Regulation and risk taking in the banking industry: evidence from Tunisia

Transcription

1 Afro-Asian J. Finance and Accounting, Vol. 3, No. 1, Regulation and risk taking in the banking industry: evidence from Tunisia Amel Belanes* Department of Finance, Faculty of Economics and Management of Nabeul, University of Carthage, Campus Universitaire Mrezga Route Hammamet, 8000, Nabeul, Tunisia *Corresponding author Afef Ben Hajiba Research Unit ( FIESTA), High Institute of Management of Tunis, University of Tunis, 41, Rue de la Liberté, Cité Bouchoucha 2000 Le Bardo, Tunis, Tunisia ben_hajiba_afef@yahoo.fr Abstract: This research highlights the impact of capital regulation and franchise value on bank risk taking in the Tunisian context. Using a panel set of Tunisian commercial banks during the period spanning from 1997 to 2007, our study puts in evidence that stringent capital requirements help deterring the bank risk taking. Similarly, banks with high franchise value may have an incentive to avoid risky business strategies. However, empirical findings clearly bring forth that the intensity of such decrease varies with the risk taking measurement. Capital regulation allows reducing both the market risk and the insolvency risk but has no effect on the specific risk of the bank; while the franchise value has a negative effect on only the market risk of the bank. Overall, our study indicates that capital allocation should be more risk sensitive and that the whole bank risk taking ought to be deeply investigated and quantified. Keywords: bank risk taking; capital regulation; franchise value; Tunisia. Reference to this paper should be made as follows: Belanes, A. and Ben Hajiba, A. (2012) Regulation and risk taking in the banking industry: evidence from Tunisia, Afro-Asian J. Finance and Accounting, Vol. 3, No. 1, pp Biographical notes: Amel Belanes is an Associate Professor of Finance at the Faculty of Economics and Management of Nabeul, University of Carthage, Tunisia. She belongs to the Research Laboratory of Finance and Business Strategy (FIESTA Finance et Stratégie des Affaires) at the High Institute of Management of Tunis, University of Tunis, Tunisia. Her research interests are corporate finance, corporate governance, risk taking, dividend policy, management risk, banking industry and Islamic finance. She has published many articles, namely in the International Review of Finance, Economics and Copyright 2012 Inderscience Enterprises Ltd.

2 90 A. Belanes and A. Ben Hajiba Finance Review, International Journal of Economics and Finance, Markets, Bankers & Investors, Management International, Journal of Emerging Market Finance, Corporate Ownership and Control, Journal of Global Business Administration, etc. She is a frequent presenter in national and international conferences namely ICF Tunisia, ATSG Tunisia, ICEF Turkey, ATINER Greece, AGB USA, etc. Afef Ben Hajiba is a Doctoral candidate at the High Institute of Management of Tunis, University of Tunis, Tunisia. Her research interests are risk taking, banking industry, corporate finance and corporate governance. 1 Introduction Bank regulation has been an issue of great interest in the finance literature. While most financial economists have analysed the impact of bank regulation on risk taking from a theoretical perspective, a little bit of recent research tries to provide quantitative guidance. The multiplication of banking sector problems has raised widespread concern and bank risk taking behaviour is considered as the main cause of such distress. In response to financial crisis, the Basel Committee on Banking Supervision has unceasingly attempted to review and improve the guidelines of the Basel Accords. The committee always tries to update their guidelines for capital and banking regulations. The objective is to set forth a robust banking sector that is able to cope with problems and shocks arising from financial and economic stress and reduce the externalities created by systemically important institutions. Since 1988, central bankers of the Group of Ten deliberated a set of minimal capital requirements for banks. This is known as the 1988 Basel Accord which was enforced by law in Basel I primarily focused on credit risk. Assets of banks were classified and grouped in five categories according to credit risk. Banks are required to hold capital equal to 8% of the risk-weighted assets. Basel I Accord is now widely considered as outdated. A more comprehensive set of guidelines, known as Basel II are in the process of implementation by several countries and new updates in response to the financial crisis commonly described as Basel III. Basel II, which was initially published in June 2004, aims at improving the minimum capital requirements, the supervisory review and the market discipline. For instance, it addresses that the capital allocation should be more risk sensitive and that operational risk and credit risk ought to be separated and quantified. Besides, economic and regulatory capital should be thoroughly aligned in order to reduce the regulatory arbitrage. Basel III, which will be implemented in December 2012, includes tighter definitions of Tier 1 capital, the introduction of a leverage ratio, a framework for counter-cyclical capital buffers, measures to limit counter party credit risk, and short and medium-term quantitative liquidity ratios. This framework has been progressively introduced in most countries all over the world. But the efficiency with which Basel Accords are enforced is variable. Some countries have adopted, at least in name, the principles prescribed under either the Basel I, or II. Basel III is actually under process. Tunisia has recently witnessed major structural and policy changes. Capital requirements in Tunisia have been dictated by Basel I since But Tunisian banks capital represented only 5% of the risk-weighted assets at that time. By the end of 1999,

