The Impacts of Market Structure on Profitability: An Application. to China s Banking

Transcription

1 The Impacts of Market Structure on Profitability: An Application to China s Banking Yanjun Huang China Foreign Affairs University yhuang_6@163.com Jiawen Yang the George Washington University jwyang@gwu.edu 1

2 The Impacts of Market Structure on Profitability: An Application to China s Banking Abstract In this paper, we study the structure-performance relationship by employing a unique panel data of 95 banks in China covering period. Contrary to the previous work in developed countries, our results show that market concentration leads to lower profitability and no significant relationship is observed between market power and profitability, indicating little evidence to support SCP and RMP hypotheses in China s banking. By investigating the structure-performance relations across banks with different size, we find that concentration accounts for profitability in the largest banks, and the undermining effects of market share on profitability is significant; inversely, concentration seems to decrease profitability in the rest of banking sector, rather than market share. In addition, we find that unexpectedly both of concentration and market power increase bank profitability during financial crisis. By exploring the effects of foreign minority ownership, we confirm that foreign entry tends to improve banking profitability under the tremendous banking system reform in China. And we also find that listed banks are prone to gaining less profits. Key words: market structure, banking profitability, foreign acquisition 2

3 1. INTRODUCTION Many studies have been are devoted to the relationship between market structure and profitability in regulated and concentrated industries such as banking from the perspective of industrial organization theory. Berger (1995) organizes two hypothesis for such a relationship in banking. One interpretation, the traditional structure-conduct-performance hypothesis (SCP), constructs a model of oligopolistic behavior of firms, stating that high market concentration decreases the cost of collusion and firms are able to extract monopolistic rents. Market concentration will also price the products inefficiently, which means in imperfectly competitive markets higher concentration enables firms to set prices less favorable to consumers. The other interpretation is the relative-market-power hypothesis (RMP), which indicates that only firms with larger market shares and well-differentiated products are able to be price-makers who can profitably raise the market price of a product or service over marginal cost and control a large portion of market to gain high profits by owning market power. Many works test these two hypothesis on the structure-performance relationship in banking but empirical evidence about the effects of market structure on profitability reaches no completely consensus. Some studies find a positive relationship between market structure and profitability (Jeon & Miller, 2005;Claeys& Vander-Vennet, 2009), which implies both SCP paradigm and RMP hypothesis are supported, while others take the opposite view(berger, 1995; Goldberg & Rai, 1996). These previous studies indicate that although the substantial differences are involved in operational structure and regulatory pattern between banks across the fast-changing world, there is still a possibility to conduct a meaningful analysis of the impact of market structure on profitability in banking. 3

4 China s economy has witnessed an annual growth around 10% and become one of the world s largest economy over the last decade. Under globalization environment and unprecedented reform, new policies are imposed on banking sector to open up the market for foreign investors. Research on developing nations and the financial growth indicates that high growth rates are associated with the reform of banking system and financial infrastructure. With respect to the structure-performance relationship, this paper investigates whether these two hypothesis are applied to banking in China or whether banking sector is benefitting from concentration during the reform of China s banking as the same as banking in advanced economies. Unlike the structure-performance relationship in developed countries, our results show SCP and RMP hypotheses are lack of support in China s banking. We are also interested in the structure-performance relations in banks with different size and we find that the relationship between market structure and profitability is heterogenous among banks with different size. Another interesting result is that concentration accounts for bank profitability during financial crisis. We also explore other determinants of the relations(i.e. foreign capital), and confirm that foreign investment is very crucial to banking profitability, moreover unlisted banks are unexpected to outperform listed banks, especially for sample excluding these largest banks. Our paper contributes to the existing literature in a number of dimensions. First, studies measuring the structure-profitability relations in developing economies are limited. This paper fills the gap by testing these two hypotheses for China, as a representative case of emerging countries. Second, reflecting the tremendous changes of banking reform and innovation environment in the last decade, we develop a model of the relationship between market structure and profitability for China with the widening scope of explanatory variables and the latest data. An understanding of the structure-performance relationship could give implication to policy-making related to the current reform 4

