Basel II Accord and Loose Coupling: Risk
Taking in European Commercial Banks, 2002-2010
DIRK BOERSMA
2 Abstract
Data on the Basel II Accord capital requirements of 121 commercial banks in Europe from 2002 to 2010 were analysed to test to what extent the capital requirements as set out by the Basel II Accord – measured by Tier 1 capital ratio - lower the risk appetite of European banks. The results show that the capital requirements as set by the Basel II Accord lower the risk appetite of European commercial banks, but leave room for bank loose-coupling from the stated capital requirements and are therefore able to increase their risk appetite in their actual working structure.
Key-words: Basel II Accord, Tier 1 capital ratio, loose-coupling, risk appetite. Author: Dirk Boersma is a Master student IB&M, Rijksuniversiteit Groningen Thesis supervisor: R. Kozhikode
3 The capabilities and the very existence of the Basel II Accord are heavily debated in the academic field. Moosa (2010) criticizes and points to the Basel II Accord as the causality of the global financial crisis: “Regulatory capital requirements are supposed to protect financial institutions from insolvency, but the crisis has shown no relation whatsoever between capital ratios and the incidence and severity of losses”. Furthermore, Doerig (2003) points to the fact that implementing formal rules to diminish bank risk taking emerge in an unstable financial system, since risk creates value for banks. In the academic field, a “wedge” is recognized between regulators and shareholders (e.g. Blum, 1999; Hakenes & Schnabel, 2010). Regulators perceive excess capital as a desirable buffer (Bank of International Settlements, 2006), whereas in the eyes of the shareholder, excess capital is inactive capital that is reducing return on equity which needs to be put to work. Banks try to fill up this gap by combining the Basel II capital requirements and the shareholders’ wants. More specific, a grey area is implemented by creating a gap between the formal structure of a bank and their actual work activities; on the one hand keeping up with the Basel II Accord, and on the other aligning with the efficiency criteria stated by shareholders and thereby not lowering their risk appetite as proposed by the Accord.
As a solution to this agency problem, banks could start ‘loose-coupling’ (Meyer & Rowan, 1977) there performance, disguising their liquidity status more and more as a solution to keep pace with the Basel II Accord, but also with their shareholders. This loosely-coupling state of banks makes their performance opaque and the effectiveness of the Basel II Accord
4 loose-coupling (Meyer & Rowan, 1977), wherein a bank brings in capital on different terms1. This does not harm the Basel II Accord terms, but does harm the specific bank’ risk appetite. Therefore, this phenomenon of ceremonial behavior could undermine the Basel II Accord, and especially the reason of existence of this Accord. In the U.S. Bear Stearns and Northern Rock were both adequately capitalized, but were able to increase their risk appetite to such levels making it impossible to save them (Moosa, 2010). In Europe French/Belgium bank Dexia collapsed as the first major European bank, however French and Belgian government saved Dexia from total collapse. Previous banks all complied to the standards of the Basel II Accord, however all failed to deal effectively with their risk appetite and were unable to manage their risky assets well. This is an example of how bank behavior adhering to capital rules still may propel the risk appetite to extreme levels.
This raises the question if the Basel II Accord is really effective enough in curtailing risk taking among banks or do banks have more leeway in risk taking than expected? It is important in seeking an answer to first understand how organizations behave under institutional change and then to understand the possible response to such behavior.
Institutional theory provides rich insights into the behavior of organizations as a reaction to institutional change (Ingram & Clay, 2000; Meyer & Rowan, 1977; Scott, 2004a). Aseries of banking theory studies focused on the impact of capital rules on risk taking and draw
significant insights from financial theory and corporate political theory to explain the consequences of the implementation of the capital rules (Blum, 1999; Hakenes & Schnabel, 2010; Saunders et al., 1990). Following the lead of many earlier studies of the diffusion of structural models and procedures, this study is designed to use insights from financial theory and new institutional theory to identify firm-level heterogeneity in their propensity and capacity to loose-couple from the Basel II Accord among commercial banks in Europe.
1
5 This study tested the theoretical proposition that banks show ceremonial behavior when adhering to the Basel II Accord but simultaneously increase their risk appetite on different terms. The research used the Tier 1 capital ratio as presented in the Basel II Accord – that is, the banks’ core capital level expressed as a percentage of their debts – as a proxy for
conformity to the Basel II Accord, since the Tier 1 capital ratio is the only element common to all countries’ banking systems and disclosed on all banks’ annual reports (Bank for
International Settlements, 2006). The research of panel data of 121 commercial banks spread over 14 countries in Europe between 2002 and 2010 provides strong evidence for the
theoretical predictions. This study has implications for both financial theory and loose coupling practices. Practically, the results suggest that banks respond to institutional changes by loose coupling their formal structure from their actual work structure.
Background
The banking puzzle under Basel II
The Basel II Accord2 is adopted by the European Union in 2008 in order to strengthen the stability of the financial system. The Basel Committee predicted a decrease in the probability of default, resulting in a more stable banking industry constructing a stable basis for the economy (Bank for International Settlements, 2004a). Major institutions and governments support the sound argument of increasing stability on the financial market and among banks (Bank for International Settlements, 2006). However, previous scholars have discussed the impact of the Basel II accord, doubting the validity of the argument of stabilizing and
strengthening the position of banks (Hakenes & Schnabel, 2011). Moreover, several scholars argued that the implementation of the Basel II accord will increase the risk probability of a bank and thereby increasing the probability of bank default (Moosa, 2010; Blum, 1999).
2
6 Another point of critique on the Accord is the lack of regulations regarding liquidity and leverage, identified by Moosa (2010) and Doerig (2003). The effects of high leverage are potentially disastrous; adverse market movements may bankrupt a bank in no time since these adverse movements are amplified by the high level of leverage. A bank may comply with the Basel II Accord capital requirements and have the sufficient capital ratio on its balance, though when confronted with a bank run due to its illiquid position, it might collapse just as fast. One of the effects of imposing capital regulation is the reduction of a bank’s profits, resulting in a smaller incentive for banks to circumvent risk taking (Blum, 1999; VanHoose, 2007). More formally, Blum (1999) reasoned that due to the capital requirements, banks have to attain the same results (profits) with limited means, resulting in a trade-off between
solvency risk and returns as proposed in financial theory. Therefore, the sounding argument of Blum is that capital requirements lead to an increase in the risk appetite of banks.
On the other hand, Hawkins and Turner (2000) justify capital regulations by stating that it makes banks more careful and provides an incentive to avoid excessive risk for fear of significant losses. However, Caruana and Narain (2008) stated that, once fully implemented, the Basel II Accord will go a long way toward addressing many of the weaknesses in bank risk management and its supervision that lie at the root of the turmoil in mature financial markets. Although, they do acknowledge that capital requirements cannot prevent banks from making mistakes. Furthermore, Caruana and Narain (2008) addressed that the capital
regulations should not be seen as a substitute for a banks’ own risk assessment and
7 because of their status, influence and incumbency. Low-status parties are more indifferent to conformity and the eventual consequences of non-conforming.
