest of the shareholders. Independent board members, or outside directors, do not have family, financial or business ties with management and are generally thought to be more effective in monitoring the CEO (Hermalin and Weisbach, 2003). The relationship between board independence and performance, however, is not that clear-cut. Adams and Mehran (2012), using a proxy for Tobin’s Q, find that board independence is not related to performance of bank holding companies. Erkens, Hung, and Matos (2012) report a negative relationship between firms with more independent boards and stock returns in a study of financial firms. Cooper (2009) shows that insider representation on the board has a positive influence on bank performance. Andres and Vallelado (2008) find an inverted U-shaped relation between bank performance and the proportion of outside directors in the board, indicating that incorporation of outsiders improves bank value up to a certain boundary. When a high proportion of outside directors over inside directors is reached, Tobin’s Q falls.
9 argue that performance should increase when firms have more independent boards because of the better monitoring capacity of such boards. Although the corporate governance literature generally views independent boards as beneficial for performance, this might not hold for the banking sector because independent directors will not have sufficient expertise to monitor the CEO (Adams, 2009). This was exacerbated at the onset of the crisis when CEOs had to make important decisions about the bank’s strategy, having consequences for the bank’s performance. Thus, the first hypothesis follows:
Hypothesis 1a: Banks with more independent boards took more risk during the crisis period.
A possible explanation of why bank performance is worse during the crisis for banks with more independent boards is that external monitoring by such boards encouraged managers to increase shareholder returns by taking more risk prior to the crisis (Erkens, Hung, and Matos, 2012). What has not been tested, however, is whether banks with more independent board members took more risk during the crisis period. Without the financial expertise that dependent board members have, independent board members could have adopted risky strategies to prevent stock returns to go down. With their fiduciary duty to shareholders, independent directors understand the residual nature of equity claims and could be more willing to increase risk. Moreover, with their unfamiliarity with complex financial mechanisms they could be more willing to let the bank participate in more risk-taking activities even in crisis periods to prevent stock returns from falling down, but not overseeing the possible consequences of the risk-taking activities. As a result, independent board members could have continued their strategy of excessive risk-taking leading to reduced performance. The following hypothesis is made:
Hypothesis 1b: Banks with more independent boards performed worse during the crisis period.
2.2 Ownership structure
10 make managers more amenable to their demands and their incentive to monitor is greater because they cannot always sell the shares of underperforming firms (Shleifer and Vishney, 1997).
Ownership identity is an issue that has been well developed in the literature on non-financial firms, but empirical evidence based on non-non-financial firms do not show conclusive evidence on the effect of institutional investors on firm performance (Acker and Athanassakos, 2003). Barry, Lepetit, and Tarazi (2009) argue that on the one hand institutional investors can exert greater control and better supervision for reasons of economies of scale. On the other hand they argue that a possible alliance between institutional investors and managers could result in lower firm value as insider interests take priority. Moreover, because institutional investors typically have a diversified portfolio of investments, they are likely to have lower incentives to exert control. An institutional owner, as a shareholder, is only concerned about expected return of an investment and indifferent to the riskiness of it (Barry, Lepetit, and Tarazi, 2009). The analysis of Erkens, Hung, and Matos (2012) shows that financial firms with greater institutional ownership experienced worse stock returns in the crisis period. They give as a potential explanation that institutional shareholders encouraged managers to increase shareholder returns by engaging in more risk-taking behavior before the crisis. Consistent with this finding, Beltratti and Stulz (2011) find that banks with more shareholder-friendly boards performed significantly worse than other banks during the crisis.
Literature about securitization also largely agrees that credit risk transfer techniques undermine financial stability and underlines the misalignment of incentives between investors and banks in the process of securitization (Casu et al., 2013). Misaligned incentives, or conflicts of interest, refer to the situation where some participants in the securitization process followed solely their own interests and engaged in behavior that is not in the interest of others. These misalignments and conflicts are generally thought to have contributed to the loss of investor confidence in securitization. Securitization did not enhance financial stability since the imperative to expand assets decreased lending standards (Shin, 2009). Institutional investors could have influenced banks to maximize shareholder value by exposing banks to risks through securitization activities that manifested themselves during the crisis and had a negative impact on bank performance. Banks with more institutional ownership were possibly making decisions that increased value for shareholders before the crisis, but resulted in unforeseen poor performance during the crisis. Institutional investors who have significant voting power can thus shape the nature of bank risk-taking by increasing risk during the financial crisis. The
11
Hypothesis 2a: Banks with higher institutional ownership took more risk during the crisis
period.
Hypothesis 2b: Banks with higher institutional ownership performed worse during the crisis
period.
A firm is widely held when ownership of capital is dispersed among small shareholders but control is concentrated in the hands of managers (Berle and Means, 1932). An important agency problem appears when a firm is not widely held. Contrary to the situation where shares are widely held, a large investor has the ability to expropriate resources of the firm at the cost of minority investors and stakeholders in the firm (Shleifer and Vishney, 1997). The cost of concentrated ownership becomes more significant when the large investor has an incentive to force the firm to take too much risk because as an equity holder he will share in the benefits, whereas costs of failure will be borne by other investors (Jensen and Meckling, 1976). Expropriation of minority shareholders by the controlling shareholders can take a variety of forms, such as transfer pricing, asset stripping, investor dilution, or simply stealing profits (La Porta et al., 2000). There is no reason to assume that these problems disappear when it comes to banks. When controlling shareholders use the profits of the bank to benefit themselves instead of returning the money to the bank or the other investors this will become value-reducing for the bank.