5 and the subsequent stage. Third, we investigate the structure-performance relationship by dividing banks by size and examine financial crisis and non-crisis separately. In this paper, Section 2 describes the background of China s banking system. Section 3 organizes the literature review and develops the hypotheses. Section 4 defines the variables and data. Section 5 conducts the methodology. Empirical results and robustness check are showed in Section 6 and Section 7 separately. Section 8 concludes. 2. CHINA S BANKING SYSTEM Prior to 1978, mono-bank model is operated in China s banking system where People s Bank of China (PBC) was established not only to issue currency as a central bank but also have commercial service as a financial hub. Since the first economic reform, a two-tiered banking system was substituted for monobanking model to improve economic efficiency and resource allocation. Big Four state-owned banks, including Bank of China (BOC), China Construction Bank (CCB), Agricultural Bank of China (ABC) and Industrial and Commercial Bank of China (ICBC), were established, whereas they were initially restricted to serving assigned sectors separately. Until 1985, Big Four was permitted to widen the scope of raising funds and allocating capital so that they were able to compete with each other in all the sectors. They expanded to universal banks with kinds of financial affiliates in Nonetheless, they were lack of incentives to compete due to national policy. Then bank reform was incrementally focused on ownership. New small and medium sized jointstock commercial banks which were allowed to offer services to households and firms were established by government, such as Bank of Communications, CITIC Industrial bank and China Merchants Bank. In 1981 Nanyang Commercial Bank set up branches in Shenzhen, as the first foreign bank entry since the reform and opening up. 5

6 From 1993, a second stage of banking reform was launched for alleviating the deteriorating asset quality of state-owned banks. Numerous measures were implemented as follows: Big Four was recapitalized through the special bonds issued by the Ministry of Finance in 1998 and NPL was transferred to newly established asset management companies which are also SOEs. Under the Central Bank Law, PBC, to reduce the influence of government on credit allocation decisions, was confirmed to be responsible for the implementation of monetary policy. Urban and rural credit unions started to merge then formed city commercial banks, though they are still controlled by the state and SOEs. This means one of the reform subjects was to create a competitive commercial banking environment. Government turned to encourage foreign bank entry by offering the tax-exempt incentives available to all the foreign companies. Since Asian Development Bank acquired a 1.9% equity in China Everbright Bank in 1996, foreign investor were further allowed to purchase stakes in domestic banks but limited proportion with restricted regulations. After WTO entry in 2001, China makes further efforts to open up the banking system with a new set of rules. Interest rates are more liberalized, and ownership takeovers and M&A are less restricted. Foreign banks were allowed incrementally to conduct RMB business with enterprises in certain regions from China Banking Regulatory Commission (CBRC) was created in 2003 to monitor the banking industry. CBRC launched a set of new rules to encourage foreign acquisitions with a limited share up to 25%. China s financial market is attractive to strategic foreign investors. Many commercial banks transferred a proportion of equity to foreign investors. In addition, foreign investors participated in establishing of Chinese bank by offering foreign minority stake. Partial privatization has been overwhelming across the country though commercial banks are still state-owned. Management improvement in banking has been put on the agenda of reform. 6