Loose-coupling as a solution to Basel II
Organizations confronted with institutional change pursue their own interests in line with the choice-within-constraints new institutionalism, explaining the differences in performance of organizations when adhering to institutional rules (Ingram and Clay, 2000). Additionally, Oliver (1991) recognized that organizations not simply respond to institutional demands with passive compliance, but have several strategic responses including agreement. However, compromise, avoidance, defiance, and manipulation are included as well.
According to institutional theory, institutional change is driven by organizational competition, which forces organizations to continually develop skills and knowledge to survive (North, 1993). Mezias (1995) suggested that compliance to institutional rules varies as a function of the resources allocated to implementation. Meyer and Rowan (1977) stated that the formal organizational structure of an organization is a reflection of the rationalized institutional rules, split from its actual working activities. Hence, due to decoupling work activities from its formal structure, conformity to the Basel II Accord may give a false comfortable feeling of security and stability while banks are engaging in risky behavior. Moreover, conformity to institutional rules does not align with the efficiency criteria of an organization, thus if an organization would promote its efficiency than they could not comply with the institutional rules. Therefore, to maintain ceremonial conformity to the amended Basel Accord, organizations start decoupling their formal structures from the uncertainties of technical activities, also called “loose-coupling”: building gaps between their formal
8 systems (Scott, 2004a). Due to this ceremonial conformity structural elements are often
loosely linked to each other and start becoming opaque. Since companies should adhere to institutional rules on the one hand, and practical activities on the other, Meyer and Rowan (1977) stated that a stable solution for companies is to maintain the company in a loosely coupled state and conform to the new institutional rules in a ceremonial manner. It is now important to integrate insight from research in institutional theory with those from research in financial and banking regulation theory to better understand banking behavior after
institutional change.
The impact of the Basel II Accord on bank behavior
The Basel II Accord and Bank Risk Taking
The aim of the Basel II Accord. The amended Basel Accord is in 2008 imposed in the
European Union regulation to increase the stability of the European banking sector. The Accord promotes the safety and soundness of the financial system, furthermore it includes approaches to capital adequacy that are sensitive to risk involvement in bank’s positions and activities. Lastly, it focuses on internationally active banks continuing to enhance competitive equality (Bank for International Settlements, 2001). Due to the stricter capital requirements and many other valuable improvements, banks are expected to decrease their risk appetite. In the Basel II Accord, amended minimum capital requirements are presented for credit, market and operational risk to diminish the risk appetite of commercial banks (Bank for International Settlement, 2004b). In institutional theory, organizations are recognized to be rational
systems, not simply reflecting institutional rules (Meyer & Rowan, 1977). Recent studies in institutional theory suggest that agreements are not predetermined solely by natural
9 2004a). Even if the apparent subject of the Basel II capital requirements is stability and order, attention towards conflict and changes in social structures cannot be avoided (Scott, 2004b). Furthermore, many scholars discuss the proposed impact of capital regulations on the risk taking behavior of banks (e.g. Blum, 1999; Hakenes & Schnabel, 2010). Prior research of banking capital regulations confirmed that capital regulations may increase bank risk taking, since imposing capital rules is expensive for banks and therefore decrease profits of banks (Blum, 1999). Furthermore, the baseline of the Basel II Accord is stressed in research since many banks adhering to the Basel capital requirements have defaulted yet (e.g. Moosa, 2010; Hakenes & Schnabel, 2010).
Complying. The original aim of the Basel II Accord is to promote safety and soundness of
the financial system. The minimum capital requirements, measured by the Tier 1 capital ratio as stated in the Basel II Accord, are expected to diminish the risk appetite of banks, since banks are the spill of the financial system. Banks conform to the Basel II Accord minimum capital requirements if their Tier 1 capital ratio exceeds 8 percent (Bank for International Settlements, 2004). Furthermore, the amended capital rules follow up the Basel I Accord in order to increase the stability of the financial system. Following up the relation between the minimum capital requirements and bank risk taking, conformity to the Basel II Accord is expected to decrease the risk appetite of banks (Bank for International Settlements, 2001).
Hypothesis 1. Compliance to the Basel II Accord has a negative relationship with bank risk taking.
Basel II Accord conformity and loose coupling
Bank size. Additionally, considering the industry, the size of the bank differs enormously
10 Accord: size. According to Hakenes and Schnabel (2011) the implementation impact of the Accord varies among the industry, seeming to bring smaller banks closer to default than larger ones. This variability can be related to the impact of status on conformity pressures. Philips and Zuckerman (2001) stated that those parties that enjoy a status in the mid-section, not being either high or low, are most like to conform to institutional rules. This inverted U-shaped relationship between status and conformity is explained by certain means. Those parties enjoying high status are likely to have relatively more freedom in regard to
conforming. This is a result their status, influence and incumbency. Low-status parties are more indifferent to conformity and the negative or positive consequences of non-conformity. Middle-status parties are most often the pivotal point of policy makers and therefore are under close scrutiny. Hakenes and Schnabel (2011) recognized different implications under the Basel II Accord for larger and smaller banks, leading to a more risk taking behavior by smaller banks, since they have to deal with larger costs, but still have to be competitive towards larger banks. They will offer higher returns and therefore should take on higher risk levels. This contrast effect seems to have a negative impact on smaller banks, bringing them on a weaker position opposed to larger banks. Furthermore, larger banks have greater potential to diversify its asset risk, leading to less risk taking (Saunders et al., 1990). Not to mention the information available for large bank investors. Information of different subsets hold by multiple investors, leads to limits in their diversification of their securities in the market. Moreover, larger banks are better able to generate larger amounts of information, resulting in investors holding fewer securities of smaller banks (Banz, 1981).
11
Profitability. Imposing the Basel II Accord induce high costs for banks. Due to the strict
requirements on capital and the high implementation costs of the Basel II Accord, banks have to impose larger buffers in order to comply with the standards as set in the Accord (Hakenes & Schnabel, 2011). Furthermore, larger buffers of capital should be set to comply with the obligated core capital level by the Basel Committee. In reaction to this, bank’s profits will be reduced by capital adequacy requirements (Blum, 1999). Additionally, due to the leverage effect of the capital requirements, the equity capital of the bank will raise, which in turn will make funding new equity capital expensive. Despite falling profits, shareholders require the same returns, resulting in a trade-off between risk taking and returns. Banks therefore, need to fill up the “wedge” between regulators and banks’ shareholders wherein regulators are willing to see strict capital buffers, whereas shareholders do not see any benefit in these buffers. Imposing regulatory rules combined with rigorous supervision leads to disposal of moral hazard incentives out of the banking system (Park & Peristiani, 2007).
