Market Structure and Competitive Conditions: A comparative analysis of Islamic and conventional banking. Rima Turk Ariss* Lebanese American University

Transcription

1 Market Structure and Competitive Conditions: A comparative analysis of Islamic and conventional banking Rima Turk Ariss* Lebanese American University January 5, 2009 * Please address all correspondence to Rima Turk Ariss, Lebanese American University, Business School, P.O.Box: , Beirut, Lebanon; rima.turk@lau.edu.lb; X1644 1

2 Market Structure and Competitive Conditions: A comparative analysis of Islamic and conventional banking Abstract We perform a comparative analysis of banking structures and competitive conditions prevailing in both Islamic and conventional global banking markets, and investigate differences in profitability levels. We select 13 countries where both segments of the banking industry prevail over the period Univariate statistics show that Islamic banks allocate a significantly greater share of their assets to financing or loans compared to conventional banks; they are significantly better capitalized and achieve significant higher returns. We compute different measures of market structure and find that the global Islamic banking market is less competitive compared to the conventional banking segment. Multivariate regression analyses show that bank profitability is significantly increased where market power prevails, possibly explaining the higher return levels observed for Islamic banks. JEL Classification Numbers: D4, G21, L1, N25 Keywords: Market Structure, Bank Competition, Panzar and Rosse methodology, Lerner Index, Islamic banks 2

3 1. Introduction Competition in banking has intensified over the past decades and is putting increasing pressure on bank returns. Major financial institutions are strategically entering new markets and/ or offering a diverse spectrum of products and services to stay abreast of recent developments and boost their profitability. Since 1975, Islamic banking started expanding in several countries and it is now widely recognized as a viable mode of financing. Islamic banks have proliferated in the Far East and the Arabian Gulf and a large number of banking firms have diverted some of their operations away from conventional practices by setting up Islamic windows or establishing fully-fledged Islamic banks. There are now about 270 Islamic financial institutions which operate from China to Europe and USA, including banks, mutual funds, mortgage companies and Takaful or insurance firms. Countries like Malaysia and Bahrain are striving to be established as regional hubs for providing Islamic financial services and Britain has announced plans to turn London into the world centre of Islamic finance (Financial Times, June 28, 2007). Notwithstanding the religious underpinnings of the provision of Islamic financial services, it is rational to assume that there must be an incentive for banking institutions (conventional or otherwise) to choose to cater for the needs of the Islamic banking market clientele. International banks such as Citigroup, BNP Paribas, HSBC, ABN Amro Bank and others have recognized the growth potential of this new segment of the financial industry and have long joined the drive to better serve customers with different financial needs and expectations. The present study analyzes competitive conditions in both Islamic and conventional banking assesses differences in bank profitability. From a structural point 3

4 of view, Islamic banks operate alongside with conventional banks and a market for Islamic financial services has developed in parallel with each conventional banking market in several countries. Deficit and surplus units have the option to use the services provided by either segment of the banking industry. While it is not uncommon for bank clientele to establish relationships with both types of institutions, Islamic banks have limited dealings with conventional banks because they are prohibited from transacting based on a predetermined rate of interest. In this respect, it is reasonable to assume that the two legs of the banking market are distinct and that inter-industry linkages are rather limited. From a regulatory perspective, central banks of countries where both Islamic and conventional banks operate have issued special circulars and promulgated new laws for the regulation and supervision of Islamic banks. Because each banking segment has its own specificities, the prevailing structures are likely to differ across the two markets. Compared to conventional banking, Islamic banking is relatively at its early stages of development and it is likely that a higher degree of market power prevails in the industry. If market players in the Islamic finance industry do command a higher degree of market power compared to their peers, are profitability conditions also significantly different? Is the embryonic Islamic banking a more lucrative business compared to the more mature conventional banking industry? This study investigates competitive conditions in Islamic and conventional banking, and assesses their implications on bank profitability. The literature on market structure and competitive conditions and for conventional banking segments is voluminous. Market structure in banking is relevant for at least two reasons. On the one hand, the degree of market power is believed to bear serious implications on financial stability. After the seminal article by Keeley (1990), 4