7 In 2005, China established a managed floating exchange rate regime based on market demand and supply with reference to a basket of currencies. Then commercial banks confronted with a more volatile environment full of more new requirements, new markets and new financial tools, are accelerating the transformation of management theories. To deepen the reform of shareholding systems and simultaneously increase state-owned banks operational efficiency, three of the Big Four (ICBC, CCB and BOC) were listed on stock exchanges in Hong Kong and Shanghai from 2005 to After preparation and trimming down the NPLs, the ABC, was finally listed in Hong Kong and Shanghai in mid-2010 to allow more market discipline to monitor its performance, which is a landmark achievement for financial system reform in China. In 2012, Wenzhou Financial Comprehensive Reform Area was approved to establish for encouraging private financing. In 2013, policymakers gradually lowered the lower limit of lending rates and finally ended all restrictions on lending rates. Interest rate liberalization will facilitate the private sector an access to credit, which will benefit China s economy in a long term run. However, the government also realizes, as a key component of economic reform policy, a market-based interest rate environment will likely put enormous pressure on reduction of financial market volatility. Despite of the need to maintain the financial stability, interest rate liberalization is expected to improve the allocation of capital resources and make a historical progress for China s economy development. 3.THEORY AND HYPOTHESES The bulk of previous literature related to banking concentrates on exploring the statistical relationship between profitability and market structure. Researches fall into two distinct industrial organization theories: SCP paradigm and RMP hypothesis. For the former hypothesis, the market structure imposes an influence on the conduct of firms (i.e. price-setting, trade policy) and finally causes the changes of individual performance (i.e. profitability, efficiency). According to SCP hypothesis, high level of 7

8 concentration lowers the cost of collusion, which fosters collusions among firms and gives rise to monopoly rents. Then it affects the way firms behave, enabling them to gain supernormal profits. So concentrated market structure causes higher profitability for firms. As for the highly concentrated banking industry, a bank can set higher interest spreads by offering lower deposit rates and charging higher loan rates due to competitive imperfections, therefore get high profits. With respect to SCP paradigm which focuses on the top largest firms in the industry, RMP hypothesis pays more attention to market share of individual banks. Based on the latter approach, market share is able to capture market power arising from product differentiation, which means only firms with a large market share have higher quality products and thus set higher prices and increase profitability. Shepherd (1982) asserts that banks with a higher market share exert more market power and earn higher profits, independent of the level of market concentration. The empirical evidence in the existing literature on the structure-profitability relationship is mixed. Although SCP studies utilizing manufacturing data find a positive relationship between concentration and profitability, Smirlock (1985) doubts that there exists similar correlation in banking industry. Using a panel data from 2,700 unit state banks over a period , Smirlock finds that there is no evidence to support SCP hypothesis once market share is controlled. In his paper, market share, rather than concentration, has a significant and positive impact on bank profitability. As reviewed previous studies using bank call report data, Gilbert (1984) finds only one of the five price measures produces a coefficient on concentration which is statistically significant. After examining the determinants of international bank profitability and reviewing the relevance of expense preference behavior theories, Bourke (1989) finds that concentration is positively related to profitability in banking of Europe, North America and Australia. In 1990s, some researches focus on the relationship between market structure and profitability incorporating efficiency-structure hypothesis including X-efficiency and scale-efficiency. Berger (1995) creates and implements tests that containing direct measures of both market structure and efficiency in the 8

9 same model to distinguish the effects on profitability in banking of RMP and Efficient Structure (ES) hypotheses. He finds limited support for all the four hypotheses by employing 1980s data of US banks. Berger and Hannan (1997) use US bank data and find more support for SCP hypothesis than the rest of all the hypotheses. And they also prove that the quiet life hypothesis where firms with larger market power are less efficient as a result of slack management, is supported. After US banking has been studied extensively, Goldberg and Rai (1996) find there is no significantly positive relationship between concentration and profitability for large banks across 11 European countries from 1988 to Their finding suggests that RMP hypothesis plays more significant role than SCP hypothesis, and only in banks located in low concentration countries, ES hypothesis is supported. This is opposite to the findings of the study by Goddard et al. (2004), who assert that evidence in their paper favoring the SCP hypothesis of a positive association between concentration and profitability, but little apparent relationship between X-inefficiency and profitability in bank level. The research papers that follow in this special issue push significantly further in expanding research on the structure-performance relationship in developing countries. Claeys and Vander Vennet (2009) review banking in Central and Eastern European nations and suggest that SCP hypothesis is applicable in these countries by utilizing SFA technique. In Jeon and Miller (2005), the findings indicate that there is strong evidence supporting for a positive relationship between concentration and profitability through a state-by-state basis. Furthermore, by temporal causality tests, they conclude that SCP paradigm is an explanation for bank profitability, therefore market power hypothesis is supported. This is the same as the findings of the research by Tregenna (2009). In this study, Tregenna finds that the concentration-profitability relationship in banking sectors is positive and robust, but the efficiency-profitability relationship is weaker. 9