The Basel II Accord leads to higher capital buffers at the expense of profitability, ceteris paribus, profitability weakens the negative relationship between Basel II Accord conformity and bank risk taking.
Hypothesis 3. The negative relationship between Basel II Accord conformity and bank risk taking weakens as profitability in a bank increases per unit.
Performance. The discussion thus far has treated capital regulations binding and induced
12 change is driven by organizational competition, therefore banks may seek for alternative ways to comply with the Basel II Accord. More specific, banks may decouple their formal structure from accounting and controlling activities. Furthermore, rationality plays a central role in the creation of formal structures (Ingram & Clay, 2004; Meyer & Rowan, 1977). Research suggested that if expectations fall below aspirations, organizations want to raise expected performance. Organizations with high expectations about future performance increased risk-taking, and the higher their aspirations, the more risks they took (Bromiley, 1991). Prior research confirmed that capital regulations confront banks with higher costs and thence lower performance expectations, resulting in a tradeoff between complying with capital regulations and risk-taking (Hakenes & Schnabel, 2011; Blum, 1999). Additionally, social
accountabilities in banks may lead to an increase in risk taking by banks in order to level their competitors and comply with their shareholders’ demands (Haigh, 2006; Park & Peristiani, 2007). This would impose that high performing banks may display ceremonial behavior to the Basel II Accord, while simultaneously increase their risk appetite. Therefore, high performing banks may increase their risk appetite even when they have regulatory constraints.
Hypothesis 4. The negative relationship between Basel II Accord conformity and bank risk taking weakens when the bank is high performing.
Basel II Accord and state aid. The Basel II Accord is implemented on the intercession of
both bankers and governments. The standards as set in the Basel II Accord constitute a basis for the regulation of banks by individual countries. Both banks and states pursue their
13 organizations produce private-centralized institutions. Hence, banks try to influence the state institutions (Ingram & Clay, 2000). Since commercial banks are subject to regulations and governments oversight the amended Basel Accord is imposed based on mutual trust and confidence, assuming that a breakdown of that perception would induce great risk for the financial system. However, it was not thought to be possible that failing banks be protected by guarantee schemes or safety nets (Gary et al., 2004). Furthermore, some banks play a central role in a nation’s financial system, causing major disruption to the solvency of the nation when it defaults and therefore guaranteed by the government.
Prior research demonstrated that bank’s failure diminishes when governments are involved in their activities (Baum & Oliver, 1992). Therefore, public guarantee schemes are expected to distort the banking sector. An important consequence of public guarantee schemes by the state are a banks reduction in market discipline, because creditors have lower incentives to monitor the bank’s activities if their bank is guaranteed for capital aid. Furthermore, recent research confirmed that a high degree of guarantee to competitor banks increases a bank’s risk-taking incentives (Gopp et al., 2010). In a transparent system, these competitor effects reduce the need of bail-out, since all banks are pushed towards higher risk-taking. However, in an opaque system, protected banks will reduce risk-taking, whereas competitors increase risk-taking (Hakenes & Schnabel, 2010). Hence, public governmental intervention may distort the financial system by increasing the overall risk appetite of commercial banks. If so, commercial banks receiving state guarantees, whether it is by public guarantee schemes or direct capital injections (bail-outs), may increase risk-taking while simultaneously conforming to the Basel II Accord.
Research methods
Data and sample
The hypotheses are tested using a sample of 121 European (Union) listed banks for which Bankscope had data available on the Tier 1 capital ratios. However, data was available on Tier 1 capital ratios covering the period 2006-2010. Therefore, all data on the Tier 1 capital ratios is obtained from the 121 individual banks, covering the period 2002-2010. Additionally, data on government intervention per bank is found in annual reports. All other variables are obtained from Datastream and Bankscope.
Reason for the specific European sample selection is that the US is still in transition to adopting the IFRS accounting standards which are mandatory for European countries. Also, Basel II is for many countries a voluntary agreement, but enshrined in EU law and is therefore mandatory for European Banks. For this reason and the sake of comparability this study solely focuses on European banks. The sample consists of various banks in the European banking sector; both relatively small and big, retail banks, savings banks and banks that combine these activities. Other financial institutions and players are not considered because the Basel II accord is only applicable to banks. In addition, the banks should be listed on a stock exchange for the data collection and have data available for more than halve the time period studied, this reduces potential biases of single year observations. This leads to a sample of 121 European banks3 spread over 14 countries (Appendix A9).
Considering the study-period, the period of 2002 up onto 2010 is chosen in order to give a proper overview of the impact of the Accord on the risk taking behavior of banks. In Europe, the point in time Basel II was completely and mandatorily implemented is the first of January 2008. In order to detect a change due to the implementation of the Basel II standards, a yearly time-series panel study is conducted. This test enables the researcher to compare the period
3 Of these 121 banks, 40 were multinationals such as Deutsche Bank and BNP Paribas; the rest were domestic
15 after the Basel II with the period before the implementation, detecting changes in values which could relate to a change in risk taking behavior. All covariates are measured annually, but some covariates were not available for every year, so the data were sorted into 12,844 bank-country-year observations, and the unit of analysis was a particular bank in a particular country in a particular year.
Measures
Bank risk taking. The key dependent variable in this study is bank risk taking, measured
16 outstanding as a percentage of total assets. The higher the loan to assets ratio indicates a bank is loaned up and its liquidity is low. Additionally, the loans to assets ratio measures the composition of bank assets on the balance sheet (Karels & Weber, 2004). This fairly simple and clear bank risk measure, measures the coincident risk taking by banks. The higher this ratio, the more risk is taken by bank and the higher the probability to default.
Basel II Accord: Tier 1 Capital. Key independent variable in this study is the measure used as a proxy for the Basel II Accord. The measure used as a proxy for the Basel II Accord is the core capital level of banks measured by the Tier 1 capital ratio (T1CR) - stating eight per cent as the minimum total capital ratio by the Basel II Accord under the Internal Rating Based approach (Bank for International Settlements, 2006). The Tier 1 ratio is the core capital of banks, expressed as a percentage of its debts. The Bank for International Settlements (2006) stated that the key element of a bank’s capital on which the main emphasis should be placed is equity capital and the disclosed reserves of a bank. Furthermore, the Tier 1 capital ratio is the only element common to all countries’ banking systems and disclosed on all banks’ annual reports. Measuring the impact of the Basel II Accord is therefore done by the Tier 1 capital ratios for all banks in the sample.