5 many studies have shown that competition encourages moral hazard in banking (Hellmann, Murdock, and Stiglitz 2000; Jimenez, Lopez, and Saurina 2007), while a counter trend provides theoretical predictions and empirical evidence that more market power may result in higher bank risk (Stiglitz and Weiss 1981; Koskela and Stenbacka, 2000; Boyd and De Nicolo, 2005; Schaeck, Cihak, and Wolfe, 2006). On the other hand, the literature has also shed the light on the implications of market structure on bank performance and efficiency 1, and other researchers show that banks tend to hold higher capital when operating in a more competitive environment (Schaeck and Cihak 2007.). This paper differs from previous work on various stands. First, it is assumed that there is a global market for Islamic financial services that is distinct from conventional banking and which is not geographically limited to one country. With the emergence of Islamic banks competing side by side with conventional banks, two parallel markets have developed and it is now possible to examine differences in banking structures across the two segments of the industry. Countries where both types of banks operate are selected and two distinct samples of banks are formed, one is conventional and the other is Islamic. No previous study has compared the market structure of both types of banking segments of the industry. Second, previous research on the market structure prevailing in related countries has used traditional measures of concentration and the contested Panzar and Rosse (PR, 1987) H-statistic. Abdul Majid and Sufian (2007) report that market conditions are monopolistically competitive in the Islamic financial industry in Malaysia using traditional measures of concentration and the PR methodology. Al-Muharrami, 1 For a detailed review of the literature on banking efficiency, see Berger and Humphrey (1997), and for an updated review of the efficiency literature, see Berger and Mester (2003). 5

6 Matthews, and Khabari (2006) also use traditional concentration ratios and the H-statistic and find that competitive conditions in banking vary across Gulf Cooperation Council countries. Turk-Ariss (2008) similarly use the PR model to evaluate competitive conditions in Middle East and North Africa conventional banking. This study employs a variety of measures to proxy for the degree of market power in countries where both Islamic and conventional banks operate, including traditional concentration measures, the PR H-statistic, as well as the Lerner index. Third, the analysis goes beyond investigating the degree of competition to examining differences in bank profitability across the two market segments. The aforementioned studies by Abdul Majid and Sufian (2007), Al-Muharrami, Matthews, and Khabari (2006), and Turk-Ariss (2009) only assess the prevailing degree of market power, and do not examine its implications on banks operations. The rest of the paper is structured as follows. Section 2 provides a background overview on Islamic finance; section 3 presents the methodology; sections 4 and 5 discuss the data and the empirical findings respectively, and section 6 concludes. 2. Background on Islamic Finance The Islamic financial system rests on the promotion of entrepreneurship, the protection of property rights and the transparency and sanctity of contractual obligations. It is based on a commerce law known as fiqh al-mu amalat which considers issues of social justice, equity and fairness in all business transactions. Islamic finance is governed by the precept of Shari a or the Islamic legal code, whereby a commercial transaction is permissible as long as it is free from Riba (interest), gharar (uncertainty), maisir 6

7 (gambling), non-halal (prohibited) food and alcoholic drinks and immoral activities such as prostitution, alcohol, nightclubs and narcotics. Because of its socially responsible and ethical underpinnings, the new class of Islamic investments is appealing to both Muslim and non-muslim who seek to invest in socially responsible products 2. While Islamic banks cannot pay or receive a pre-determined rate of interest on deposits and loans, capital is instead rewarded based on the rate of return on investments. Islamic financing services are developing phenomenally around the world, and Islamic financial institutions managed to efficiently compete with their conventional peers in countries that do not generally support riba-free environments. Recent figures indicate that global Shari a assets under management stand at about $500 billion (Financial Times, June 28, 2007). While the size of the Islamic financial industry is at very low levels when compared to the $1.5 trillion of assets for some of the largest commercial banks (including Barclays Bank Plc, UBS A.G, HSBS, Citigroup, BNP Paribas and others), its rate of growth has been impressive, averaging around 15% over the past three decades. The development of the Islamic finance industry was matched over the past few years by progress on the legal, accounting and auditing, regulatory and governance fronts. An architecture of institutions has developed to fuel the growth and development of the industry. In 1991, the Accounting and Auditing Organization for Islamic Financial Institutions (AAOIFI) was mandated to prepare accounting, auditing, governance, ethics and Shari'a standards. In 2002, the Islamic Development Bank based in Jeddah took the lead in establishing the International Islamic Financial Market (IIFM) in April, the 2 London has become a major trading centre for Islamic funds and a quarter of all Islamic banking business in Malaysia is conducted by non-muslims (Bahrain Tribune, 29/06/07). 7

8 Liquidity Management Center (LMC) in July and the Islamic Financial Services Board (IFSB) in November of the same year. The IIFM has the mandate to take part in the establishment, development, self-regulation and promotion of Islamic capital and money markets. The LMC was established with the purpose to facilitate the liquidity mismatch of Islamic financial institutions through quality financial instruments structured in accordance with Shari a principles. The IFSB is an active international standard setting body with a mission to ensure soundness and stability of the Islamic financial services industry by developing prudent and transparent standards and codes 3. More recently in 2005, the International Islamic Rating Agency was set up to assist Islamic financial institutions and instruments in gaining recognition locally and internationally by adhering to greater standards of disclosure and transparency. Early studies on Islamic finance focused primarily on the conceptual viability and sustainability of interest-free financing 4 and later work addresses different problems faced by Islamic banks and on the development of an Islamic financial system 5. The finance and economic growth literature has established a positive relationship between financial sector development and economic growth, although the direction of causality remains an issue of debate 6. In countries where Islamic finance is gaining importance, capital markets are relatively underdeveloped and banks remain the major source of funding for individuals and firms. The recent developments of Islamic financial services 3 Officially inaugurated on November 3, 2002, the IFSB was initiated by a group of senior officials of central banks and monetary authorities of various countries, together with the support of the Islamic Development Bank, the International Monetary Fund, and the Accounting and Auditing Organization for Islamic Financial Institutions. 4 See for example Khurshid (1981) and Karsen (1982) 5 See for example Omar and Iqbal (2000) and Dar and Presley (2000) 6 Levine (2004) provides an excellent review on the research in this area. 8