Contrary to the expropriation argument mentioned above, there is no evidence that financial firms with a large shareholder have encountered worse stock returns (Erkens, Hung, and Matos, 2012). Caprio, Laeven, and Levine (2007) find that banks are generally not widely held and that larger cash-flow rights of the controlling shareholder improves performance. There is evidence, however, that greater cash flow rights by a large controlling owner are associated with more bank risk (Laeven and Levine, 2009). This is consistent with the view that large shareholders have more power and incentives to increase bank risk-taking than small shareholders (Demsetz and Lehn, 1985; Jensen and Meckling, 1976). Therefore it can be assumed that banks with a large shareholder took more risk than banks without a large shareholder during the crisis period. In the build-up to the crisis large shareholders might have taken actions that they believed the market would welcome but ex post these actions turned out to be negative for bank performance. Consequently, banks with dispersed ownership ought to perform better during the financial crisis.
12
Hypothesis 3a: Banks with more widely held shares took more risk during the crisis period. Hypothesis 3b: Banks with more widely held shares performed worse during the crisis period.
2.3 Country characteristics and bank specialization
Countries differ in the way they protect investors and could therefore have an influence on bank governance and performance during the crisis. The rule of law plays a crucial role in the control of self-dealing and the legal protection of minority shareholders against expropriation by managers or controlling shareholders (Djankov et al., 2008). Minority shareholder rights will mostly be determined at the country level because the possibility for a bank to obtain and maintain good governance depends on the state investor protection. Doidge et al. (2007) find that it is more costly to implement high-quality governance in less developed countries and that incentives for firms to adopt firm-level governance measures are higher with higher state investor state protection. Differences in legal protection across countries shape the ability of insiders to expropriate outsiders and has been studied in several contexts and crises (Johnson, Boone, Breach and Friedman, 2000a; Glaeser, Johnson, and Shleifer, 2001; La Porta et al., 2003).
Yet, these country characteristics and the role they might have played on bank performance during the recent financial crisis have not been investigated thus far for European countries and banks. It would be interesting to see what the effect of country-level rules is with respect to the 15 European countries investigated in this study, because all countries were affected by the crisis. Research indicates that strong shareholder protection laws enhance the governance of nonfinancial firms and increase firm valuations (La Porta et al., 2012), but protection of investor rights plays an important role for bank governance and performance as well, especially during crises. Although not everyone agrees that shareholder protection will hinder expropriation, because of the complex and opaque nature of banks, I argue that banks countries with better protection of minority shareholders had better stock returns. Banks in countries with good investor protection might have taken less risk before the crisis because strict regulations have prevented them from doing so. Furthermore, these banks could have fared better during the crisis because investors were less afraid of expropriation because of the heavy regulations and sound governance banks had to adhere to. This might have caused these banks to outperform banks in countries with a weaker legal systems and investor rights because they encountered shareholders who were less willing to finance these banks.
13 returns and the equation on risk-taking incentives. In Table 2.2 (see Appendix) an overview is given of the indexes used. It is hypothesized that banks in countries that have a better legal environment for investors are likely to have better governance and, therefore, might have performed better during the crisis. Hence, the following hypotheses can be posited:
Hypothesis 4a: Banks in countries with better legal protection took less risk during the crisis. Hypothesis 4b: Banks in countries with better legal protection performed better during the
crisis.
The type of specialization of a bank could have had an impact on risk-taking behavior and performance of banks. In this paper, the level of risk and performance of three types of specialization of banks are being compared (cooperative, savings, and commercial banks). Historically, cooperative and savings banks have played a big role in almost all European banking systems (Bülbül, Schmidt, and Schüwer, 2013). Savings and cooperative banks are generally regional banks. The main difference between savings and a cooperative banks is, besides that savings banks have a focus on savings and savings mobilization, their legal structure. A cooperative bank is a self-governed private organization and the owners and equity holders are called members. These members are the main clients of the banks and each member has only one vote in the annual meeting. Because members cannot sell their shares at a higher market price, their incentives to monitor the board are weak. Moreover, they cannot exert pressure on the board because they cannot accumulate voting rights. On the other hand, there is no expropriation of small members and incentives to incur high risks to achieve high profits are also low. Commercial banks provide services such as accepting deposits, making loans, and offering several investment products and are more profit-oriented than cooperative and savings banks. Commercial banks make their profits by taking small, short-term, relatively liquid deposits and transforming these into larger, longer maturity loans.
Cooperative and savings banks were viewed as outdated and old-fashioned for the past years, but the financial crisis has changed this former critical view (Bülbül, Schmidt, and Schüwer, 2013). Contrary to large shareholder-oriented commercial banks, cooperative and savings banks incurred much less risk during the crisis. Commercial banks had too high a focus on profit maximization and financial sophistication. Generating too much risk and with a higher exposure to systemic risk, it is likely that commercial banks suffered more during the financial crisis than cooperative and savings banks, leading to the last two hypotheses.
14
Hypothesis 5a: Cooperative and savings banks took less risk during the financial crisis. Hypothesis 5b: Cooperative and savings banks performed better during the financial crisis.
3. Data and Methods
This section presents the data collection method, sample formation, and measures used and will elaborate on the employed statistical tests.
3.1 Data collection, sample and measures
The data used in this research was taken from Bankscope available at Bureau van Dijk Electronic Publishing. It contains comprehensive information on banks across the globe about financials, interim data, ratings and rating reports, country risk and country finance reports, stock data, directors, contracts, original filings, and detailed bank structures suitable for this particular research. To be included in the sample, a bank had to be active in the European Union (15) and Switzerland, with as specialization being a commercial, savings, or cooperative bank.1 The obtained result is a sample of 350 banks from 16 European countries, covering the period 2006-2009. Because Bankscope does not contain information about buy-and-hold stock returns, this information is obtained from Datastream. Not all banks are covered in that database and therefore 270 banks had to be dropped from the sample to obtain a balanced panel.