According to the Basel Committee, all European banks should adhere to the standards set in the Basel II Accord and must have a Tier 1 capital ratio above the eight percent. In order to measure hypothesis 1, a dummy Conformity measures if banks comply with the Basel II standards as set by the Basel Committee. Conformity is coded 1 if banks have a Tier 1 capital ratio greater than eight percent.
Profitability. Second independent variable is profitability, measured by either the return on
17 indicating that profitable banks improve capitalization by retained earnings (Rime, 2001). Current profits are calculated by the return on assets and are expected to have a positive effect on risk taking. Furthermore, return on equity (ROE) is used as a measure of profitability as well (Gilbert & Wheelock, 2007). Measuring profitability as well from the equity perspective has an expected positive effect, since capital is raised by equity and asymmetric information affects a bank’s value negatively (Rime, 2001).
Size. Third independent variable is size, reflecting the size of the bank. Size is measured by
the natural logarithm of total assets (Laeven & Levine, 2009). The larger a bank the greater the possibilities are to diversify its assets risk. Furthermore, large banks do have greater possibilities to forecast and analyze the risk invited by taking possessions (Saunders et al., 1990). Size is expected having attraction effect on investors, since larger banks are expected to be assumed as ‘too big to fail’ for national governments and are therefore expected to receive capital aid when results face a downturn. Therefore, size is expected to have a negative effect on bank risk taking.
Control variables. Prior research involving bank risk taking and capital adequacy
requirements has established that a number of firm-, and location-level variables can influence it. Hence, several firm-level covariates need to be included in the model. Financial Leverage controls for the leverage effect of a bank. According to Saunders et al.(1990) highly leveraged firms tend to positively affect bank risk taking. The leverage effect is calculated by the book value of total capital over total assets. A higher capital asset ratio (low leverage ratio) should be negatively associated with bank risk taking. A bank’s loan loss provision ratio Loanloss
Provisions affects a bank Tier 1 capital ratio, since it is decreasing the nominal account for the
18
Revenue Growth controls for the recent performance of a bank (Laeven & Levine, 2009). It is
calculated by this year total revenue sales minus previous year’s total revenue sales divided by previous year’s total revenue sales. Performance is expected to have a negative influence on risk taking (Bromiley, 1991). Also age tend to influence bank risk taking, since older banks have had more time to diversify their risky assets (Laeven & Levine, 2009). To control for the age of a bank, measured is the year of foundation (Founded).
Apart from firm-specific covariates that might influence banks’ risk taking, several
country specific variables might influence bank risk taking as well. A dummy is implemented in the equation for government support. Government aid is a dummy that measures if a bank’s home country government backs up the bank with either a guarantee scheme or with direct capital injections in the form of cash or equity. The dummy variable takes a value of one if the government supports the home country bank. Government aid is expected to moderate bank risk taking, since in case banks received publicly capital injections or
guarantee schemes, banks are able to refinance at more favorable terms their funds (Gropp et al., 2011; Hakenes & Schnabel, 2010). Government control is a dummy variable for
government control. It takes a value of one if the government is the largest shareholders in the bank or has over 10% shareholders votes (Laeven & Levine, 2009).
Data on the Tier 1 capital ratio, dates of establishment, and data about government aid and control is extracted from individual bank annual reports covering 2002 to 2010. The
19
Modeling
Bank-risk taking was modeled measuring the z-score of each bank in each country. Additionally, the loan to asset ratio measured risk taking as a robustness check. An Ordinary Least Squares (OLS) regression clustering at bank level is used to model panel data using the z-score as a dependent variable, following the methods of previous studies (Laeven & Levine, 2009). A fixed-effect regression model is generally specified as follows:
Zbc = σbc + Xβbc + εbc
where Zbc is the z-score of bank b in country c as a measure for bank risk taking, σbc is the bank specific fixed effects, X is a vector of covariates, β is a vector of unknown coefficients,
εbc is the independent error term, which follows a gamma distribution. Similarly, for random
effects, the model is:
Zbc = Xβbc + µbc + εbc
Where Zbc is the z-score of bank b in country c as a measure for bank risk taking, X is a vector of covariates, β is a vector of unknown coefficients, µbc is the bank specific random effect accounting for the differences between banks in country c, εbc is the independent error term containing within bank variance. This study interprets both fixed- and random-effects, however the fixed effects are leading in this study. The random effect models are reported in Appendix A6 and A7.
In addition, for checking robustness the dependent variable is replaced for the loan to asset ratio (LTA). A random effect tobit model4 computes the maximum likelihood estimates for the models including the loan to asset ratio, since the dependent variable used in this model clustered at a limiting value - zero in this study (McDonald & Moffit, 1980). Moreover, the sample concentrates between a lower limit of zero and an upper limit of one, implying that the
4 The “Xttobit” routine fits a random-effect tobit model. There is no command for a conditional fixed-effect
20 assumptions of the multiple regressions are realized when using a tobit model (Tobin, 1958). The random-effect model is:
Rbc = Xβbc + µbc + εbc
Where Rbc is the LTA of bank b in country c as a measure for bank risk taking, X is a vector of covariates, β is a vector of unknown coefficients, µbc is the bank specific random effect accounting for the differences between banks in country c, εbc is an independently distributed error term to be normal with zero. Data patterns have to be detected in order to identify the time series components, testing whether changes in capital requirements lead to permanent changes in risk taking.
The STATA 9.2 software package was used to estimate the models in this study. More specific, the fixed and random effects were estimated in the OLS regression models using the package’s “xtreg” routine, accompanied with “fe” and “re” options and “xttobit” routine for the Tobit regression, added manually.
--- Insert table A1 here --- --- Insert table A2 here ---
Results
21 including interaction terms, the maximum VIF was 3.73. This eased concerns about
multicollinearity. All the variables were mean-centered before forming interaction terms. Table 3 presents the estimates of fixed-effect regression models on bank risk taking and the requirements of the Basel II Accord as specified in the Tier 1 capital ratio. Model 1 provides the baseline formulation with all the control variables excluding government intervention. Model 2 provides the baseline formulation including government intervention. Model 3 and 4 provide the estimates of fixed-effect regression models for bank risk taking in relation to Basel II conformity predicting the relationships in hypothesis 1. In table 4 the models 1-5 are hierarchical tests of the relationships predicted in the hypothesis 2-5.
--- Insert table A3 here ---
--- Insert table A4 here --- Hypothesized effects
Table 3 presents the results of hypothesis 1. Hypothesis 1 states that compliance to the Basel II Accord is likely to have a negative relationship with bank risk taking. In model 2 of table 3, the coefficient of the Tier 1 capital ratio, the proxy for the Basel II Accord minimum capital requirements, showed a negative and significant (p<.10) sign. This indicates a negative relationship between the Basel II Accord and bank-risk taking. Table 3 model 4 shows that conformity to the Basel II Accord has a negative and significant (p<0.01) relationship with the level of bank-risk taking. Hypothesis 1 was thus supported.