9 are likely to channel previously idle resources to productive uses and contribute to economic development. In the next sections, we examine differences in Islamic and conventional banking structures. In principle, Islamic banks operate under the precepts of Shari a, with the objective of ensuring social and economic justice rather than being guided by the principle of profit maximization. Classical industrial organization theory assumes profit maximization, and it can be argued that New Empirical Industrial Organization techniques such as PR H-statistic and Lerner index cannot be applied to Islamic banks. However, the social objective of Islamic finance can mainly be achieved through the promotion of risk-sharing financing techniques. A close look at Islamic banks balance sheet shows that credit-based financing (murabaha or cost-plus sales) is the dominant form of uses of funds, while profit-and-loss (or risk-sharing) financing in the form of mudaraba and musharaka financing represent on average less than 10% of assets. This, in fact, may represent Achilles heel for Islamic banks which have been criticized over the past three decades for not abiding by the social aspect of their mission, but rather seeking quick more and secure profit through murabaha financing. In line with other studies, we assume that Islamic banks behave as profit-maximizing firms. They also compete with conventional banks to provide financial services to their clientele. Under this assumption, the derivation of the H-statistic and Lerner index from the first-order condition of the bank s maximization problem can be applied to the context of Islamic banking. 3. Methodology 9

10 We aim at analyzing the banking structures of both Islamic and conventional segments of the industry and evaluate competitive conditions in each segment of the industry. The literature on market structure is inconclusive regarding the best measure of the degree of competition. Different indicators of market power are estimated, and multivariate analysis is conducted to investigate differences in profitability levels across the two types of banks. 3.1 Traditional Measures of Concentration Traditional measures of concentration include concentration ratios and the Herfindahl-Hirschman Index (HHI). We use the n-bank concentration ratio, in particular C3 and C5 ratios, which show the concentration ratios of the biggest 3 and 5 banks respectively according to the share of their assets, deposits and loans in the banking sector. Concentration ratios, however, do not consider information about the remaining banks. The HHI is index calculated by adding up the squares of the market shares of all banks, using each of total assets, deposits and loans. While a number of studies use measures of concentration such as the HHI or n- firm concentration ratio to indicate market power, these have been shown to be ambiguous indicators of competitiveness (e.g. Berger, Demirguc-Kunt, Levine, and Haubrich 2004; Beck, Demirguc-Kunt, and Levine 2006). Other studies employ the PR H-statistic to assess the degree of competitiveness in banking (e.g., Claessens and Laeven 2004, Schaeck, Cihak, and Wolfe 2006, Molyneux and Nguyen-Linh 2008) and the Lerner index (Jimenez, Lopez, and Saurina, 2007; Berger, Klapper, and Turk-Ariss, 10

11 2009). In this paper, we compute traditional measure of market structure in addition to estimating the H-statistic and the Lerner index. 3.2 The H-statistic The PR methodology rests on the estimation of the following reduced-form revenue equation on pooled samples for each country: ln(tr it ) = α + β 1 ln(w L,it )+ β 2 ln(w F,it )+ β 3 ln(w K,it ) + γ 1 ln(y 1,it ) + γ 2 ln(y 2,it ) + ε it (1) The dependent variable TR it indicates total revenues measured by the ratio of interest and non-interest revenues to total assets, following Shaffer (1982), Nathan and Neave (1989) and Casu and Girardone (2006) 7. Three input prices are included in equation (1), W L,it as the cost of labor proxied by the ratio of personnel expenses to total assets, W F,it as the cost of funds represented by the ratio of interest expenses to total deposits and W K,it as the cost of fixed capital calculated as the ratio of other operating and administrative expenses to total assets. Consistent with Molyneux, Thornton, and Lloyd- Williams (1996), Bikker and Haaf (2002), Gelos and Roldos (2004) and Claessens and Laeven (2004), other bank-specific control variables are included in the analysis. Y 1,it and Y 2,it represent the ratio of equity to total assets and net loans to total assets respectively, and control for business and portfolio mix of the bank. While there is no expectation about the sign on total assets, the results of the estimation would provide information whether banks face economies or diseconomies of scale. The subscripts i and t refer to bank i operating at time t. 7 For Islamic banks, the category of loans is substituted by financing activities and interest revenues are called financing revenues. Similarly, the interest expense item is labeled financing expenses. 11