The director data is being sourced from BoardEx. BoardEx is the leading database on board composition of publicly listed firms and contains profiles on approximately 500.000 business leaders globally. In line with Doidge et al. (2007) a variable measuring the quality of investor protection is calculated by multiplying the anti-director rights index constructed by Djankov et al. (2008) by the rule of law index reported by La Porta et al. (1998). As pointed out by Spamann (2010) there are several inconsistencies in the coding of La Porta et al. (1998) and Djankov et al. (2008) because corporate charter provisions and corporate practice are not taken into account. I will therefore use his index as a robustness check.
The final sample consists of 80 banks from 15 European countries in the period 2005-2009. See Table 3.1 (see Appendix) for a summary of the data collection method, and Table 3.2 for a list of the banks and the countries used in this research.
15
3.2 Methods and working data set 3.2.1. Dependent variable
To test the hypotheses multiple regressions will be estimated. The first dependent variable is bank performance and in the second regression risk-taking will be the dependent variable. Buy-and-hold stock returns are commonly used as a proxy for bank performance. Following Beltratti and Stulz (2011) and Erkens, Hung and Matos (2012) the primary measure of bank performance is the buy-and-hold stock returns for the period 2007-2009, although I add the period after 2008 because the consequences of the crisis continued after that period. Buy-and-hold stock returns are defined as shares that an investors holds for the long-term regardless of short-term market fluctuations. The buy-and-hold stock returns are calculated as percentage returns on the assets on a yearly basis:
.
As a robustness test I will use stock returns of quarterly data from the period 2007Q3-2008Q4. To measure bank risk the Z-score of each bank will be calculated. The Z-score equals the return on assets plus the capital asset ratio divided by the standard deviation of asset returns. Laeven and Levine (2009) define insolvency as a state in which losses surmount equity ( , where π is profits). The probability of insolvency is expressed as probability ( ), where ROA is the return on assets ( ⁄ ) and CAR is the capital assets ratio ( ⁄ ). The inverse of the probability of insolvency thus equals
⁄ , where is the standard deviation of ROA. The inverse of the probability of insolvency is defined as the Z-score and is often referred to as a measure of stability or distance to default. A higher Z-score implies a lower probability of failure.
3.2.2. Independent variables
Board independence is defined as the percentage of independent directors. Using the SD ratio from BoardEx directors are independent if they are non-executives.2 Institutional ownership is measured as the percentage of shares held by institutional investors and obtained via annual reports of the banks. Widely held shares are measured as a dummy variable equal to 1 if the bank does not have a large owner with voting rights greater than 25%. The 25% cutoff is usually large enough for complete control in the bank and the data are obtained via Bankscope. For the country characteristics I will use the anti-director rights variable from
16 Djankov et al. (2008) as a measure of shareholder rights. The anti-self-dealing index of Djankov et al. (2008) is constructed as an index of the strength of minority shareholder protection against self-dealing by the controlling shareholder. The index reflects the hurdles that the controlling shareholder must jump in order to get away with a self-dealing transaction. The higher the hurdles, the higher the index. The rule of law is a proxy for law enforcement and will be obtained from La Porta et al. (1998). Following Durnev and Kim (2005) and Doidge et al. (2007) I will define ADRI as the product of the anti-director rights variable and the rule of law. Possible specialization effects are taken into account by using two specialization dummy variables, representing the three types of specialization (Spec1=cooperative bank and Spec2=savings bank). The dummy variable equals 1 if the bank has that type of specialization and 0 otherwise. Commercial banks are the benchmark case.
3.2.3. Control variables
17
3.3 Theoretical model
To test the hypotheses two regression equations will be used as expressed below. The theoretical model of this study is specified in equation 1 and 2, where Table 2 explains the concepts of the employed variables and their respective measurements.
(1) ̅̅̅̅ where , , and (2) ̅̅̅̅ where , , and Table 3.3
Definitions of variables and concepts of the theoretical model
3.4 Methodology
This paper uses generalized least squares regression (GLS) to make inferences about the panel data to overcome the problem of dependence between the random variables. To investigate the determinants of bank performance and bank risk-taking multiple regressions are
Dependent variables Abbreviation Proxy
Bank performance Bank performance Buy-and-hold stock returns Risk-taking Risk-taking Z-score
Corporate governance variables
Board independence BOARD ind % of non-executive directors
Institutional ownership INST own % of shares owned by institutional investors
Widely held WIDE held Dummy variable equal to 1 if a firm does not have a large owner with voting rights > 25% Country and specialization variables
Country characteristics Ln(ADRI) Natural log of Anti-director rights variable (Spamann, 2010) and rule of law (La Porta, 1998) Bank specialization SPEC Dummy variables indicating whether the bank is a savings, cooperative or commercial bank Control variables
Firm size Ln(FIRM size) Natural log of total assets in euros of 2006 Board size Ln(BOARD size) Natural log of the number of directors in the board
18 estimated. The data is strongly balanced, indicating that each panel contains the same number of observations. The sample is short and wide.
Before making inferences about the data, however, a few tests needs to be checked. The Breusch-Pagan test is performed and tabulated in Tables 3.4. and 3.5 (see Appendix) to test for the presence of random effects. The test leads to the conclusion that random effects are present for both bank performance and risk-taking. More importantly, the Hausman test is performed to see whether a fixed effects or a random effects model should be applied. In Tables 3.4 and 3.5 (see Appendix) this test is performed for both performance and risk-taking. The Hausman test is used to check for correlation between the error component and the regressors in the random effects model. The test compares the coefficient estimates from the random effects model to those from the fixed effects model. The high p-value in Table 3.4 leads to accept the hypothesis that the coefficient estimates are equal to one another. The random effects estimator is consistent for risk-taking. The small p-value of performance in Table 3.5 leads to reject the hypothesis that the coefficient estimates are equal to one another. The difference suggests that the random effects estimator is inconsistent for bank performance. This difference is probably due to a misspecification. Because the model contains many time invariant variables, like board independence, board size, ADRI, and widely held shares, the Hausman-Taylor model is used as an alternative.