22 inclusion of size in the interaction term with conformity shows a non-significant positive relation with bank risk taking.
Hypothesis 3 states that the negative relationship between Basel II Accord conformity and bank risk taking weakens as profitability in a bank increases per unit. Model 2 and 3 show that the negative and significant (p <.01) relationship between banks adhering to the Basel II standards (conformity) and bank risks weakens when profitability increases per unit. The interaction term coefficient for conformity and profitability measured by either return on assets or return on equity was negative and significantly (p <.01) related to bank-risk taking, showing that higher profitability weakens the relation between Basel II conformity and risk-taking. Thus, Hypothesis 3 was supported.
Hypothesis 4 suggests that the negative relationship between Basel II Accord conformity and bank risk taking weakens when the bank is high performing. In model 4, the coefficient for the interaction between conformity to the Basel II Accord and bank revenue growth was negative and significant (p <.05). This showed that the negative relationship between the Basel II Accord weakens strongly when the bank is high performing. Thus, Hypothesis 4 is supported.
Model 5 tested hypothesis 5, which predicts that the negative relationship between Basel II Accord conformity and bank risk taking weakens when the bank receives capital support from the government. The estimated coefficient for the interaction between government aid and Basel II conformity was non-significant with a negative sign. This would suggest that the negative relationship between Basel II conformity and bank risk taking strengthen when the bank receives capital support from the government. However, Hypothesis 5 is not supported.
23 between the variables. Therefore, robustness tests are conducted. The predictions are
examined replacing the dependent variable for the loan to asset ratio. The models were examined with the maximum likelihood estimates of the random effects tobit models and the random effect linear regression models presented in Table 5 and 6, respectively, in Appendix A.
Additional Analyses: Basel II Accord and ‘loose coupling’
The Basel II Accord strengthened the minimum capital requirements and set the Tier 1 capital ratio at 8 percent. Replacing T1CR for Conformity, measuring if a bank’s Tier 1 capital ratio exceeds 8 percent, examines whether the imposed capital requirements results in a more stable bank as intended by the Basel II Accord. Table 3 shows the results for the relation between Basel II conformity and bank risk taking. After having confirmed that a higher level of Tier 1 capital is negatively related to bank risk taking, the subsequent step is to assure that the imposed Tier 1 capital level is negatively related to bank risk taking. Table 3 shows that the 8 per cent capital level as set by the Basel II Committee is negative and
significant (p<.01) related to bank risk taking (Model 3 and 4). The regression coefficient for conformity (0.067) presented in Table 3 shows an upturn comparing to the test with the Tier 1 capital ratio. This indicates that conformity is stronger related to less bank risk taking. This confirms the intentions for implementing the Basel II Accord by the Basel Committee.
24 replicated when using random effect OLS regression models. The coefficients for the
interaction terms changed substantially, when using the loan to asset ratio (Appendix table 7). However, when replicating the analysis using a random effect OLS regression models, the results were robust.
--- Insert table A5 here ---
Having examined the relation between bank risk taking and the capital adequacy measures as stated by the Bank for International Settlements (2006), institutional theory suggest that the feeling of security as deployed by the capital requirements in the Basel II Accord leads to camouflage play by the banking industry (Haigh, 2006). According to Haigh and Meyer and Rowan (1977) conformity to institutional rules may give a false feeling of security and stability. Therefore, this research conducted additional analysis. Basel II’s minimum capital requirements and balance sheet variables, bank’s leverage (finlev) and the loan loss provisions (loanloss) were examined in relation to bank-risk taking. Furthermore, age is included in the interaction terms, since older banks are better able to diversify their risky assets (Laeven & Levine, 2009). Additionally, an interaction term containing Basel II conformity and
government control is examined, testing whether government controlled banks display ceremonial behavior.
Model 6-7 (Table 4) show the hierarchical tests of the relationships between balance sheet variables, Basel II Accord conformity and bank-risk taking. All the other relations denoted no significant effect on bank-risk taking. Older banks do not influence the relationship between Basel II conformity and bank-risk taking, nor does governmental control.
25 This study examined banks’ responses to the implementation of the Basel II Accord – bank behavior related to institutional change. It tested the proposition if the Basel II Accord is really effective enough in curtailing bank risk taking or do banks have more leeway in risk taking than expected. Furthermore, the study was designed to identify firm-level
heterogeneity among commercial banks in Europe in their propensity and capacity to loose-couple from the Basel II Accord. Using both the z-score and the loan-to asset ratio as proxy’s for bank risk taking, the empirical assessments support this proposition. The main findings show that bank’s conforming to the Basel II Accord have a negative relationship with bank-risk taking. Furthermore, the negative relationship between Basel II Accord conformity and bank-risk taking weakens strongly when the bank’s profitability level increases or when the bank is high performing. These findings have some important implications for both academic research and regulatory practices.
First, in terms of bank regulation research, these findings confirm the reasoning behind imposing the amended Basel Accord. The minimum capital requirements as stated in the Accord are sufficient to enhance bank risk taking. Questions about the baseline effect of the Basel II Accord have led some research to conclude that the Basel II Accord does not enhance bank-risk taking, but as a consequence lead to an increase in bank-risk taking. The findings in this study demonstrate that the Basel II Accord, measured by the minimum capital
requirements, decreases the risk appetite of banks in the European Union, as intended by the Basel Committee (Bank for International Settlements, 2001; 2004).
26 evaporated by the capital requirements imposed. Ceteris paribus, when profitability raises the Basel II Accord is less effective in curtailing a banks’ risk appetite. The capital buffers
introduced in the Basel II Accord are at the expense of the profitability of banks. Furthermore, as a direct consequence, past performance of a bank weakens the negative relationship
between banks conforming to the Basel II regulations and risk taking. This implements that high performing banks will continue to take more risk even when they have regulatory constraints. More specific, this study demonstrated that the introduction of the Basel II Accord induced banks to trade-off between Basel II conformity and risk-taking. This study took as its starting point that organizations loose couple their formal structure from their actual work structure in regard to capital regulations under institutional change (Meyer & Rowan, 1977). As suggested in prior research, organizations not simply reflect passive compliance to capital regulations, but act in a rational way to create formal structures to reflect the minimum capital requirements imposed in regulatory constraints (Meyer & Rowan, 1977; Oliver, 1991). Indeed, the findings in this study confirm that banks loose couple their formal structure from their actual work activities in order to deal with the trade-off between Basel II conformity and risk-taking. This has been the first large-scale study examining bank behavior in relation to capital regulation, specifically focusing on decoupling of bank formal structures. Hopefully, this study can be a stepping stone for future studies examining the relation between bank loose coupling and bank-risk taking after institutional change.