12 The PR H-statistic is computed as the sum of the input price elasticities of total revenues. Panzar and Rosse show that the H-statistic can reflect the structure and conduct of the market to which the firm belongs and can be interpreted as follows. Under long run competitive equilibrium, an increase in input prices will lead to an equivalent increase in total revenues, and firms that cannot cover the increase in input prices will exit the market; therefore the H-statistic would be equal to one. By contrast, if the firm operates as a monopoly, the H-statistic will be negative since an upward shift in the marginal cost curve will be associated with a decrease in revenues. Finally, if the market structure is characterized by monopolistic competition, the H-statistic will lie between zero and one. Three assumptions are needed to apply the PR framework to the context of the banking industry. First, banks are single product firms that produce interest revenues using labor, capital and deposits as inputs (De Bandt and Davis, 2000); second, higher factor prices are not correlated with higher revenues generated by higher quality services; and the third assumption is about profit maximization and normally shaped cost and revenue functions (Gelos and Roldos, 2004). More importantly, banks should be observed from a long run equilibrium perspective. This is addressed by using a panel data specification and by testing that observations are in long-run equilibrium using the following model: ln(roa it ) = α + β 1 ln(w L,it )+ β 2 ln(w F,it )+ β 3 ln(w K,it )+ γ 1 ln(y 1,it ) + γ 2 ln(y 2,it ) + ε it (2) where ROA is the return on assets. The equilibrium statistic E is calculated as the sum of the input price elasticities, and the hypothesis that its value is 0 is tested where, if 12

13 rejected, the market is not in equilibrium. The intuition behind the test is that, in the long-run, return on assets is not related to input prices. 3.3 The Lerner index While the magnitude of the H-statistic can be interpreted as a direct indicator of the degree of competition, this proxy is contested as a continuous measure of the degree of competition. Panzar and Rosse show that the H-statistic is a decreasing function of market power in the case of a pure monopoly, but the interpretation of its value is less straightforward because the magnitude as well as the sign of the statistic may be of interest (PR, p. 446). Shaffer (2004 a and b) also casts doubts on the use of the H- statistic as a continuous measure of competition and concurs that the interpretation of the specific value of H may be ambiguous (2004 b, p.297). Another shortcoming of the H- statistic is that it is calculated on the national level and it cannot be used to assess the decisions of a bank on the firm level. Instead, market structure is better captured by a proxy that is allowed to vary at the bank level and over time like the Lerner index. For all these reasons, the Lerner index is the preferred measure of market structure in this study, although results for the H-statistic are also reported. The Lerner Index offers a direct measure of the degree of market power because it focuses on the pricing power apparent in the difference between price and marginal cost (Jimenez, Lopez, and Saurina 2007; Berger, Klapper, and Turk-Ariss, 2008). It is a more accurate indicator of market power than standard concentration measures because it measures market power by capturing the degree to which a firm can increase its marginal 13

14 price beyond marginal cost. The computation of the Lerner index requires the estimation of a translog cost function with bank fixed effects and time dummies as follows: ln Cost it β β W + ε β1 lnqit + lnqit + γ kt lnwk, it + φk lnqit lnwk, it + lnwk, it ln 2 k= 1 k= 1 k= 1 j= 1 = j, it it where Q it represents a proxy for bank output or total assets for bank i at time t, and W k,it are the three input prices defined above 8. Marginal cost is then derived as: MCTA it 3 Costit = β 1 + β2 lnqit + φk lnw Qit k = 1 k, it And the Lerner index is computed as: Lerner it = (P TAi t - MC TAit ) / P TAit In computing Lerner it, P TAit is the price of total assets proxied by the ratio of total revenues (interest and noninterest income) to total assets for bank i at time t, and MC TAit is the marginal cost of total assets for bank i at time t. The Lerner index ranges between 0 and 1. When P TA = MC TA, the Lerner Index is zero and the firm has no pricing power. A Lerner index closer to 1 indicates relatively weak price competition and hence market power for the firm. 3.4 Multivariate Analysis In a second-stage analysis, the Islamic and conventional banking samples are combined together and differences in profitability levels are examined. A dummy variable named Islamic is created and is set to 1 when the activities of the bank are Shari a-compliant. We run bank fixed effect regression with time dummies for a panel of 1,173 observations to analyze the relationship between profitability and market 8 A potential problem with the Lerner index as a measure of market power, as we calculate it, is that the cost of funds or ratio of interest expenses to deposits W 2, may itself embody market power in the deposit market. 14