19 an indicator of a banks’ financial strength, a positive relation with bank performance would expected.
A random effects estimator can estimate the effects of variables that are individually time-invariant, such as many of the governance variables that do not change over time. Furthermore, the random effects estimator has greater precision in estimating the data because it uses information on how changes in the dependent variable across different individuals could be attributable to the different independent values for those individuals and the GLS estimator has a smaller variance than the least squares estimator. However, the GLS does not control for unobserved firm heterogeneity.
There are a few assumptions of the random effects model and they are as follows. First of all, the combined error term has zero mean . Secondly, it has a constant, homoscedastic variance:
. There are also several correlations that can be considered. Errors for individual are correlated, the errors for different individuals are uncorrelated, errors are uncorrelated with the independent variables and the random effects are uncorrelated with the independent variables.
4. Empirical Results 4.1 Summary statistics
An overview of bank characteristics before and after the crisis is given in Table 4.1 (see Appendix). Table 4.1 compares common bank characteristics (i.e. buy-and-hold stock returns, total assets, equity to total assets, total capital ratio, return on average assets (ROAA), cost to income ratio, loans, subordinated debts, loan loss provisions, net interest income, net income to average total equity, and leverage) of the sample banks right before the crisis and during the crisis. The mean of both stock returns and ROAA declined during the crisis years. The cost-to-income ratio increased, indicating that on average banks were less efficient during the crisis. The amount of subordinated debt also increased. Subordinated debt holders are less likely to be paid off in a bank bailout, so implicit government support plays no role here. Loan loss provisions almost tripled, leverage remained roughly the same, and return on equity dropped as expected.
20 returns it is shown that Ireland performed worst with a mean stock returns of -32.78. Austria was the only country with positive stock returns during the crisis. The average mean of board independence is relatively low in this sample compared to other studies. See for example Erkens, Hung, and Matos (2009) who report an average of 82% but this might be due to the small sample. Firm size is comparable in all countries, and the same holds for board size and leverage. It is interesting to note that Austria had the lowest stock returns for 2006, whereas Germany had the highest returns. Austria performed best during the crisis where Germany had mean stock returns below the average.
4.2 Model estimation and discussion
The random effects estimates for risk-taking behavior of banks are tabulated in Table 4.3. Because the random effects estimator utilizes both between and within individual variation, it is possible to estimate the effects of board independence, widely held shares, institutional ownership, ADRI, and the specialization of the bank on buy-and-hold stock returns and risk-taking behavior. The problem of exact collinearity between these variables and individual dummy variables no longer exists with random effects, as opposed to the fixed effects model.
Table 4.3
Relation between risk-taking and corporate governance
(1) (2) (3) (4) (5) (6) (7) Board independence -0.751** -1.022*** (0.382) (0.339) Institutional ownership -.0115* -0.016** (0.007) (0.007) Widely held 0.087 -0.077 (0.143) (0.323) Cooperative -0.554*** -1.123** (0.202) (0.492) Savings -0.143* -0.094 (0.077) (0.126) ADRI 0.485 0.135 (0.302) (0.495) Firm size 0.0589 .0979 0.164 0.173 0.168 0.168 0.081 (0.0601) .1390 (0.105) 0.1075 (0.107) (0.106) (0.096) Board size -0.492*** -.3524 -0.395* -0.299 -0.385* -0.335* -0.494** (0.184) .2851 0.214 (0.204) (0.209) (0.202) (0.219) Leverage -153.509*** -158.6*** -161.18*** -163.94*** -161.51*** -165.16*** -156.32*** (4.453) (6.655) (6.666) (6.848) (6.811) (6.868) (7.531)
21
(0.027) .04027 (0.040) (0.041) (0.041) (0.039) (0.041)
N 147 91 204 204 204 204 91
Adj-R² 0.700 0.594 0.711 0.714 0.712 0.716 0.596
Wald chi 2814.29 1620.75 4790.38 4468.15 4633.78 5134.48 2198.31 Robust standard errors in brackets
*** if P>|z| <0.01, ** if P>|z| <0.05, * if P>|z| <0.1
Table 4.3 reports the relation between risk and corporate governance. Board independence and institutional ownership are both significant at 1% and 5% levels, respectively. Thus, board independence and institutional ownership are associated with greater risk as predicted in hypothesis 1a and 2a. These results are consistent with the view that institutional owners and independent board members encourage managers to take more risks. Institutional owners, despite their greater monitoring capabilities, influenced banks thus to take excessive risk to increase shareholder value. Independent board members are also thought to have better monitoring capabilities but let banks increase risk during the crisis. This could be because of their lack of financial expertise. The implication of the results are that banks should rely on less independent boards during periods of recession and institutional ownership should be avoided to reduce risk. Cooperative as well as savings banks increased risk-taking behavior during the crisis. This is not what was expected in hypothesis 5a because it was expected that these banks would have a lower level of risk. The coefficient of the anti-directors rights index (ADRI) is positive, as expected, but insignificant at conventional levels. The control variables indicate that board size, leverage, and Tier 1 capital ratio increased risk-taking and are all significant. The Wald chi is high and the Wald chi test is used to test the hypothesis that at least one of the predictors' regression coefficient is not equal to zero.
22 performance, or that these variables did not significantly change performance of banks during the crisis for the sample used in this research.