3 Regulation and risk taking in the banking industry 91 the Central Bank of Tunisia (CBT) increased the minimum capital requirements for the banking industry to 8%. Even though the rate of commercial banks is nowadays higher than 8%, Tunisian banks do not seem to comply with prudential rules of Basel II. The Tunisian Central Bank is actually reviewing the regulations guidelines in order to ensure safety and soundness of this sector and enable it to cope with problems and shocks arising from financial and economic crises. In this regard, it should be pointed here that Tunisian banks were not seriously harmed by the late-2000s financial crisis like other banks all over the world though this crisis is argued to be the worst financial crisis since the Great Depression of the 1930s. There is some debate that arises: to what extent does the alignment with the Basel capital requirements influence the bank risk taking? Capital regulation represents one of the instruments of the prudential regulations and is recognised to maintain financial stability. Capital requirements are intended to limit the insolvency risk of banks. Whether more stringent capital requirements lead to reduced or to increased bank risk taking has been discussed controversially in the academic literature. Similarly, the franchise value depends on the banking organisation and regulation and is expected to constrain bank s risk taking as banks would not put at stake their valuable charters. Although the literature on this subject is growing and one can think of the cost of this issue, the research undertaken to analyse the impact of the franchise value on risk taking has been limited by data availability and methodology issues. Furthermore, the Tunisian banking sector has undergone many profound changes over the last two decades. On the one hand, advances in technology, financial liberalisation, the ongoing economic and regulatory integration should incite bank risk taking. On the other hand, the wave of bank mergers and acquisitions may lead to the opposite effect as the number of competitors significantly decreases. Therefore, the combined overall impact of these changes on bank risk taking is uncertain. To properly address these questions, it has become necessary to thoroughly analyse the effect of both the capital regulation and the franchise value on risk taking within the Tunisian banking industry. The contribution of this paper is threefold. First, our study is the first for Tunisia that analyses the effects of both capital regulation and franchise value on banks risk taking. But above all, our investigation wants to shed some light on the importance of the both micro and macro aspects of the regulation to the stability of the banking industry in Tunisia. Second, our study provides a comprehensive framework to assess the influence of regulation on four measures of banks risk taking: total risk, specific risk, market risk and risk insolvency as it is worth dealing with different features of risk. In fact, in the Basel I only one risk, that is credit risk, that was dealt with in a simple manner while market risk was an afterthought; operational risk was not dealt with at all. The Basel II Accord tries to address this shortcoming by separating specific from market risk. The leverage ratio will be also added as a supplementary measure of risk in the Basel III. Third, we use panel data regression to investigate the influence of the regulation on bank risk taking. Such an econometric technique is adapted for countries in which the number of listed companies is very small, such as Tunisia. The remainder of the paper is organised as follows. Section 2 covers the relevant literature on relationship between the bank regulation and risk taking. Section 3 outlines the specificities of the Tunisian institutional background. Section 4 describes the data and the research design. Section 5 discusses the empirical results. A brief conclusion follows with implications of the findings and suggestions for future research.

4 92 A. Belanes and A. Ben Hajiba 2 Literature review A large stream of finance literature investigates the effects of capital regulation on bank risk choice but there is an ongoing controversial debate on the sign of the relationship between both dimensions. In this paper we investigate the impact of both the capital adequacy requirement and the charter value, which are the potential drivers of bank system stability, on the bank risk taking. This paper introduces a new argument which has been neglected so far. It puts in evidence that that the sense and the intensity of this influence vary with the bank risk taking measurement. Both theoretical and empirical papers, ranging from portfolio models (Kahane, 1977; Koehn and Santomero, 1980; Kim and Santomero, 1988) to state-preference models and models based on option pricing theory (Furlong and Keeley, 1989) to incentive-based approaches (Besanko and Kanatas, 1996; Blum, 1999; Milne, 2002) remain inconclusive. The answers vary due to the different analytical frameworks. In portfolio theory-based approaches, the bank acts as a portfolio manager. Tighter capital regulation influences the bank s leverage ratio; which incites the bank to reallocate its portfolio by selecting riskier assets in order to maintain its expected return on equity at an optimal level (Kahane, 1977; Koehn and Santomero, 1980; Kim and Santomero, 1988). In fact, stringent capital requirements provide a utility-maximising bank the incentive to increase the risk of its portfolio. The increase in asset risk may offset the desired effect of reduced leverage. However, in models based on state-preference theory and option pricing theory, capital regulation might either reduce risk taking incentives (Furlong and Keeley, 1989) or induce higher risk taking for value-maximising banks. For instance, Furlong and Keeley (1989) point out the relevance of the deposit insurance guarantee and prove that it is not optimal for value-maximising banks to raise asset risk as a reaction to a higher regulation. Moreover, in a mean-variance framework, improperly chosen risks weights may induce banks to take on additional risks (Koehn and Santomero, 1980; Kim and Santomero, 1988; Rochet, 1992). In incentive-based approaches, tighter regulation can mitigate conflicts between bank managers and different groups of claimholders, namely shareholders, depositors and other creditors; and efficiently allocate control rights between them (Dewatripont and Tirole, 1995; Besanko and Kanatas, 1996; Calem and Rob, 1999; Anderson and Fraser, 2000; Milne, 2002). It compels banks to have more of their own capital at risk in order to internalise the inefficiency of gambling or investing in high-risk assets (Ben Naceur and Kandil, 2009; Yilmaz, 2009). Capital requirement are thus expected to influence the bank managers incentives; thereby, regulating and moderating risk taking levels; which helps preserving the financial stability and maintaining the banking system soundness. In this respect, the Capital Accord has been acknowledged for its influence on bank risk behaviour and its contribution to the widespread use of risk-based capital ratios despite the several distortions to the banking business. In an inter-temporal context, capital regulation is expected to have two effects on bank risk taking incentives (Blum, 1999). Capital adequacy requirement might influence both the marginal cost and the marginal return of risk taking. On one hand, stricter regulation lowers the marginal cost of risk taking and hence the increase of risk taking. In fact, stricter regulation reduces both the expected profits of the bank in case of success; and the expected loss if it defaults. On the other hand, higher requirements may affect the marginal return of taking risk; but the effect depends on the regime whether the regulation is only binding in the first or the second period. A stricter regulation today