27 behavior in relation to regulatory constraints. More precisely, this study focused on whether capital regulations have led to decoupling of a bank’s structure in order to manage their risk level themselves while simultaneously conforming to the Basel II Accord. This analysis (model 6-9 in Table 4) was inconclusive owing to data limitations, however this question deserves further research attention when larger scale data are available.
Secondly, the results imply a focus shift of European banking policy makers. Loose coupling is a serious threat to institutional rules, organizations decoupling their formal structure from actual work activities have demonstrated that single banks might increase risk taking, while adhering to the Basel II Accord capital requirements. These results extend the concerns raised in previous research stressing the impact of the Basel II Accord on bank risk taking (Hakenes & Schnabel, 2010; Hawkins & Turner, 2000). This study highlighted a nonmarket consideration important for organizations in relation to institutional rules. This study projects the influence of bank behavior in relation to bank capital rules in a large empirical assessment and identifies the firm-level heterogeneity in their propensity and capacity to loose-couple from regulatory rules. Furthermore, it supplements studies of ceremonial behavior with empirical results, considered from both institutional and financial theory perspectives. Such an approach adds value to Scott’s (2004a) proposition that
organizations can and do decouple work activities from accounting control and other review systems, in a way that it reflects the extent to which this occurs.
28 Despite its potential contributions, this study has some limitations that are important for interpreting its empirical findings and suggest further research directions. First, due to the limitation on variables able to measure the minimum capital requirements as set in the Basel II Accord, Tier 1 capital ratio was used as a single proxy for the Basel II Accord. Future studies may adopt a more creative way of measuring institutional change and more specific adherence to the stated minimum capital requirements. Secondly, the control variables used in this study do not control for many bank and industry specific effects on the corporate
governance of a bank, the distribution of the cash-flows or cross-shareholding among the shareholders, due to inability to collect adequate data on these controls. The corporate governance board is considered having a main effect on bank risk taking, even as the
shareholders of the bank. However, failing to include them did not give inconclusive results in this study. Neither biased it the findings, since both effects would influence the proxy’s for risk measuring, though would only suggest additional validation of the relation between Basel II conformity and bank risk taking.Future studies can examine these effects to get a more detailed insight about the intentions of decoupling structures are established. This area of research would benefit if comprehensive models could be developed to understand the precise intentions of loose coupling behavior.
29 study thus bring loose coupling due to regulatory constraints to the forefront of research on the impact of regulatory rules on bank behavior.
30
References
Altman, E. 1968. Financial ratios, discriminant analysis and the prediction of corporate bankruptcy. The Journal of Finance, 4: 189-209.
Avery, R. & Berger, A. 1991. Risk-based capital and deposit insurance reform. Journal of
Banking and Finance, 15: 847-874.
Bank for International Settlements, 1999. Capital requirements and bank behaviour: The Impact of the Basel Accord. Working Papers, 1.
Bank for International Settlements, 2001. Overview of the new Basel capital accord. Basel
Committee on Banking Supervision, consultative document.
Bank for International Settlements, 2006. International convergence of capital measurement and capital Standards. Basel Committee on Banking Supervision, Basel consultative document.
Bank for International Settlements, 2004a. International convergence of capital measurement and capital standards. Basel Committee on Banking Supervision, Basel consultative
document.
Banz, R.W. 1981. The relationship between return and market value of common stocks.
Journal of Financial Economics, 9: 3-18.
Bebchuk, L. and Fried J. 2003. Executive compensation as an agency problem. Journal of
Economic Perspectives, 17: 71-92.
Blum, J. 1999. Do capital adequacy requirements reduce risks in banking? Journal of
Banking and Finance, 23: 755-771.
Bromiley, P. 1991. Testing a causal model of corporate risk taking and performance.
Academy of Management Journal, 34: 37-59.
Caruana J. & Narain A. 2008. Banking on more capital. Finance & Development, 45: 24-28. Doerig, H.U. 2003. Operational risks in financial services: an old challenge in a new
environment. Credit Suisse Group Working paper. (Available at: https://www.credit-suisse.com/governance/doc/operational_risk.pdf)
31 Esty, B. 1998. The impact of contingent liability on commercial bank risk taking. Journal of
Financial Economics, 47: 189-218.
Furlong, F.T. & Keeley, M.C. 1989. Capital regulation and bank risk-taking; A note. Journal
of Banking and Finance, 13, 883-891.
Gennote, G. & Pyle, D. 1991. Capital controls and bank risk. Journal of Banking and
Finance, 13: 883-891.
Gilbert, R.A. & Wheelock, D.C. 2007. Measuring commercial bank profitability: proceed with caution. Federal Reserve Bank St. Louis Review, 22: 515-532.
Gropp, R. Hakenes, H. & Schnabel, I. 2010. Competition, risk-shifting, and public bail-out politics. The Review of Financial Studies, 24: 2084-2120.
Haigh, M. 2006. Camouflage play: making moral claims in managed investments.
Accounting Forum, 30: 267-283.
Hakenes, H. & Schnabel, I. 2010. Banks without parachute: competitive effects of government bail-out policies. Journal of Financial Stability, 6: 156-168.
Hakenes, H. & Schnabel, I. 2011. Bank size and risk-taking under Basel II. Journal of
banking & Finance, 35: 1436-1449.
Hawkins, J. & Turner, P. 2000. International financial reform: regulatory and other issues.
Paper presented at a conference on International Financial Contagion, 3-4 February,
Washington.
Hermsen, O. 2010. The impact of the choice of VaR models on the level of regulatory capital according to Basel II. Quantitative Finance, 10: 1215-1224.
Ingram P, & Clay, K. 2000. The choice-within-constraints new Institutionalism and implications for sociology. Annual Review of Sociology, 26: 525-546.
Karels, G.V. & Weber, M. 2004. Do deposit insurances premiums affect bank risk-taking?
Investment Management and Financial Innovations, 1: 72-87.
Kwan, S. & Eisenbeis, R. A. 1997. Bank risk, capitalization, and operating efficiency.
Journal of Financial Services Research, 12: 117-131.
Laeven, L. & Levine, R. 2009. Bank governance, regulation, and risk taking. Journal of
Financial Economics, 93: 259-275.
McDonald, J.F. & Moffitt, R.A. 1980. The uses of tobit analysis. The Review of Economics
32 Mathuva, D.M. 2009. Capital adequacy, cost income ratio and the performance of commercial banks: the Kenyan Scenario. The international Journal of Applied Economics and Finance, 3: 35-47.
Meyer, J.W. & Rowan, B. 1977. Institutionalized organizations: formal structure as myth and ceremony. American Journal of Sociology, 83: 340-363.