15 structure, after controlling for bank and country differences. The baseline regression equation has the general form: Profitability = f(degree of Competition, Bank Controls, Islamic, Economic Development) (3) We use return on assets (ROA) and return on equity (ROE) as measures of profitability. Two proxies for market structure are included, the H-statistic and the Lerner index. Bank control variables include bank market share and size. We also add a dummy variable Islamic to distinguish between Islamic and conventional banks. Following Martinez & Repullo, (2008) and Berger, Klapper & Turk-Ariss (2009), a quadratic term is added to equation (3) to allow for a nonlinear relationship between measures of market structure and profitability in banking. All regressions include the natural logarithm of GDP per capita to control for the level of economic development. 4. Data Bank-level financial data on Islamic banks for the years is retrieved from the BankScope database provided by Fitch-IBCA (International Bank Credit Analysis Ltd) and is matched by data on conventional commercial banks that operate in the same countries as Islamic banks. The list of Islamic banks is checked against the website of each institution because some banks classified as Islamic under BankScope operate in fact as conventional banks. Only fully-fledged Islamic banks are considered, and countries where there is only one Islamic bank are dropped. The full sample includes 58 Islamic and 192 conventional banks operating in 13 different countries. Table 1 lists the number of banks and observations in each country. 15

16 [Table 1 about here] The figures indicate that, except for the regional Islamic banking hub Bahrain, the number of Islamic banks which operate in the sampled countries is very small compared to their peers, reflecting the embryonic stage of the Islamic banking industry. Descriptive statistics on the average loans to assets, equity to assets, return on assets and return on equity for Islamic (Panel A) and conventional (Panel B) banks in each of the 13 countries considered are provided in Table 2, and the data are summarized and grouped by year in Table 3 9. [Table 2 about here] [Table 3 about here] Despite growing interest in the Islamic financial services industry, Islamic banks are by far outnumbered by conventional banks in the profiled countries, and the average size of Islamic banks (in terms of total assets) is smaller than their peers. Over the sample period, the average loans to assets ratio of Islamic banks stood at 52.78% versus 43.96% for conventional banks and the equity to assets ratios for the two banking segments are and 12.42% respectively. The figures seem to indicate that Islamic banks are more engaged in financing economic activity compared to conventional banks and that they are highly more capitalized as well. In terms of indicators of profitability, the average ROA and ROE of Islamic banks is 2.04 and 14.20% respectively, while the corresponding figures for conventional banks are 1.85 and 14.04% respectively. In order to determine whether there are significant differences in asset composition, capitalization and profitability across Islamic and conventional banking 9 Table 2 will be shortened in a revised version of this paper to conserve space. 16

17 segments of the industry, tests of differences in means across banks, years and countries are conducted. The results are reported in Table 4. [Table 4 about here] Univariate statistics appearing in Table 4 indicate that Islamic banks allocate a significantly higher portion of their assets to loans compared to conventional banks, indicating higher exposure to credit risk. Larger portfolio risk is, in turn, balanced by lower financial risk, since Islamic banks are significantly better capitalization compared to their peers. This result holds for the bank, year and country levels t-tests conducted. In countries where Islamic finance is growing, people generally prefer to obtain financing from a Shari a compliant institution rather than from a riba-based firm, and regulators impose larger capital requirements for the establishment of an Islamic bank compared to a conventional bank. The findings of the tests of differences in profitability are not consistent across the ROA and ROE measures. While ROA of Islamic banks is significantly higher than for conventional banks, ROE differences are not pronounced. 5. Empirical Findings Table 5 lists comparative traditional measures of concentration for each of Islamic vs. conventional banks. Panel A includes the 3-bank and 5-bank concentration ratios using deposits, loans and assets, and Panel B shows the Hirschmann Herfindahl Index (HHI) using deposits, loans and assets. [Table 5 about here] In comparison with the more mature conventional banking segment of the industry, the Islamic financial industry is at its early stages of development. The first 17

18 Islamic bank was established in 1975 and the number of Shari a compliant financial institutions started to rise soon after. In this light, it is expected that a few institutions dominate the Islamic banking global market, and this is confirmed by the figures of Panel A. The HHI calculations appearing in Panel B also show that concentration in the Islamic financial global market is higher than that prevailing among their competitors. A close look at the figures shows that concentration ratios are three times higher for Islamic banks than for conventional banks, and that all HHI are six times as large. In order to gain a comprehensive understanding of competitive conditions across the two banking segments, two more widely used measures of competition, the PR H- statistic and the Lerner Index, are estimated and the results are reported in Table 6. [Table 6 about here] The calculated H-statistics indicate that monopolistic competition best describes the market structure of both the global Islamic and conventional banking segments and are in line with those reported by previous studies 10. While the figures for conventional banks slightly exceed those computed for Islamic banks, it cannot be assumed that the conventional banking market is less competitive because the H-statistic is not recognized as a continuous measure of the degree of competition (Shaffer, 2004a, b). In contrast, the Lerner index is a preferred measure because it provides a more accurate insight into the prevailing degree of market power at the bank level. The estimations indicate that Islamic banks Lerner indices are higher than their peers, suggesting a considerable higher degree of market power. 10 The results (not reported) of the model using ROA as a dependent variable following equation (2) indicate that observations are in long-run equilibrium. 18