5 Regulation and risk taking in the banking industry 93 tends to reduce risk while the introduction of a capital requirement for tomorrow induces a higher risk today (Blum, 1999) as raising equity is excessively costly. Likewise, if a value-maximising bank additionally issues subordinated uninsured debt and has to pay a flat-rate deposit insurance premium, its reaction to a higher capital requirement may change substantially (Homölle, 2004). For instance, banks are induced to increase asset risk due to the enforcement of a more stringent capital requirement. Using the seminal model proposed in Blum (1999), Silva (2007) puts in evidence that the level of taking risk depends on capital requirement thresholds; and consequently upon the initial equity of the bank. By applying constant capital requirements and at very high levels of the requirement, regulated banks are tempted to take lesser risks than the unregulated bank, and can even achieve the zero bankruptcy cost. Beyond these theoretical and empirical researches, there is little conclusive evidence that an unregulated bank is tempted to take higher risks than a regulated bank. Besides, there is no consensus on how best to design the regulation of bank capital (Santos, 1999). A part from the capital regulation, the charter value is considered as well as a bank s self-imposed risk discipline device (Buser et al., 1981; Marcus, 1984; Keeley, 1990). In banking, the franchise value arises from regulatory restrictions on entry and competition. Thus, regulatory restrictions have a positive effect on bank charter value (Gonzláez, 2005; Laeven and Levine, 2009; Fonseca and González, 2010). Since bank owners have much to lose if the bank becomes insolvent, a bank with high franchise value may have an incentive to avoid risky business strategies. Indeed, franchise value can counterbalance asset substitution moral hazard and reduce the bank s incentives to adopt excessive risk taking strategies (Marcus, 1984). In this regard, higher charter values are associated with a higher market power; which is correlated to higher banking solvency ratios and lower risk of credit (Salas and Saurina, 2003; Konishi and Yasuda, 2004; Gonzláez, 2005). Additionally, the franchise value and subordinated debt are suggested to work together to constrain banks risk taking (Niu, 2008). Finally, insight in the evolution of the franchise value reflects the financial stability and perhaps the mechanisms available to supervisors and regulators to maintain stability. It is clear that competition and stability issues are interrelated and should be considered accordingly. Previous research has analysed the link between charter value, bank risk taking and financial stability. In this vein, the third pillar of the Basel II framework advocates the adoption of market discipline mechanisms for prudential supervision and suggests investigating the adjusted Tobin s Q, which is the common proxy for the franchise value. This approach is based on the assumption that well-informed market participants will reward a risk conscious management strategy by credit institutions in their asset allocation decisions. Therefore, banks with higher franchise value have lower default risk as measured by lower risk premiums on their large. Similarly, risk at individual banks is inversely related to value, indicating that high franchise value banks have less total and bank specific risk (Demsetz and Strahan, 1997; Anderson and Fraser, 2000). In contrast, low charter-values should encourage managers to take on additional risk. Increases in competition reduced the charter value of banks and hence the increase of default risk through increases in asset risk and reductions in capital (Keeley, 1990). Increase in competition, caused by market deregulation, may erode market power and monopoly profits, and hence the decline of charter values (Galloway et al., 1997). Similarly, when the industry is unhealthy, the franchise values are expected to be low; which encourages

6 94 A. Belanes and A. Ben Hajiba bad managers to take excessive risk (Gorton and Rosen, 1995). In fact, intense competition that lowers the franchise value of incumbent banks intensifies incentives for both stockholders and managers to increase risk. 3 Tunisian institutional background Tunisia has recently been ranked the most competitive country in the African continent, the fourth country in the Arab world and the 35th worldwide. It has succeeded to maintain an average growth rate of 3.1% versus 5% during the last years despite the recent international financial crisis. To achieve such performance, the Tunisian authorities undertook several reforms towards a more liberal economy. For instance, Tunisia has shifted to an open-door policy with an active role for the private sector in economic management and less governmental control. Tax structures are also simplified and a prudent approach to debt was adopted. But above all, Tunisia has made important progress over the last decades in order to improve its financial sector. In fact, the banking sector is the major component of the Tunisian economy and its soundness is great to revive the whole economic activity and sustain a high level growth. These reforms include advances in technology, financial liberalisation, the ongoing economic and regulatory integration, which should increase the degree of competition and efficiency in the Tunisian banking sector, and hence the whole Tunisian economy. To achieve such objectives, Tunisia has adopted the progressive liberalisation since 1987 through ongoing structural adjustment programmes. Before, most banks were state-owned and are assigned the task of providing long-term finance for the whole economy. They were concentrated and segmented, which weakens competition. Several reforms have been implemented so as to promote the equity market and enhance banks privatisation. Interest rates and credit allocation have been liberalised since The financial sector was also opened to foreign investors and financial institutions in 2000 and the banking sector witnesses a lot of joint-ventures. Following these developments, real GDP growth over the last ten years averaged 5% and real GDP per capita increased by 45% between 1997 and However, the rapid growth of the banking sector during the 1990s and the progressive liberalisation of the whole economy added extra burden on the CBT as the sole regulator and supervisor of the banking industry. Therefore, the CBT was compelled to develop its supervisory framework and staff. The Tunisian authorities have enhanced the financial sector regulations and strengthened the bank supervision on the basis of internationally accepted standards to deal with the risk inherent in the new policy. Most regulatory reforms implemented by the CBT complied with the Basle Core Principles for Effective Banking Supervision. The prudential regulation includes the risk repartition, its coverage by provisions and the adequacy of capital. Prior to reforms in the early 1990s, the banking sector was heavily regulated through credit controls and portfolio restrictions. In 1999, the CBT implemented the minimum capital adequacy standard and increased the minimum capital requirements versus their risk-weighted assets to 8%, along the lines proposed by the Basel Committee on Banking Supervision. Other supervision and prudential regulations were applied so as to urge the banks to comply with the new minimum capital adequacy ratio. Provisioning levels for loans were monitored very closely. The provisioning ratio, for instance, has approximately increased to 53.8% in The solvency ratio of the Tunisian banks has also dramatically