Mezias, S. J. 1995. Using institutional theory to understand for-profit sectors: the case of financial reporting standards. In: The Institutional Construction of Organizations:
International and Longitudinal Studies. W. Richard Scott and Søren Christensen, ed.
Thousand Oaks, CA: Sage: 164-96.
Montgomery, D., Peck, A. & Vining, G. 2007. Introduction to linear regression analysis. 4th Edition, John Wiley, New York, 34-38.
Moosa, Imad A. 2010. Basel II as a casualty of the global financial crises. Journal of
Banking Regulation, 11: 95-114.
North, D.C. 1993. Institutions and credible commitment. Journal of Institutional and
Theoretical Economics, 149: 11-23.
Oliver C. 1991. Strategic responses to institutional processes. The Academy of Management
Review, 16: 145-179.
Smith, G. 2012. Why I am leaving Goldman Sachs. New York Times, 15-03-2012.
Stern, G.H. & Feldman, R.J. 2004. Too big to fail: the hazards of bank bailouts. Brookings
Institution Press, Washington, D.C.
Park, S. & Peristiani, S. 2007. Are bank shareholders enemies of regulators or a potential source of market discipline? Journal of Banking and Finance, 31: 2493-2515.
Philips D.J, & Zucherman, E. 2001. Middle‐status conformity: theoretical restatement and empirical demonstration in two markets. American Journal of Sociology, 107: 379-429. Repullo, R. 2004. Capital requirements, market power, and risk-taking in banking. Journal of
Financial Intermediation, 13: 156–182.
Rime, B. 2001. Capital requirements and bank behaviour: empirical evidence for Switzerland.
Journal of Banking & Finance, 25: 789-805.
Saunders, A., Strock, E. & Travlos, N.G. 1990. Ownership structure, deregulation, and bank risk taking. The Journal of Finance, 45: 643-654.
Scott, R.W. 2004a, Institutional theory: contributing to a theoretical research program,
Chapter prepared for Great Minds in Management: The Process of Theory Development,
Oxford UK
33 Singh, V. J. 1986. Performance, slack, and risk taking in organizational decision making.
Academy of Management Journal, 29: 562-585.
Sinkey Jr., J.F. & Greenawalt, M.B. 1991. Loan-loss experience and risk-taking behaviour at large commercial banks. Journal of Financial Services Research, 5: 53-59.
Thadden Von, E.L. 2004, “Bank capital adequacy regulation under the new Basel accord,”
Journal of Financial Intermediation, 13: 90–95.
Thomas, A.B. 2004. Research Skills for Management Studies. Routledge, UK.
Tobin, J. 1958. Estimation of the relationship for limited dependent variables. Econometrica, 26: 24-36.
VanHoose, D. 2007. Theories of bank behavior under capital regulation. Journal of Banking
APPENDIX Table A1 Descriptive statistics
This table presents the summary statistics of the main variables. LTA is the loan to assets ratio; the higher the ratio the more risk is taken by the bank. Z-score is the z score of the bank; the higher the z score, the lower the probability of default. Tier 1 Capital Ratio is the Tier 1 capital ratio of the bank. Size is the log of total assets. Return on Equity is the return on equity for the bank. Return on Assets is the return on assets for the bank. Financial Leverage is measured by the book value of total capital over total assets. Revenue Growth is the growth in total revenues of the bank. Loan Loss Provisions Ratio is the ratio of a bank’s loan loss provisions to net interest income. Age are the years of the bank since founding. Government Aid is a dummy variable which takes the value of one when the government banked up or supported the bank with a capital injection or guarantee scheme. Government Control is a dummy which takes the value of one if the government has more than 10% of the shareholder rights. Conformity is a dummy variable which takes a value of one if the bank’s Tier 1 Capital Ratio exceeds eight per cent. All variables are measured over the period 2002-2010.
Variable Number
of banks
Mean Std. Dev. Min. Max.
LTA 110 0.69 0.16 0.00 1.00
Z-score 110 1.11 0.73 -2.75 2.88
Tier 1 Capital Ratio 109 11.07 6.02 2.25 70.53
Size 107 7.41 0.95 5.34 9.72
Return on Equity 114 9.28 16.69 -230.31 60.15
Return on Assets 115 0.01 0.02 -0.49 0.13
Financial Leverage 115 0.24 0.15 0.02 0.99
Revenue Growth 112 0.06 0.20 -0.55 0.99
Loan Loss Provision Ratio 104 0.18 0.50 -14.20 0.96
Founded 112 101.48 87.21 1 538
Government aid 119 0.31 0.46 0 1
Government control 121 0.13 0.34 0 1
35 Table A2
Correlation matrix a
1
2
3
4
5
6
7
8
9
10
11
12
1. Loan to Assets Ratio
2. Z-score 0.007
3. Tier 1 Capital Ratio 0.118 -0.174
4. Size -0.305 0.085 -0.357
5. Return on Equity -0.048 -0.573 0.014 0.055
6. Return on Assets -0.018 0.316 0.135 -0.231 0.500
7. Financial Leverage 0.294 0.099 0.115 -0.165 -0.014 -0.056
8. Revenue Growth 0.086 0.250 -0.092 -0.051 0.252 0.175 0.046
9. Loan Loss Provisions 0.034 -0.121 0.015 0.001 -0.168 -0.140 0.063 -0.104
10. Founded 0.009 0.059 -0.018 0.034 0.021 0.029 0.067 0.014 -0.020
11. Government Aid 0.009 -0.333 0.017 0.076 -0.043 -0.004 -0.071 0.014 0.037 0.160
12. Government Control -0.114 -0.131 -0.140 0.202 -0.035 -0.055 0.004 -0.052 -0.007 -0.108 0.267
13. Conformity -0.010 -0.196 0.432 -0.308 0.034 0.101 -0.063 -0.090 -0.027 -0.126 0.032 -0.181
a
36 Table A3
Fixed Effects Linear Regression Models Predicting Capital Adequacy Requirements and Bank Risk Taking by European Commercial Banks, 2005-2010a
a