Table 4.4
Hausman-Taylor model
Time variant, exogenous
Institutional ownership -0.443 (0.812)
Time variant, endogenous
Tier 1 capital ratio -20.106*** (3.561)
Leverage 707.181
(883.503)
Firm size -87.655**
(40.820)
Time invariant, exogenous
Widely held 155.672 (117.086) ADRI 47.247 (99.457) Cooperative 161.578 (156.026) Savings 2.106 (178.065) Board size 87.961 (110.971)
Time invariant, endogenous
Board independence -260.117
(584.703)
N 84
Wald chi 85.11
Standard errors in brackets
*** if P>|z| <0.01, ** if P>|z| <0.05, * if P>|z| <0.1
4.3 Additional robustness tests
23 sample with 80 banks.3 Because the availability of buy-and-hold stock returns greatly reduced the sample, using ROAA as a proxy for performance, the larger sample of 206 banks can be used (see Table 3.1, Appendix for the sample collection procedure). Panel B of Table 4.3 reports the results using the Hausman-Taylor model. In the small sample, this alternative proxy for bank performance makes board independence and institutional ownership significant at 10% and 5% levels, respectively but does not alter their signs. ADRI becomes positive in the full model and even significant in column (6) when ADRI is added without the governance variables. The control variable leverage becomes highly significant throughout the model. Including more banks with ROAA as a proxy in Panel B of Table 4.3 does not make the model more significant. Overall, the results hold when using a different proxy for bank performance. As pointed out by Spamann (2010) there are some inconsistencies in the coding of the anti-director index of Djankov et al. (2008), therefore his index is used as a robustness test. The results for firm performance and risk-taking are reported in Table 4.4 (see Appendix), Panel A for bank performance, and Panel B for risk-taking. ADRI remains insignificant and there are no remarkable differences in the other coefficients.
Furthermore, a different definition of the financial crisis is used. Beltratti and Stulz (2011) focus on bank returns from the middle of 2007 to the end of 2008 and call this the crisis period. Therefore this alternative specification of the crisis period is used and shown in Table 4.5 (see Appendix). Using quarterly data, the number of observations increased because buy-and-hold stock returns for the period 2007Q3-2008Q4 are used. This different time framework does not alter the signs of most coefficients, although the coefficient of cooperative banks becomes significantly positive.
5. Conclusions
The aim of this thesis was to examine why some European banks performed better during the financial crisis. The results indicate that corporate governance did not matter that much for bank performance during the crisis. Beltratti and Stulz (2011) who found that more shareholder-friendly banks experienced worse stock returns during the crisis, showed and argued that poor corporate governance was not one of the main causes of the financial crisis. Maybe there has been put too much emphasis on corporate governance and bank performance during the crisis and further research should focus more on other aspects that contributed to the performance of banks such as short-term capital market funding and leverage levels.
24 The size of the bank, measured as the natural log of total assets, did have a negative impact on performance. This suggests that larger banks were worse off during the crisis. Tier 1 capital ratio also had a significant negative influence on bank performance. This result is a bit ambiguous because banks with more capital should suffer less from a debt overhang problem and should have more flexibility to respond to adverse shocks. In the literature a positive relation has been found (Beltratti and Stulz, 2011; Erkens, Hung, and Matos, 2012). Why banks with a higher Tier 1 capital ratio performed worse is thus not clear and warrants further research. The composition of the board had an influence on bank risk-taking during the crisis. Independent boards had a higher level of risk. The ownership structure of banks also had an influence on risk-taking in the sense that banks with institutional ownership took more risk during the crisis. Whether a bank had widely held shares or a larger owner did not have a significant influence on neither bank performance nor bank risk-taking. This finding is consistent with the findings of Erkens, Hung, and Matos (2012).
The legal environment of a country, measured as the anti-director rights index times the rule of law, suggests to improve bank performance and reduce risk-taking behavior. However, the results are not significant at conventional levels. The type of specialization for a bank does not seem to matter that much for performance and risk-taking during the crisis. In the Hausman-Taylor model, both cooperative and savings banks had better stock returns than commercial banks, as hypothesized in hypotheses 5a and 5b, but the obtained results were not significant. Since cooperative and savings banks are more old-fashioned and more conservative than commercial banks, further research should investigate the consequences of the type of specialization of a bank. Also other types of specialization should be included.
25 Limitations of this research are that the main sample is relatively small as is the period studied before the crisis. Further research could use data on a larger period before the crisis started, to account for delayed effects. Performance of earlier years could be included as a control variable to control for performance of earlier years. Performance before the crisis can give better insight in why a bank performed worse or better during the crisis. Findings of this paper could not be generalized to banks in other countries and banks that operate on an international field since only banks in European countries are studied. Further research could include banks from the United States, since the crisis originated in the United States. Also the regulatory framework of the banks studied is not taken into account and future research should consider the stance of the regulators. The non-significant findings may be due to the sample that has been collected, delayed effects or because the effect of corporate governance on performance and risk-taking is indirect. It could also be the case that corporate governance did not cause worse stock returns during the crisis, as shown by Beltratti and Stulz (2011).
26
6. References
Adams, R. 2009. Governance and the Financial Crisis. European corporate governance
institute. Working paper Nº 248/2009.
Adams, R. & Mehran, H. 2012. Bank Board structure and performance: Evidence for large bank holding companies. Journal of Financial Intermediation, 21: 243-267.
Adams, R., Hermalin, B. & Weisbach, M. 2008. The role of boards of directors in corporate governance: a conceptual framework and survey. NBER working paper series.
Working paper Nº 14486.
Adkins, L. & Carter Hill, R. 2011.Using Stata for Principles of Econometrics, fourth edition. Wiley.
Aebi, V., Sabato, G. & Schmid, M. 2012. Risk management, corporate governance, and bank performance in the financial crisis. Journal of Banking & Finance, 36: 3213-3226. Aggarwal, R. Erel, I., Stulz, R. & Williamson, R. 2009. Differences in governance practices between U.S. and foreign firms: measurement, causes, and consequences. Review of
Financial Studies, 22: 3131-3169.
Andres, de P. & Vallelado, E. 2008. Corporate governance in banking: The role of the board of directors. Journal of Banking & Finance, 32: 2570-2580.