7 Regulation and risk taking in the banking industry 95 improved and reached in 2007 a rate of 13, 7%. However, the interest margin rate of Tunisian banks in percentage of the Gross National Product falls from 62% in 2001 to 57% in 2004; following the decline of the monetary market rate (MMR). Since the recovery of the MMR in 2005, this rate strengthened to pass in 59, 2% at the end of Also, the intermediation rate of listed banks on Tunisian Stock Exchange (TSE), knew a continuous decline to pass from 112, 6% in 2002 to 93% in The introduction of capital adequacy rules has strengthened bank capital and improved the bank soundness. A well-organised banking sector may increase the bank value and hence improve the resilience of banks to negative shocks and avoid adverse budgetary consequences for governments, which often bear a significant part of the costs of bailouts. Nonetheless, banks may not fulfil adequately their capital requirements and hence their risk may not decrease as expected. 4 Data and research design We build a new database on bank risk taking, bank regulation, franchise value and other bank specific characters. Data used is gathered from three sources: the official bulletins of the TSE, the annuals reports of the Council of Capital Market, and the annual reports of the Tunisian listed banks. The daily data of stock prices, the TSE index, the amount of dividend and the number of shares were collected from TSE reports as well as the annual reports of banks activity through the website of the Professional Association of Tunisian Banks. Finally, we referred to the circular published by the CBT and which is dealing with the implementation of solvency ratio in Tunisia so as to deduce the capital adequacy variable. In this paper, we mainly use daily banks stock returns of the ten Tunisian commercial banks that are listed on the TSE. The period covers ten years and spans from 1999 to We choose the period because it includes the year 1999 when the international standard of capital requirements was implemented in Tunisia. We use also banks balance sheet and income statements data. 4.1 Bank risk taking To capture bank s risk taking, we use: 1 the bank insolvency risk 2 the bank s stock return volatility. Furthermore, to confirm our results, we add two other variables by considering whether the bank s stock return volatility is due to: 3 the bank specific risk, or 4 the market risk. To account for the bank default risk, we use the z-score which indicates the distance from insolvency. Specifically, the z-score points out the number of standard deviations that the bank s return on assets has to drop below its expected value before equity are depleted

8 96 A. Belanes and A. Ben Hajiba (Hannan and Hanweck, 1988; De Nicolo, 2000; Laeven and Levine, 2009). A higher z-score reveals that the bank is more stable and less insolvent. z = ( ROA + CAR)/ σ ( ROA) (1) The z-score is the inverse of the probability of insolvency. The bank is insolvent when losses surmount equity (Equity <- Profits). The probability of insolvency can be expressed as prob(car <- ROA); CAR (capital-asset-ratio) = Equity / Total Assets. While the z-score has been used widely in the financial and non-financial literature, it is subject to several caveats. First, it considers only the first and second moment of the distribution of profits and ignores the potential skewness of the distribution (De Nicolo, 2000). Secondly, the z-score relies on accounting data whose quality might vary with the degree of institutional development (Leuz et al., 2003). This, however, should bias the results against finding a significant relationship between regulations and bank risk. Alternatively, as a proxy of risk taking, we also consider the volatility of the bank equity returns in order to test our results. This equals the annualised volatility of daily equity returns during , which is also used by Saunders et al. (1990), Demsetz and Strahan (1997), Esty (1998), and Laeven and Levine (2009). The results are robust to estimating equity volatility over different periods. Additionally, we confirm our results when splitting the whole volatility of equity returns into volatility due either to the specific risk or to the market risk. In fact, in portfolio theory-based approaches (Kahane, 1977; Koehn and Santomero, 1980; Kim and Santomero, 1988), the volatility of equity returns, which is called the total risk, is not a good measure of equity risk. Thus, it would make sense to measure the equity risk in terms of how it moves with the market and the bank s characteristics. Theoretically, according to the CAPM, most of stock variance is bank specific and hence a specific and diversifiable risk while less is market specific and hence systematic and non-diversifiable risk. To calculate the bank-specific risk and the market risk, we apply the capital asset pricing model as follows: R = β + β R + ε (2) it 1 2 Mt it i and t indicate the bank and the period. R it is the daily stock return of the bank; R Mt is the daily stock return of the aggregate TSE index; and ε it is the term error. The bank specific risk is defined as the standard deviation of the residual term ε it; and the market risk is cov ( Rit ; RMt ) given by β 2 = which measures the sensibility of the bank s stock to the var ( RMt ) market. 4.2 Bank regulations To underline the Tunisian bank regulation, we use two variables: the capital regulation (CAPR) and the franchise value (FRVA). Tobin s Q is a good proxy for the health of the individual banking firm and hence for franchise value (Keeley, 1990). Market value of equity + Book value of leverage Q = Book value of total assets