10 Nowadays, new papers generally have an international orientation that includes different groups, such as developed and developing nations, a significant departure from the vast majority of the studies in the extant literature. But papers related to emerging groups are still limited. Mirzaei et al. (2013) examine the effects of market structure on profitability in banking from emerging economies and advanced economies separately. They find evidence that RMP hypothesis is supported in advanced economies. However, there is little evidence to prove that SCP hypothesis is applicable to both of groups. In addition, the impact of concentration on profitability is unexpectedly negative in emerging economies. For Chinese case, Fu and Heffernan (2009) investigate the relationship between market structure and profitability in China s banking sectors by using two subsample panel data to observe the impacts in banking before and after the reform in China banking. Through the test, they find that prior to the reform, RMP hypothesis is applicable to China s banking. In contrast with RMP hypothesis, concentration affects profitability negatively after financial liberalization started in 1992, which indicates that SCP is not supported with China s banking system. Garcia-Herrero et al. (2009) analyze the determinants of low profitability of China s banking sectors for the period from 1997to 2004 and find that profitability is affected negatively by market share and there is no significant relationship between concentration and profitability. We investigates empirically profitability of China s banking in SCP framework. Our first series of hypotheses are: Hypothesis 1A. Bank concentration improves bank profitability in China. China. Hypothesis 1B. There is no positive relationship between bank concentration and profitability in 10

11 Based on the SCP hypothesis, concentration results in a lower collusion cost so firms are able to extract monopolistic rents, which means bank concentration is expected to lead to higher profitability in China. Alternatively, focusing on RMP hypothesis, second series of hypotheses are: Hypothesis 2A. Banks with larger share tend to be more profitable in China. banking. Hypothesis 2B. There is no positive relationship between market share and profitability in China s In RMP hypothesis, firms with large market shares and well-differentiated products can exert their market power in pricing products, therefore they are able to control the market and earn high profits. In line with RMP hypothesis, market share is expected to have a positive impact on profitability in China s banking. 4. METHODOLOGY below: Following Smirlock(1985), the traditional hypothesis can be tested by the profit equation shown Π it =f (Market Structure it ) (1) Where Π it is the dependent variable measuring the profitability of ith bank at time t; Market Structure it refers to either market share (MS) of ith bank at the firm level or the concentration ratio (CR) at the industry level. The expanded version is as follows: Π it =β 0 +β 1 MS it +β 2 CR t +µ it (2) 11

12 Where MS measures market share of bank ith; CR is the index which measures the top largest firm concentration ratio. This formula is assumed to reflect the two hypothesis of SCP and RMP. We are following the reasonable expansion of model constructed by Mirzaei et al. (2013), employing different variables which is much more applicable to China s case. Π it =β 0 + β 1 MS it +β 2 CR t + B it + M t +µ i +u t +v it (3) B it is the first vector of control variables which account for bank-specific characteristics; M t is the second vector of control variables which denote financial structure and macroeconomics characteristics. An error term for each bank i at the time t includes three components: µ i is standing for the unobserved individual specific effects; u t is the unobserved time effects; v it is the normal stochastic disturbance. 5. DATA AND VARIABLES 5.1 Data We employ an unbalanced panel data obtained from Bankscope and annual reports from banks. The macroeconomic level data are retrieved from the World Bank. The full sample runs from 2003 to 2013with 95 observations in China. 5.2Variables All the variables are presented by Table Dependent variables In the empirical analysis performance measures of banking are focused on as dependent variables. In the previous banking literature the profitability variables of banking are represented by two alternative measures: the ratio of profits to assets, i.e. the return on average assets (ROAA) and the profits to equity 12