19 In order to explain differences in profitability levels among Islamic and conventional banks, the two separate samples are combined to obtain panel data 1,173 observations. Multivariate regressions are run following equation (3) and the results are reported in Table All regressions are estimated using bank fixed effects with time dummies. [Table 7 about here] Each regression also controls for the bank s market share in terms of total assets, its size, type of activities (Islamic vs. conventional) and for the country s natural logarithm of GDP per capita. For each of the contested H-statistic and preferred Lerner index, two different specifications are run - where one assumes a linear relationship between market structure and bank profitability and the other includes a quadratic term. The reported significant coefficient estimates of the quadratic terms indeed show that the relationship between market structure and bank profitability is U-shaped and cannot be considered as linear. The inflection point of each quadratic equation is calculated and compared to the empirical distribution of the data in order to establish the sign of the relationship between bank competition and profitability across the profiled countries. In all regressions, the coefficients of the measure of competition are highly significant and can be interpreted in a consistent manner. The significant negative sign of the H-statistic parameter indicates that a lower degree of competition is associated with higher bank profitability, and the significant positive association between the Lerner index and profitability suggests that bank returns are increased with a rise in the bank s market power. The results indicate 11 The reported results in Table 7 consider ROA as dependent variables. Similar findings (not reported) obtain when ROE is the proxy for bank profitability. 19

20 that a higher degree of market power allows market players to command a pricing power which, inturn, is translated into higher rates of return. The findings thus lend support to recent strategies by major conventional banks to enter new market segments where the degree of competition is low such as the embryonic Islamic financial industry. Finally, the coefficient on the Islamic dummy variable is positive in all specifications, and it is significant when using the preferred measure of market structure or the Lerner index. This result seems to confirm that Islamic banks, on average, achieve higher records of profitability compared to their peers. Neither firm size nor market share, however, are found to be significant determinants of banks profitability. 6. Summary and Conclusions Over the past three decades, the Islamic banking industry has developed into a viable mode of finance. The number of institutions that operate along the Islamic jurisprudence has multiplied and major international players like Citigroup, HSBC and others have recognized the growth potential of this new segment of the industry and have joined the drive by establishing Shari a-compliant products, windows or subsidiaries. In this paper, we analyze the banking structures of both Islamic and conventional global markets and attempt to explain differences in profitability among the two segments of the industry. A sample of Islamic banks is selected in 13 different countries and data on conventional banks that operate in the same countries are also retrieved over the period With these data, we first calculate univariate statistics and conduct some simple tests for differences in means of performance and portfolio composition among Islamic and conventional banks. These tests provide broad evidence on the 20

21 differences in portfolio composition and bank returns of the two segments of the industry. First, Islamic banks appear to allocate a significantly greater share of their assets to financing or loans compared to conventional banks, leading to a greater exposure to credit risk. In turn, the higher portfolio risk of Islamic banks is at least partly cushioned with significantly lower financial risk with higher capitalization levels. Consistent with the basic risk and return relationship, this is translated into significantly higher returns as measured by ROA. Different proxies of market structure are estimated and they all indicate that the global Islamic banking market is both more concentrated and less competitive compared to the conventional banking segment. Next, we examine the implications of bank competition on profitability in a multivariate setting, while controlling for bank and country differences in the sample. The regression results show that a higher degree of market power is significantly positively associated with higher bank returns, possibly explaining the higher profitability levels observed for Islamic banks. The study has shown that there are important asset composition and return differences among Islamic and conventional banks operating in the same country. Competitive conditions are fundamentally different across the two segments of the industry and the higher degree of market power prevailing in Islamic banking may be driving higher profitability levels. The findings confirm that banks have a strong incentive to engage in market segments where the degree of competition is low in order to achieve high rates of return. This is particularly true for the embryonic Islamic banking market which is attracting a lot of attention from its conventional counterparts. Future research can add another dimension to the analysis by addressing issues of bank 21

22 stability. If the degree of competition is low and bank profitability is relatively high in the Islamic banking segment, it would be interesting to investigate whether Islamic banks have a role to play in contributing to overall financial stability. 22