9 Regulation and risk taking in the banking industry 97 Banks with higher Q values have lower default risk, less total and bank specific risk (Keeley, 1990; Demsetz and Strahan, 1997; Anderson and Fraser, 2000). To analyse the effect of capital regulations, namely the capital adequacy ratio, we add a dummy variable, CAPR, which takes the value of one if the observation takes place after 1999 and zero otherwise. In 1999, the CBT implemented the minimum capital adequacy standard and increased the minimum capital requirements versus their risk-weighted assets to 8%, along the lines proposed by the Basel Committee on Banking Supervision. 4.3 Other bank control variables Two control variables are used, namely the bank size (SIZE) and the frequency of the shares exchange (FREQ). To assess the bank size (SIZE), we calculate the natural logarithm of the book value of total assets. Larger banks are argued to be able to diversify their risk (Demsetz and Strahan, 1997). Besides, larger banks have better access to capital markets; they may have more flexibility to cope with unexpected liquidity shortfalls. Thus, the size is expected to negatively influence the bank risk taking. We proxy the frequency of trading of the share through the variable FREQ which is the average daily volume of shares divided by the total number of outstanding shares. It is a proxy for the speed at which new information is reflected in stock prices (Anderson and Fraser, 2000). Since the speed can be positively associated with the variance of assets, liabilities and off-balance sheet portfolios (Demsetz and Strahan, 1997), it is expected that the frequency will positively influence the bank risk taking. 4.4 Econometric modelling We apply panel data to test the impact of capital regulation and franchise value on bank risk taking. More specifically, we test the following equation: Risk = α + α CAPR + α FRVA + α SIZE + α FREQ + ε (3) it 0 1 it 2 it 3 it 4 it it Risk is the proxy for the bank risk taking; CAPR indicates the capital regulation implementation; FRVA the franchise value; SIZE the bank size; FREQ the frequency of the exchange (FREQ); ε it the term error; i = the bank; and t = the year. Empirical work on the effect of capital regulation and franchise value on bank risk-taking can potentially suffer from two sources of inconsistency: omitted variable and endogeneity biases. With this in mind, we ought to combine cross-section and time-series data and use panel data estimators. It is worth combining cross-section and time-series data for at least three reasons. First, pure cross-section regression leads to biased estimates because the bank-specific error term is likely to contain unobserved bank effects. The assumption that independent variables are not correlated with error term εit is then violated. Besides, the use of panel data provides a wealth of information ignored in cross-sectional studies; and hence more variability and less colinearity among the variables (Baltagi, 1995). Furthermore, by increasing the sample size, we gain more degrees of freedom and hence more efficiency. This is particularly relevant when the number of explanatory variables is relatively large and the number of banks is small; which is our case here. Beyond these assumptions, we apply panel data to test the impact of capital regulation and the franchise value on bank risk taking within the Tunisian context. The

10 98 A. Belanes and A. Ben Hajiba heterogeneity of banks is captured by including bank-specific effects. The Hausman test is then used to verify whether theses specific effects are random or fixed (Hausman, 1978). 5 Empirical results Three sets of results will be displayed and discussed in this section: the descriptive statistics of general features of the banks, namely the franchise value, the bank size and the frequency of their transactions, the descriptive statistics of the four measures of bank-risk taking and finally the empirical influence of capital regulation and franchise value on risk taking within the Tunisian banking industry. 5.1 Summary statistics Table 1 presents a summary of statistics that describe the mean and standard deviations for all variables employed in our analysis, for the combined sample period, and in the periods before and after capital regulations. The full sample was divided into two subsamples before and after the implementation of the capital adequacy requirement in 1999 according to the Basel I. Table 1 points out that the average total assets have increased from 1, to 2, million TND respectively in the first and the second subsamples. The difference is statistically significant. The implementation of the capital regulation has increased the size of the Tunisian banks; which normally strengthen banks capital and improve their resilience to negative shocks. Table 1 also shows that the transactions of shares exchange become more frequent in the second sub-period, spreading from 2000 to Perhaps, the advances of technology, the enhancement of the used techniques and the promotion of the equity market incite the investors to usually seek the privileged information and to profit from the potential opportunities. The market transactions become therefore more frequent. Table 1 Descriptive statistics Variables t-statistics Total assets Average 1,425,801 2,186,701 1,945, *** Standard deviation Franchise value Average *** Standard deviation % frequency of trading Average Standard deviation N Notes: *, ** and *** significance at 10%, 5% and 1% levels; t-statistics test the means difference. Total assets in 1,000 TND.