Barry, T., Lepetit, L. & Tarazi, A. 2009. Ownership structure and risk in publicly held and
privately owned banks.
Beltratti, A. & Stulz, R. 2011. The credit crisis around the globe: Why did some banks perform
better?
Berger, A., Clarke, G., Cull, R., Klapper, L. & Udell, G. 2005. Corporate governance and bank performance: A joint analysis of the static, selection, and dynamic effects of domestic, foreign, and state ownership. Journal of Banking & Finance, 29: 2179-2221.
Bhagat, S. & Bolton, B. 2008. Corporate governance and firm performance. Journal of
Corporate Finance, 14: 257-273.
Bhagat, S., Bolton, B. & Romano, R. 2008, The promise and peril of corporate governance indices. Columbia Law Review, 108: 1803-1882.
Booth, J.R., Cornett, M.M. & Theranian, H. 2002. Boards of directors, ownership, and regulation. Journal of Banking & Finance, 26(10): 1973-1996.
Brunnermeier, M.K. 2009. Deciphering the Liquidity and Credit Crunch 2007-2008. Journal
of Economic Perspectives, 23(1): 77-100.
Caprio, G., Laeven, L. & Levine, R. 2007. Governance and bank valuation. Journal of
27 Carter Hill, R., Griffiths, W. & Lim, G. 2011. Principles of Econometrics, fourth edition. Wiley, p. 538-570.
Casu, B., Clare, A., Sarkisyan, A. & Thomas, S. 2013. Securitization and Bank Performance.
Journal of Money, Credit and Banking, 45(8) 1617-1658.
Cheng, I., Hong, H. & Scheinkman, J.A. 2010. Yesterday’s heroes: compensation and creative risk-taking. NBER working paper series. Working paper Nº 16176.
Ciancanelli, P. & Gonzalez, J.A.R. 2010. Corporate Governance in Banking: A Conceptual Framework. European Financial Management Association Conference, Athens, June. Cooper, E.W. 2009. Monitoring and governance of private banks. The Quarterly Review of
Economics and Finance, 49: 253-264.
Diamond, D.W. 1984. Financial Intermediation and Delegated Monitoring. Review of
Economic Studies, 51(3): 393-414.
Djankov, S., La Porta, R., Lopez-de-Silanes, F. & Shleifer, A. 2008. The law and economics of self-dealing. Journal of Financial Economics, 88(3): 430-465.
Doidge, C., Karolyi, A., & Stulz, M. 2007. Why do countries matter so much for corporate governance? Journal of Financial Economics, 86(1): 1-39.
Durnev, A. & Kim, H. 2005. To steal or not to steal: firm attributes, legal environment, and valuation. Journal of Finance, 60: 1461-1493.
Ellul, A. & Yerramilli, V. 2010. Stronger Risk Controls, Lower Risk: Evidence from U.S. Bank Holding Companies. AXA Working Paper Series Nº 1. Discussion paper Nº 646. Erel, I., Nadauld, T. & Stulz, R.M. 2012. Why Did U.S. Banks Invest in Highly-Rated
Securitization Tranches? Fisher College of Business Working Paper Series.
Erkens, D.H., Hung, M. & Matos, P. 2012. Corporate Governance in the 2007-2008 Financial Crisis: Evidence from Financial Institutions Worldwide. Journal of Corporate
Finance, 18(2): 389-411.
Fahlenbrach, R. & Stulz, R.M. 2011. Bank CEO incentives and the credit crisis. Journal of
Financial Economics, 99: 11-26.
Fahlenbrach, R., Prilmeier, R. & Stulz, R.M. 2012. This Time is the Same: Using Bank Performance in 1998 to Explain Bank Performance during the Recent Financial Crisis. The Journal of Finance, LXVII(6): 2139-2185.
Faleye, O. & Krishnan, K. 2013. When is Bank Governance Important? Board Effectiveness, Banking Industry Distress, and Borrower Risk. Electronic copy available at:
28 Fernandes, N. & Fitch, E. 2009. Does Financial Experience Help Banks during Credit Crises? Electronic copy available at:
https://www.cemfi.es/ftp/pdf/papers/wshop/WPM$60DE.pdf.
Gillan, S.L. & Starks, L.T. The Evolution of Shareholder Activism in the United States. Electronic copy available at: http://ssrn.com/abstract=959670.
Glaeser, E., Johnson, S. & Shleifer, A. 2001. Coase versus the Coasians. Quarterly Journal of
Economics, 116: 853-899.
Hermalin, B.E. & Weisbach, M.S. 2001. Boards of directors as an endogenously determined institution: A survey of the economic literature. NBER working paper series. Working
paper Nº 8161.
Jensen, M.C. & Meckling, W.H. 1976. Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. Journal of Financial Economics, 3(4): 305-360. John, K., Litov, L. & Yeung, B. 2008. Corporate Governance and Risk-Taking. The Journal of
Finance, LXIII(4): 1679-1728.
Johnson, S., Boone, P. Brach, A. & Friedman, E. 2000. Corporate Governance in the Asian Financial Crisis. Journal of Financial Economics, 58: 141-186.
Kirkpatrick, G. 2009. The Corporate Governance Lessons from the Financial Crisis. OECD
Report, Paris, France.
Kumar, N. & Singh, J.P. 2013. Global Financial Crisis: Corporate Governance Failures and Lessons. Journal of Finance, Accounting and Management, 4(1): 21-34.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A. & Vishny, R. 2000. Investor protection and corporate governance. Journal of Financial Economics, 58: 3-27.
Laeven, L. & Levine, R. 2009. Bank governance, regulation and risk taking. Journal of
Financial Economics, 93: 259-275.
Lemmon, M. & Lins, K. 2001. Ownership Structure, Corporate Governance, And Firm Value: Evidence from the East Asian Financial Crisis. William Davidson Working Paper,
Number 393.