23 References Al-Muharrami, S., K. Matthews and Y. Khabari, Market structure and competitive conditions in the Arab GCC banking system, Journal of Banking and Finance, 30, Abdul Majid, M. Z, and F. Sufian, Market Structure and Competition in Emerging Market: Evidence from Malaysian Islamic Industry, Journal of Economic Cooperation 28 (2), Bahrain Tribune, 16/06/2007. Beck, T., A. Demirguc-Kunt and R. Levine, Bank concentration, competition, and crises: first results, Journal of Banking and Finance 30, Berger, A. N., and Humphrey, D. B., 1997, Efficiency of Financial Institutions: International Survey and Direction for Future Research, European Journal of Operation Research, 98, Berger, A. N., and Mester, L. J., 2003, Explaining the Dramatic Changes in Performance of U.S. Banks: Technological Change, Deregulation, and Dynamics Changes in Competition, Journal of Financial Intermediation, 12, Berger, A.N., A. Demirguc-Kunt, R. Levine and J. Haubrich, 2004, Bank concentration and competition: An evolution in the making, Journal of Money, Credit and Banking 36, Berger, A.N., Klapper, L.F and Turk Ariss, R., 2009, Bank Competition and Financial Stability, Journal of Financial Services Research 35(2), forthcoming. Bikker, J.A. and K. Haaf, 2002, Competition, Concentration and Their Relationship: An Empirical Analysis of the Banking Industry, Journal of Banking and Finance 26, 23

24 Boyd, J. and G. De Nicolo, 2005, The Theory of Bank Risk Taking Revisited, Journal of Finance 60, Casu, B. and C. Girardone, Bank competition, concentration and efficiency in the single European market, The Manchester School 74 (4), Claessens, S. and L. Laeven, L., 2004, What drives Bank Competition? Some International Evidence, Journal of Money, Credit and Banking 36(3), Dar, H. and J. Presley, 2000, Lack of profit sharing in Islamic banking, International Journal of Islamic Financial Services, 2 (2), De Bandt, O. and E.P. Davis, 2000, Competition, Contestability and Market Structure in European Banking Sectors on the Eve of EMU, Journal of Banking and Finance 24, Financial Times, June 28, 2007, (accessed on June 28, 2007), Gelos, R. G. and J. Roldos, 2004, Consolidation and Market Structure in Emerging Market Banking Systems, Emerging Markets Review 5, Hellmann, T.F., K. Murdock and J. Stiglitz, 2000, Liberalization, moral hazard in banking and prudential regulation: are capital requirements enough?, American Economic Review 90, Jimenez, G., J. Lopez and J. Saurina, 2007, How does competition impact bank risk taking?, working paper, Banco de Espana. Koskela, E. and R. Stenbacka, 2000, Is there a tradeoff between bank competition and financial fragility?, Journal of Banking & Finance 24(12),

25 Karsen, I., 1982, Islam and Financial Intermediation. IMF Staff Papers. Keeley, M., 1990, Deposit Insurance, Risk and Market Power in Banking, American Economic Review, December, Khurshid, A., 1981, Studies In Islamic Economics, International Centre for Research in Islamic Economics, The Islamic Foundation, United Kingdom. Levine, Ross Finance and growth: theory, evidence, and mechanisms. NBER Working paper No Martinez-Miera, D. and R. Repullo, 2008, Does competition reduce the risk of bank failure? Unpublished manuscript, CEMFI. Molyneux. P., J. Thornton, and D.M.Lloyd-Williams, 1996, Competition and Market Contestability in Japanese Commercial Banking, Journal of Economics and Business 48, Molyneux, P. and H. Nguyen-Linh, 2008, Competition and risk in the South East Asian banking, Bangor Business School working paper, Bangor, Wales. Nathan, A. and E. Neave, 1989, Competition and Contestability in Canada s Financial System: Empirical Results. Canadian Journal of Economics 22, Omar, F. and Iqbal, M., Some Strategic Suggestions for Islamic Banking in the 21st century, Review of Islamic Economics, No. 9, pp Schaeck, K, M. Cihak, and S. Wolfe, 2006, Are More Competitive Banking Systems More Stable?, Unpublished Working Paper No. 143, International Monetary Fund, Washington, D.C. Schaeck, K, and M. Cihak, 2007, Banking Competition and Capital Ratios. IMF Working Paper No. 07/

26 Shaffer, S., l982, A Non-structural Test for Competition in Financial Markets. In Bank Structure and Competition. Conference Proceedings, Federal Reserve Bank of Chicago, Shaffer, S., 2004a, Comments on What Drives Bank Competition: Some International Evidence, by Stijn Claessens and Luc Laeven, Journal of Money, Credit and Banking 36, Shaffer, S., 2004b, Patterns of competition in banking. Journal of Economics and Business 56(4), Stiglitz J. E and A. Weiss, 1981, Credit Rationing in Markets with Imperfect Information., American Economic Review, Vol. 71, No. 3, pp Turk-Ariss, R., 2009, Competitive Behavior in Middle East and North Africa Banking Systems, Quarterly Review of Economics and Finance, forthcoming. 26

27 Table 1: Sample size of the Islamic and conventional banks datasets, Islamic banks Conventional banks Country Nb of banks Nb of obs Nb of banks Nb of obs Bahrain Brunei Darussalam Iran Jordan Kuwait Malaysia Pakistan Qatar Saudi Arabia Sudan Turkey United Arab Emirates Yemen Total