11 Regulation and risk taking in the banking industry 99 Nonetheless, Table 1 puts in evidence the decrease of the franchise value after tightening capital requirements. Two main causes can explain such a decrease. First, this decline can be explained by the strengthening of the rivalry during the period Competition between Tunisian banks has become more tough and harder due to the deregulation of Tunisian banking sector, the progressive financial liberalisation and the ongoing economic integration as a whole. Secondly, capital regulation might have restricted banks activities, adversely affecting credit expansion and credit growth. Regulatory restrictions on bank activities may increase net interest margins or overhead costs. Table 2 reports the average variation of the four bank risk taking proxies across all banks. The full sample was divided into two subsamples before and after 1999; that is prior and subsequent to the implementation of the capital adequacy requirement, namely the solvency ratio of 8%. Table 1 puts in evidence that the volatility of equity returns has declined in average from a to after 1999; which might predict the decrease of fragility across banks. The volatility is measured by the annualised volatility of daily equity returns (Saunders et al., 1990; Demsetz and Strahan, 1997; Esty, 1998; Laeven and Levine, 2009). Our findings are confirmed when splitting the whole volatility of equity returns into volatility specific either to the banks characteristics or to the market status (Kahane, 1977; Koehn and Santomero, 1980; Kim and Santomero, 1988). In other words, our results are approved of when the total risk is divided into bank-specifir risk and market risk. However, the implementation of capital regulation has decreased the market risk of about more than 65% while the bank-specific risk of only 7%. In fact, the bank specific risk has decreased in average from to while the market risk has declined in average from to in Thus, it would make sense to measure the equity risk in terms of how it moves with the market and the bank s characteristics. Table 2 Total risk Bank risk taking t-statistics Average Standard deviation Bank-specific risk Average Standard deviation Market risk Average *** Standard deviation Downside risk (z-score) Average * Standard deviation No. of observations Notes: *, ** and *** indicate statistical significance at the 10%, 5% and 1% level; t-statistics test the means difference.

12 100 A. Belanes and A. Ben Hajiba The important decrease of market risk is specifically due to the strengthening of prudential regulations and banking supervision which have started since Banks can fulfil their capital requirement ratios by reducing their risk-weighted assets or by increasing their capital; and hence the decline of the bank specific risk. However, by liberalising the financial sector and opening it to foreign financial institutions, banks can undertake additional risks through other subordinates activities; and perhaps manage them with highly sophisticated tools. Besides, the introduction of capital adequacy rules may cause a shift from making loans to the private sector to providing credit to the public sector (Ben Naceur and Kandil, 2009). In sum, a bank facing binding capital rules has a higher incentive to decrease equity volatility. Our estimates of bank z-scores display similar findings. Banks with higher z-scores display lower equity volatility. Table 2 also summarises the average of bank z-scores before and after 1999, when capital regulation was first enacted. The averaged z-score increases from in the first subsample to in the second subsample. The z-scores indicate that profits have to fall by more than 24 times their standard after 1999 to deplete bank equity, but profits only need to fall by about 19 times their standard deviation before 1999 to eliminate bank equity. However, Table 2 points out a wide variation in bank fragility across banks. A high value of z-score reflects a low level of downside risk. Therefore, the higher the z-score, the more stable the banks are. The endorsement of capital regulation has then decreased the bank risk of insolvency. Therefore, if regulators are mainly concerned about reducing the insolvency risk of banks, introducing capital rules may limit bank risk taking. 5.2 Determinants of bank risk taking Table 3 summarises the main results on the relationship between capital regulation and franchise value and bank risk taking. For each risk taking measure, we test the null hypothesis that an individual effect does not exist using the F-test. The null hypothesis is rejected for the market risk and the downside risk at the level of significant of 1% and 5% respectively. However, we accept the hypothesis of homogeneity of the sample for the total risk and the specific risk. Therefore, we estimate the regression equations using the ordinary least squares for the total risk and the market risk. For the market risk and z-score, we further test the null hypothesis that the individual effect is correlated with explanatory variables using Hausman test. The null hypothesis was rejected for the market risk measure and hence the fixed effects model is applied. The null hypothesis was in contrast for the insolvency risk measure. That is why we choose the random effect model. Adjusted R² varies between 30.6% and 73.4%. The models undertaken respectively the downside risk and the market risk offer the higher and the lower explanatory power. Table 3 puts in evidence the negative correlation between the variable CAPR and the market risk, suggesting that the implementation of the capital requirement reduced the equity volatility due to the market status. However, the setting-up of the international standard of capital requirements is positively associated with the bank default score. But above all, binding regulation has decreased the equity returns volatility of banks and increased their z-scores. Banks with highly volatile equity returns are expected to follow risky strategies and invest in various opportunities widely risky. Unregulated banks undertake therefore additional risks compared to regulated banks. Besides, banks with higher z-scores have less downside risk; and hence less risky. Thus, unregulated banks