Levine, R. 2004. The Corporate Governance of Banks: A Concise Discussion of Concepts and
Evidence. World Bank Policy Research Working Paper, 3404.
Macey, J.R. & Miller, G.P. 1995. Corporate Governance and Commercial Banking: A Comparative Examination of Germany, Japan, and the United States. Stanford Law
Review, 48(1): 73-112.
29 Mehran, H., Morrison, A. & Shapiro, J. 2011. Corporate Governance and Banks: What Have We Learned from the Financial Crisis? Federal Reserve Bank of New York Staff
Reports. Staff Report Nº 502.
Minton, B., Taillard, J.P. & Williamson, R. 2010. Do Independence and Financial Expertise of the Board Matter for Risk Taking and Performance? Fisher College of Business
Working Paper Series.
Mitton, T. 2002. A cross-firm analysis of the impact of corporate governance on the East Asian financial crisis. Journal of Financial Economics, 64: 215-241.
OECD. 2009. Corporate Governance and the Financial Crisis: Key Findings and Main
Messages.
Parrino, R., Poteshman, A.M., & Weisbach, M.S. 2005. Measuring investment distortions when risk-averse managers decide whether to undertake risky projects. Financial
Management, 34: 21-60.
Pathan, S. 2009. Strong boards, CEO Power and bank risk-taking. Journal of Banking &
Finance, 33: 1340-1350.
Pathan, S. & Skully, M. 2010. Endogenously structured boards of directors in banks. Journal
of Banking & Finance, 34: 1590-1606.
Saunders, A. & Cornett, M.M. 2006. Financial Institutions Management: A Risk Management
Approach. McGraw-Hill Companies Inc., New York.
Shleifer, A. & Vishny, R.W. 1997. A Survey of Corporate Governance. The Journal of Finance, LII(2): 737-783.
Shin, H. 2009. Securitization and Financial Stability. Economic Journal, 119: 309-332. Spamann, H. 2010. “The Antidirector Rights Index” Revisited. The Review of Financial
Studies, 23(2): 467-486.
Stefanescu, C.A. 2011. Do corporate governance “actors” features affect banks’ value? – Evidence from Romania. Procedia Social and Behavioral Sciences, 24: 1311-1321. Stiglitz, J.E. & Weiss, A. 1981. Credit Rationing in Markets with Imperfect Information. The
American Economic Review, 71(3): 393-410.
30
7. Appendix Table 2.1
Author(s) Period Performance indicatorsRisk indicators Empirical finding(s) Econometric model/Methodology
Adams , Hermalin, Weisbach (2010) n/a n/a n/a Overview article. n/a
Adams and M ehran (2012) 1965-1999 Tobin's Q n/a Board independence is not related to performance, board size is positively related to performance. OLS Aebi, Sabato, and Schmid (2012) 2006-2008 Buy-and-hold returns, n/a Banks in which CRO reports directly to the board of directors and not the CEO performed better. OLS
ROE, ROA No significant or negative relationship between performance and corporate governance.
Andres and Vallelado (2008) 1995-2005 Tobin's Q, ROA, SM R n/a Inverted U-shaped relation between bank performance and board size, and between the proportion two step system estimator of non-executive directors and performance.
Barry, Lepetit, and Tarazi (2009) 1999-2005 ROA, ROE SDROA, SDREO, M _LLP, A shift in equity from institutional investors to either individuals/families or banking institutions OLS Z-score, ZP Score implies a decrease in asset risk and default risk, but no change in profitability.
Beltratti and Stulz (2010) 2006-2008 Buy-and-hold stock Idiosyncratic volatility, log Performance of banks with shareholder-friendly boards performed significantly worse during the M ultiple OLS regressions returns of distant to default, tangible equity crisis and were not less risky before the crisis.
Caprio, Laeven, Levine (2007) 2000-2001 Tobin's Q, market value n/a Larger cash-flow rights by the controlling owner boost bank performance and weak shareholder OLS to book value of equity protection laws lower bank valuations.
Cheng, Hong, and Scheinkman (2010) 1992-2008 Cumulative excess Beta, volatility of stock, Residual pay is correlated with price-based risk-taking measures and compensation and risk-taking OLS returns correlation of stock returns with ABX incentives are not corelated to governance variable.
Cooper (2009) 2006 ROA n/a Insider representation on the board has a positive influence on both director and executive OLS compensation in commercial banks.
Ellul & Yerramilli (2010) 20101-2008 Stock performance Risk M anagement Index BHCs with a high RM I in 2006 generally fared better in terms of operating performance and had Simultaneous equation model (RM I) lower downside risk during the crisis.
Erkens, Hung, and M atos (2012) 2007-2008 Buy-and-hold stock EDP (expected default prob), Stock Firms with high institutional ownership and more independent boards performed worse. OLS and Tobit regression returns return volatility
Fernandes and Fitch (2009) 2006-2007 Stock returns n/a Financial experience of the banks' outside directors is positively related to the financial Probit regressions institutions' stock return performance during the credit crisis.
Fahlenbrach and Stulz (2011) 2007-2008 Buy-and-hold stock Equity risk sensitivity No evidence that CEOs whose incentives were less well aligned with the interests of shareholders OLS
returns performed worse.
Fahlenbrach and Stulz (2012) 2007-2008 Buy-and-hold stock Equity beta Stock market performance of banks in the recent crisis is positively correlated with that of banks in Probit regressions
returns the 1998 crisis.
Faleye and Krishnan (2010) 1994-2006 n/a Non-investment grade rating by S&P Banks with more effective boards are less likely to lend to risky commercial borrowers. Probit regressions
Laeven and Levine (2009) 2001 Revenue growth Z-score Bank risk taking varies positively with the comparative power of shareholders within the corporate OLS and simultaneous equitations system governance structure of each bank.