Although our panel covered 12,844 bank-country-year observations of both European banks between 2005 and 2010, only 7,603 were used in this analysis due to the restrictions for the fixed effects of the linear regression and the limitations of using the z-score. Standard errors in parentheses * p < .10 ** p < .05 *** p < .01. Two-tailed tests. Dependent variable: Zscore
Covariates Model 1 Model 2 Model 3 Model 4
Tier 1 Capital Ratio 0.011**
(0.005) 0.012** (0.005) Conformity 0.067*** (0.022) 0.066*** (0.023) Size -0.049 (0.103) -0.052 (0.104) -0.077 (0.101) -0.078 (0.102) ROA 24.422*** (2.596) 24.369*** (2.613) 24.093*** (2.579) 24.078*** (2.595) ROE 0.032*** (0.002) 0.032*** (0.002) 0.031*** (0.002) 0.031*** (0.002) Financial Leverage 0.853*** (0.132) 0.851*** (0.133) 0.852*** (0.127) 0.851*** (0.129) Revenue Growth 0.022 (0.042) 0.022 (0.043) 0.036 (0.041) 0.036 (0.042)
Loan Loss Provisions -0.005
37 Table A4
Fixed Effects Panel Regression Models with Added Interaction Terms Predicting Capital Adequacy Requirements Conformity and Bank Risk Taking by European Commercial Banks, 2005-2010a
Covariates Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9 Model 10
38 Conformity * FINLEV -0.159 (0.196) -0.131 (0.198) Conformity * FOUNDED -0.000 (0.000) -0.000 (0.000) Conformity * Control 0.063 (0.054) 0.009 (0.063) Constant 0.309* (0.170) 0.796 (0.759) 0.908 (0.772) 0.898 (0.772) 0.897 (0.777) 0.965 (0.787) 1.158 (0.777) 0.859 (0.778) 0.881 (0.775) 0.948 (0.773) R-Squared 0.8391 0.8461 0.8423 0.8406 0.8391 0.8392 0.8393 0.8393 0.8396 0.8512 F-Value 239.39 252.31 245.12 242.08 239.33 239.50 239.74 239.63 240.22 136.30 Mean VIF 1.71 1.78 1.68 1.68 2.30 2.51 1.65 2.06 1.90 3.73 a
39 Table A5
Maximum-Likelihood Estimates of Random Effects Panel Tobit Regression Models Predicting Capital Adequacy Requirements and Bank Risk Taking by European Commercial Banks, 2002-2010a
Covariates Model 1 Model 2 Model 3 Model 4
Tier 1 Capital Ratio -0.003***
(0.001) -0.003*** (0.001) Conformity -0.019*** (0.006) -0.016*** (0.006) Size -0.058*** (0.004) -0.057*** (0.004) -0.052*** (0.004) -0.044*** (0.004) ROA 1.549*** (0.555) 1.659*** (0.561) 1.639*** (0.597) 2.236*** (0.534) ROE -0.000 (0.000) -0.000** (0.000) -0.001** (0.000) -0.001** (0.000) Financial Leverage 0.353*** (0.033) 0.335*** (0.024) 0.257*** (0.021) 0.254*** (0.018) Revenue Growth -0.006 (0.012) -0.004 (0.012) 0.003 (0.012) 0.001 (0.012)
Loan Loss Provisions -0.006
(0.004) -0.005 (0.004) -0.006 (0.004) -0.007 (0.004) Founded -0.000 (0.000) -0.000** (0.000) -0.000 (0.000) 0.000 (0.000) Government AID -0.001 (0.006) 0.033*** (0.006) Government Control -0.072*** (0.010) -0.028*** (0.008) Constant 1.104*** (0.036) 1.100*** (0.035) 1.055*** (0.038) 0.964*** (0.030) Log Likelihoodb 883.076 874.473 953.755 954.872 Mean VIF 1.44 1.43 1.45 1.43
a Standard errors in parentheses * p < .10 ** p < .05 *** p < .01. Two-tailed tests. b Log likelihoods in tobit models are a log of a probability densities and not the log of a
40 Table A6
Random Effects Linear Regression Models Predicting Capital Adequacy Requirements and Bank Risk Taking by European Banks, 2005-2010a
a
Standard errors in parentheses * p < .10 ** p < .05 *** p < .01. Two-tailed tests. Only 9,676 observations were used in this analysis due to the limitations of using the z-score. Dependent variable: z-z-score.
Covariates Model 1 Model 2 Model 3 Model 4
Tier 1 Capital Ratio 0.009*
(0.005) 0.009* (0.005) Conformity 0.056** (0.022) 0.055** (0.023) Size 0.053 (0.052) 0.082 (0.051) 0.043 (0.052) 0.072 (0.051) ROA 24.592*** (2.568) 24.809*** (2.578) 24.342*** (2.560) 24.626*** (2.571) ROE 0.032*** (0.002) 0.032*** (0.002) 0.032*** (0.002) 0.032*** (0.002) Financial Leverage 0.884*** (0.126) 0.888*** (0.127) 0.874*** (0.123) 0.880*** (0.124) Revenue Growth 0.014 (0.045) 0.019 (0.043) 0.032 (0.041) 0.036 (0.042)
Loan Loss Provisions -0.005
41 Table A7
Maximum-Likelihood Estimates of Random Effects Panel Tobit Regression Models with Added Interaction Terms Predicting Capital Adequacy Requirements Conformity and Bank Risk Taking by European Banks, 2002-2010a
Covariates Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9 Model 10
42 LOANLOSS (0.028) (0.032) Conformity * FINLEV 0.004 (0.037) -0.029 (0.039) Conformity * FOUNDED 0.000 (0.000) 0.000 (0.000) Conformity * Control 0.001 (0.013) 0.010 (0.015) Constant 0.634*** (0.010) 0.970*** (0.029) 0.937*** (0.032) 0.963*** (0.030) 0.966*** (0.031) 0.963*** (0.029) 1.016*** (0.030) 0.964*** (0.030) 0.964*** (0.030) 0.713*** (0.009) Log Likelihoodb 954.573 959.177 959.759 954.980 956.307 956.025 955.322 954.873 954.877 963.228 Mean VIF 1.45 1.63 1.72 1.42 1.94 1.88 1.41 1.79 1.62 3.07 a
43 Table A8
Random Effects Panel Regression Models with Added Interaction Terms Predicting Capital Adequacy Requirements Conformity and Bank Risk Taking by European Commercial Banks, 2005-2010a
Covariates Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9 Model 10
44 Conformity * FINLEV -0.127 (0.193) -0.089 (0.198) Conformity * FOUNDED -0.000 (0.000) -0.000 (0.000) Conformity * Control 0.076 (0.054) 0.019 (0.064) Constant 0.097 (0.389) 0.078 (0.380) 0.124 (0.390) -0.095 (0.384) -0.079 (0.393) -0.073 (0.394) 0.143 (0.389) -0.113 (0.387) -0.085 (0.388) -0.081 (0.376) R-Squared 0.8382 0.8456 0.8417 0.8397 0.8382 0.8382 0.8384 0.8383 0.8387 0.8502 Mean VIF 1.71 1.78 1.68 1.68 2.30 2.51 1.65 2.06 1.90 3.73 a
45 Table A9
Bank risk, Tier 1 capital ratio, profitability, size and governmental support and control by country
Country Loan to Assets Z-score Tier 1 Capital Rat. Return to Assets