13 Regulation and risk taking in the banking industry 101 have lower z-scores and take more risk than regulated banks. The introduction of capital adequacy rules has then strengthened bank capital and enhanced their soundness; which might improve the resilience of banks to negative shocks. Regression results are in line with previous descriptive statistics reported below which confirm that equity return volatility has on average declined after tightening capital requirements. These results are also consistent with the theoretical conclusion of Kim and Santomero (1988) and Furlong and Keeley (1989) and previous empirical results including Anderson and Fraser (2000), Murinde and Yaseen (2006), and Yilmaz (2009). Table 3 Determinants of bank risk taking Variables Total risk Market risk Specific risk Downside risk Intercept ( 0.53) CAPR ( 0.19) FRVA (0.89) FREQ ( 0.47) SIZE (0.60) ( 1.06) 0.585*** ( 5.97) 1.422** ( 2.10) ( 1.22) 0.269* (1.81) ( 0.53) ( 0.07) (0.93) ( 0.48) (0.56) ( 1.05) 7.293** (2.14) ( 0.05) * ( 1.65) (0.96) F-statistic Adjusted R No. of observations Notes: *, ** and *** indicate statistical significance at the 10%, 5% and 1% level; t-statistics in parentheses. CAPR is a dummy variable that takes 1 if the observation takes place after 1999 and 0 otherwise; FRVA is the sum of market value of common equity plus the book value of liabilities divided by the book value of assets; FREQ is the average daily volume of shares divided by the total number of outstanding shares; and SIZE is the natural logarithm of the book value of total assets. Similarly, Table 3 shows the negative impact of franchise value on bank risk taking. The negative correlation between the franchise value and the market risk suggests that the franchise value can help reducing excessive risk taking at commercial banks. These findings concur with the findings of Marcus (1984), Keeley (1990), Salas and Saurina (2003), Konishi and Yasuda (2004), and Gonzláez (2005) among others. If future profits are lower, a bank has a smaller incentive to avoid default. The bank has less to lose in the event of bankruptcy. Therefore, increasing risk, and hence the probability of default, is less costly for the bank under binding regulation. However, if the bank has prosperous future opportunities, it is prohibitively costly to undertake additional risks and put at stake the bank. Table 3 exhibits a negative correlation between the frequency of shares trading and the downside risk. This result implies that banks whose equity is more frequently traded are exposed to a higher level of risk; and hence have a lower z-score. Finally, Table 3 reports a positive influence of the size on bank risk taking; meaning that larger banks take on more risks than their smaller counterparts. Perhaps, Tunisian banks adopt more risky loan portfolios and operate with more leverage as they are more vigilant,

14 102 A. Belanes and A. Ben Hajiba resilient and more able to manage risks and escape the negative shocks. Such result is consistent with the results of Demsetz and Strahan (1997) and Anderson and Fraser (2000). Last but not least, empirical findings clearly bring forth that the intensity of such decrease varies with the risk taking measurement. Capital regulation allows reducing the market risk and decreasing the insolvency risk but has no effect on the specific risk of the bank; while the franchise value has a negative effect on only the market risk of the bank. Thus, it is worthwhile to investigate and quantify all sides of bank risk taking; which is a wide and a very complex concept. It would make sense to evaluate bank risk taking not only through the risk default but also through the equity volatility due to the bank characteristics and the equity volatility because of the market circumstance. Overall, our study supports the aims of Basel III which claims that capital allocation should be more risk sensitive and that the whole bank risk taking ought to be deeply investigated and quantified. 6 Concluding remarks and future research The aim of this study is to analyse the effect of regulation on the risk taking within the Tunisian banking industry. In this paper we investigate the impact of both the capital adequacy requirement and the charter value, which are the potential drivers of bank system stability, on the bank risk taking. Four measures of bank risk taking are investigated, namely the market risk, total risk, specific risk and insolvency risk. The combined evidence provides a menu of important determinants of bank risk taking to guide policymakers towards upgrading quality and enhancing the stability of an industry that is considered by many to be the core of economic development. Empirical findings reveal that the impact of regulation on bank risk taking varies with the measurement. Our study puts in evidence that the implementation of the international standard of capital requirements in 1999 reduces the market risk and the insolvency risk of commercial Tunisian banks. In this regard, capital requirements, for instance the solvency ratio, help reducing excessive bank risk taking and preserve the resilience and the soundness of the Tunisian banking industry. Regression results show as well that the franchise value can help reduce excessive risk taking at commercial banks. Banks with high charter value have an incentive to avoid high-risk choices that might reduce their competitive advantages and may trigger a drop, even a bankruptcy. Since the franchise value has as main source the market and the competition may deteriorate it, banks tend to reduce their market risk to keep it. Furthermore, it seems that large banks are more sensitive to the market and undertake more risks. Finally, banks whose shares are more frequently traded are exposed to a higher level of downside risk. In sum, recent economic crises have emphasised the importance of bank regulations to guard against the variety of risks that banks can face, mainly the financial and operational risks. But above all, it is indispensable to find out new supervisory measures so as to promote such a vigilant and resilient banking sector. For instance, other features of risk taking are worth investigating in further researches, namely the credit risk, the operational risk and the leverage ratio. It is also interesting to shed some light in the combined effect of corporate governance and capital regulation on bank risk taking.