28 Table 2: Descriptive statistics on loans to assets (L/TA), equity to assets (E/TA), return on assets (ROA) and return on equity (ROE) by country Panel A: Islamic banks Country L/TA E/TA ROA ROE Bahrain Mean Std Dev Min Max Brunei Darussalam Mean Std Dev Min Max Iran Mean Std Dev Min Max Jordan Mean Std Dev Min Max Kuwait Mean Std Dev Min Max Malaysia Mean Std Dev Min Max Pakistan Mean Std Dev Min Max Qatar Mean Std Dev Min Max Saudi Arabia Mean Std Dev Min Max Sudan Mean Std Dev Min Max Turkey Mean Std Dev Min Max 28

29 United Arab Emirates Mean Std Dev Min Max Yemen Mean Std Dev Min Max Total Mean Std Dev Min Max Panel B: Conventional banks Country L/TA E/TA ROA ROE Bahrain Mean Std Dev Min Max Brunei Darussalam Mean Std Dev Min Max Iran Mean Std Dev Min Max Jordan Mean Std Dev Min Max Kuwait Mean Std Dev Min Max Malaysia Mean Std Dev Min Max Pakistan Mean Std Dev Min Max Qatar Mean Std Dev Min Max Saudi Arabia Mean Std Dev 29

30 Min Max Sudan Mean Std Dev Min Max Turkey Mean Std Dev Min Max United Arab Emirates Mean Std Dev Min Max Yemen Mean Std Dev Min Max Total Mean Std Dev Min Max Table 3: Summary comparative statistics by year, Islamic vs. conventional banking markets Nb. of obs. Av. TA ($ mil) L/TA E/TA ROA ROE Year Isl. Conv. Isl. Conv. Isl. Conv. Isl. Conv. Isl. Conv ,857 5, ,586 4, ,077 5, ,846 5, ,925 6, ,772 7, ,717 8, Table 4: Tests of comparisons of means, Islamic vs. conventional banking markets L/TA Islamic vs. L/TA Conventional E/TA Islamic vs. E/TA Conventional By bank By year By country By bank By year By country t value (1.6093)** (4.2767)*** (2.176)** (4.0473)*** (3.4239)*** (1.6142)* Pr (T < t) ROA Islamic vs. ROA Conventional ROE Islamic vs. ROE Conventional By bank By year By country By bank By year By country t value (1.7962)** (2.707)** ( )* Pr (T < t) Significance levels ***, ** and * correspond to 1, 5 and 10% respectively 30

31 Table 5: Comparative traditional concentration measures, Islamic vs. conventional banking markets Panel A: 3-bank and 5-bank concentration ratios using deposits, loans and assets C 3 (deposits) C 5 (deposits) C 3 (loans) C 5 (loans) C 3 (assets) C 5 (assets) Year Islamic Conv. Islamic Conv. Islamic Conv. Islamic Conv. Islamic Conv. Islamic Conv Panel B: Hirschmann Herfindahl Index (HHI) using deposits, loans and assets HHI deposits HHI loans HHI assets Year Islamic Conv. Islamic Conv. Islamic Conv Table 6: Comparative measures of competitiveness, Islamic vs. conventional banking markets Year Islamic Market H-Statistic Conventional Market Islamic Market Lerner Index Conventional Market

32 Table 7: Degree of competition and bank profitability Based upon an unbalanced panel of 250 Islamic and conventional banks in 13 banking markets where both institutions compete over the years (1,173 observations).the dependent variable measures bank profitability using ROA. Competition measures include the H-statistic and the Lerner index. A larger H- statistic implies a more competitive market, and a larger value for Lerner indicates a higher degree of market power. Both a linear and a quadratic relationship are assumed between profitability and market structure in banking. Market share is calculated using the bank s total assets. Bank Size is measured by the natural logarithm of total assets. Economic Development is controlled for using the natural logarithm of each country s GDP per capita. Regressions are run using bank fixed effects with time dummies. Robust standard errors appear in parentheses below estimated parameters. Measure of Competition H-Statistic Lerner Index Measure of Competition (0.0053)*** (0.0285)*** (0.0088)*** (0.0070)*** (Measure of Competition) (0.0158)*** (0.0058)** Market Share (0.0226) (0.0228) (0.0240) (0.0268) Bank Size (0.0020) (0.0020)* (0.0009) (0.0009) Islamic (0.0068) (0.0069) (0.0031)** (0.0030)** Economic Development (0.0015)*** (0.0015)*** (0.0008)* (0.0008) Constant (0.0141) (0.0205) (0.0063)*** (0.0059)*** Inflection point n/a n/a Sign of relationship * Significant at 10%; ** significant at 5%; *** significant at 1% 32