M agalheas, Gutiérrez, and Tribó (2008) 2000-2005 Risk-adjusted ROA Volatility of earnings, Z-score Cubic relationship between ownership concentration and bank performance, and a cubic Generalized M ethod of M oments relationship between ownership concentration and bank risk.
M ehran, M orrison, and Shapiro (2011) n/a n/a n/a Overview article. n/a
M inton, Taillard, and Williamson (2010) 2001-2008 Tobin's Q, cumulative Standard deviation stock returns, Financial expertise is associated with more risk-taking and higher performance prior to the crisis OLS stock returns received TARP funds and lower performance during the crisis.
31
Table 2.2
Country-level governance variables
ADRI (DLLS, 2008) ADRI (Spamann, 2010) Rule of Law (La Porta et al., 1998)
Austria 2.5 4 10 Belgium 3 2 10 Denmark 4 4 10 Finland 3.5 4 10 France 3.5 5 8.98 Germany 3.5 4 9.23 Greece 2 3 6.18 Ireland 5 4 7.8 Italy 2 4 8.33 Netherlands 2.5 4 10 Portugal 2.5 4 8.68 Spain 5 6 7.8 Sweden 3.5 4 10 Switserland 3 3 10 U.K. 5 5 8.57
ln(Antidirector rights index) ln(Antidirector rights index)* Austria 3.218875825 3.688879454 Belgium 3.401197382 2.995732274 Denmark 3.688879454 3.688879454 Finland 3.555348061 3.688879454 France 3.447762851 3.804437795 Germany 3.475222017 3.60875341 Greece 2.514465452 2.91993056 Ireland 3.663561646 3.440418095 Italy 2.813010637 3.506157817 Netherlands 3.218875825 3.688879454 Portugal 3.077312261 3.54731589 Spain 3.663561646 3.845883203 Sweden 3.555348061 3.688879454 Switserland 3.401197382 3.401197382 U.K. 3.757705645 3.757705645
32
Table 3.1
Sample selection
N of banks dropped Remaining banks
Bankscope European banks 350
Less:
Banks not covered in BoardEx 144 206
Banks not covered in Datastream 126 80
Final sample 80
33
Table 3.4 Risk-taking
Panel A: Breusch and Pagan Lagrangian multiplier test for random effects Risk[id,t] = Xb + u[id] + e[id,t]
Estimated results Var sd = sqrt(Var) Risk 11.5359 3.396453 Estimated results 0.4354723 0.6599033 u 0.140237 0.3744823 Test: Var(u) = 0 chibar2(01) = 0.11 Prob>chibar2 = 0.3727
Breusch and Pagan Lagrangian multiplier tests for the presence of random effects. If the null hypothesis is true, there are no random effects. Using leads to rejecting the null hypothesis. Random effects are present.
Panel B: Hausman test
Coefficients (b) (B) (b-B) sqrt(diag(V_b-V_B) fe re Difference S.E. Institutional -0.0130287 -0.0160285 0.0029998 0.016466 Firm size -0.4623643 0.0810319 -0.5433962 0.923114 Leverage -151.5495 -156.3205 4.771 18.736
Tier 1 capital ratio -0.1717609 -0.1358501 -0.0359108 0.0661969 b = consistent under H0 and Ha; obtained from xtreg B = inconsistent under Ha; efficient under H0; obtained from xtreg Test: H0: difference in coefficients not systematic
Chi2(4) = (b-B)’[V_b-V_B)^(-1)](b-B)
= 2.33
Prob>chi2 = 0.6753
34
Table 3.5
Bank performance
Panel A: Breusch and Pagan Lagrangian multiplier test for random effects
STOCK[id,t] = Xb + u[id] + e[id,t]
Estimated results Var sd = sqrt(Var) Stock 1591.242 39.89037 Estimated results 749.2909 27.37318 u 168.312 12.97351 Test: Var(u) = 0 chibar2(01) = 0.06 Prob>chibar2 = 0.4069
Breusch and Pagan Lagrangian multiplier tests for the presence of random effects. If the null hypothesis is true, there are no random effects. Using leads to rejecting the null hypothesis. Random effects are present.
Panel B: Hausman test
Coefficients (b) (B) (b-B) sqrt(diag(V_b-V_B) fe re Difference S.E. Institutional -.4426537 -.3763306 -.0663231 .6714909 Firm size -87.65456 -3.325908 -84.32865 38.27284 Leverage 707.1807 601.8153 105.3654 770.097
Tier 1 capital ratio -20.10576 -14.57558 -5.530177 2.601838 b = consistent under H0 and Ha; obtained from xtreg B = inconsistent under Ha; efficient under H0; obtained from xtreg Test: H0: difference in coefficients not systematic
Chi2(4) = (b-B)’[V_b-V_B)^(-1)](b-B)
= 90.30
Prob>chi2 = 0.0000
35 caused by correlation between the random effects and some of the explanatory variables, the Hausman-Taylor estimator will be applied to the random effects model.
Table 3.6
Panel A: Jarque-Bera test for normality for risk-taking
sktest ResidualsRisk
Skewness/Kurtosis tests for Normality
---joint---Variable Obs Pr(Skewness) Pr(Kurtosis) adj chi2(2) Prob>chi2 Residuals Risk 84 0.1015 0.7166 2.91 0.2338
The p-value 0.2338 is greater than the significant threshold level of of , and therefore, the null hypothesis of normality is not rejected. Not enough evidence to view the residuals as something other than normally distributed.
Panel B: Jarque-Bera test for normality for performance
sktest ResidualsPerformance
Skewness/Kurtosis tests for Normality
---joint---Variable Obs Pr(Skewness) Pr(Kurtosis) adj chi2(2) Prob>chi2 Residuals Performance 84 0.8824 0.3128 1.07 0.5867