15 Regulation and risk taking in the banking industry 103 In fact, previous literature based upon portfolio-theory views the bank as a whole. It fails therefore to put in evidence agency conflicts among shareholders and bankers. Bank managers are not necessarily maximising shareholder s value. While such agency conflicts are widely explored in the corporate finance literature, very few papers deal with this issue regarding banking firms. References Anderson, R.C. and Fraser, D.R. (2000) Corporate control, bank risk taking, and the health of the banking industry, Journal of Banking and Finance, Vol. 24, No. 8, pp Baltagi, B.H. (1995) Econometric Analysis of Panel Data, John Wiley & Sons, New York. Ben Naceur, S. and Kandil, M. (2009) The impact of capital requirements on banks cost of intermediation and performance: the case of Egypt, Journal of Economics and Business, Vol. 61, No. 1, pp Besanko, D. and Kanatas, G. (1996) The regulation of bank capital: do capital standards promote bank safety?, Journal of Financial Intermediation, Vol. 5, No. 2, pp Blum, J. (1999) Do capital adequacy requirements reduce risk in banking?, Journal of Banking and Finance, Vol. 23, No. 5, pp Buser, S., Chen, A. and Kane, E. (1981) Federal deposit insurance, regulatory policy and optimal bank capital, Journal of Finance, Vol. 36, No. 1, pp Calem, P. and Rob, R. (1999) The impact of capital based regulation on bank risk taking, Journal of Financial Intermediation, Vol. 8, No. 4, pp De Nicolo, G. (2000) Size, charter value and risk in banking: an international perspective, International Finance Discussion Papers, No. 689, Board of Governors of the Federal Reserve System, Washington, DC. Demsetz, R. and Strahan, P. (1997) Diversification, size, and risk at bank holding companies, Journal of Money, Credit, and Banking, Vol. 29, No. 3, pp Dewatripont, M. and Tirole, J. (1995) The Prudential Regulation of Banks, MIT Press, Cambridge, MA. Esty, B. (1998) The impact of contingent liability on commercial bank risk taking, Journal of Financial Economics, Vol. 47, pp Fonseca, A.R. and González, F. (2010) How bank capital buffers vary across countries: the influence of cost of deposits, market power and bank regulation, Journal of Banking and Finance, Vol. 34, No. 4, pp Furlong, F. and Keeley, M. (1989) Capital regulation and bank risk taking: a note, Journal of Banking and Finance, Vol. 13, No. 6, pp Galloway, T.M., Lee, W.B. and Roden, D.M. (1997) Banks changing incentives and opportunities for risk taking, Journal of Banking and Finance, Vol. 21, No. 4, pp Gonzláez, F. (2005) Bank regulation and risk taking incentives: an international comparison of bank risk, Journal of Banking and Finance, Vol. 29, No. 5, pp Gorton, G. and Rosen, R. (1995) Corporate control, portfolio choice and the decline of banking, Journal of Finance, Vol. 50, No. 5, pp Hannan, T.H. and Hanweck, G.A. (1988) Bank insolvency risk and the market for large certificates of deposit, Journal of Money, Credit, and Banking, Vol. 20, No. 2, pp Hausman, J.A. (1978) Specification tests in econometrics, Econometrica, Vol. 46, No. 6, pp Homölle, S. (2004) Bank capital regulation, asset risk, and subordinated uninsured debt, Journal of Economics and Business, Vol. 56, No. 6, pp

16 104 A. Belanes and A. Ben Hajiba Kahane, Y. (1977) Capital adequacy and the regulation of financial intermediaries, Journal of Banking and Finance, Vol. 1, No. 2, pp Keeley, M.C. (1990) Deposit insurance, risk, and market power in banking, American Economic Review, Vol. 80, No. 5, pp Kim, D. and Santomero, A.M. (1988) Risk in banking and capital regulation, Journal of Finance, Vol. 43, No. 5, pp Koehn, M. and Santomero, A.M. (1980) Regulation of bank capital and portfolio risk, Journal of Finance, Vol. 35, No. 5, pp Konishi, M. and Yasuda, Y. (2004) Factors affecting bank risk taking: evidence from Japan, Journal of Banking and Finance, Vol. 28, No. 1, pp Laeven, L. and Levine, R. (2009) Bank governance, regulation and risk taking, Journal of Financial Economics, Vol. 93, No. 2, pp Leuz, C., Nanda, D.J. and Wysocki, P. (2003) Investor protection and earnings management: an international comparison, Journal of Financial Economics, Vol. 69, No. 3, pp Marcus, A.J. (1984) Deregulation and bank financial policy, Journal of Banking and Finance, Vol. 8, No. 4, pp Milne, A. (2002) Bank capital requirement as an incentive mechanism: implications for portfolio choice, Journal of Banking and Finance, Vol. 26, No. 1, pp Murinde, V. and Yaseen, H. (2006) The impact of Basel Accord regulations on bank capital and risk behavior: evidence from the MENA region, Working paper, University of Birmingham. Niu, J. (2008) Can subordinated debt constrain banks risk taking?, Journal of Banking and Finance, Vol. 32, No. 6, pp Rochet, J.C. (1992) Capital requirements and the behavior of commercial banks, European Economic Review, Vol. 36, No. 5, pp Salas, V. and Saurina, J. (2003) Deregulation, market power and risk behavior in Spanish banks, European Economic Review, Vol. 47, No. 6, pp Santos, J.A.C. (1999) Bank capital and equity investment regulations, Journal of Banking and Finance, Vol. 23, No. 7, pp Saunders, A., Strock, E. and Travlos, N.G. (1990) Ownership structure, deregulation, and bank risk taking, Journal of Finance, Vol. 45, No. 2, pp Silva, N. (2007) Capital regulation and bank risk taking: completing Blum s picture, Working Papers No. 416, Banco Central de Chile. Yilmaz, E. (2009) Capital accumulation and regulation, The Quarterly Review of Economics and Finance, Vol. 49, No. 3, pp