13 ratio, i.e. the return on average equity (ROAE), which refers to the ability of banks asset and equity generating revenues (Lloyd-Williams et al., 1994; Fu & Heffernan, 2009) The focal variables The focal variables focus on market structure of banking sectors in China. To be taken to investigate RMP hypothesis, market share, which represents market power at a firm level, is calculated as the ratio of the individual bank asset to the total asset of the whole industry in China. SCP hypothesis employs the concentration ratio which estimates the extent to that the top largest banks take control of the entire industry. According to Demirguc-Kunt et al. (2003), we measure bank concentration as the fraction of bank assets held by the five largest banks in China. We use the 5-firm concentration ratio due to Chinese situation in which assets held by top five largest banks are much higher than assets held by the rest of the banking sector Control variables Control variables are composed of bank-specific variables, financial structure and macroeconomics variables, which are instrumental in explaining bank profitability are introduced into the regressions. First, ratio of equity to total assets is considered as a measure of bank capital strength. In perfect competitive market assumption with symmetric information, capital refers to the amount of own funds available making contribution to earnings and supporting banks business (Athanasoglou et al., 2008). In imperfect competitive market assumption, banks are expected to have better performance to transmit asymmetric information through the higher capital ratio. In addition, a bank can have more possibility to earn profits from holding capital in excess of the regulatory minimum and reducing potential risk by holding an adequate capital (Mirzaei et al., 2013). This regressor is expected to be positively correlated with profitability (Pasiouras & Kosmidou, 2007). 13

14 Second, the ratio of overheads to total assets is used to proxy bank cost which measures X- efficiency(tregnna, 2009). Overheads to total assets ratio providing information on the fluctuation of bank cost compares a bank s cost efficiency with other banks indirectly. Another popular indicator that has the same function is cost to income ratio, but in China the lending and deposit rates have not been fully liberalized over yet. We use the former indicator to represent bank cost, which is expected to be negatively correlated with bank profitability (Garcia-Herrero et al., 2009). Third, considered to be relative to bank profitability, bank size is measured by the log of total assets. Garcia-Herrero and Vazquez (2007) shows that banks with large size are able to maintain lower cost owing to scale efficiency. However, Stiroh and Rumble (2006) hold that big banks are too large to manage, which leads to less profits. Therefore, the effects of bank size are mixed. Fourth, diversification of bank income is constructed as an indispensable variable in this model. Following the literature (Stiroh, 2004; Casu & Girardone, 2005), non-interest income produced by nontraditional activities is a share of banking profitability and also increase the volatility of bank revenue (DeYoung & Roland, 2001). Fifth, based on the paper (Smirlock, 1985), market growth is likely to influence bank profits. The rate of growth in lending loan growth is employed as an indicator to represent market growth. Sixth, the effects of foreign minority ownership on profits are examined with dynamics dummies. Unlike other countries, except foreign-owned banks, China s banks are owned by either local government or state-owned enterprises. Foreign acquisitions were restricted with government policy in China, including offering a limited foreign minority stake. According to previous literature, banks with foreign investors tend to higher profits than domestic banks in developing countries(berger et al., 2009). The dummy variable is equal to 1 if foreign investors acquire a share of the particular bank in the given year. 14

15 Moreover, we include the dynamic dummy variable listed that takes the value of 1 when a bank is listed in the stock exchange in the given year, which assesses the role of increasing privatization in China s banking sector. As with macroeconomic environment in which banks operate may influence bank profits, the previous studies show that economic growth (Demirguc-Kunt& Huizinga, 1999; Bourke, 1989) and inflation (Boyd et al. 2001; Demirguc-Kunt et al., 2003) increase banking profitability. Therefore we specify GDP growth and inflation rate as macroeconomics indicators in the model. 6. EMPIRICAL RESULTS As discussed above, we examines empirically the impact of market structure on bank profitability with annual panel data for a maximum of 168 banks over the period Testing the structure-performance relationship in China s banking Table 2 reports the empirical results for the baseline specification. Both market share of individual banks and industry concentration have been correlated negatively with ROAA. With respect to ROAA, the result shows that the coefficient of 5-firm concentration is negative and statistically significant at the 5% level, suggesting that an increase in market concentration leads to a decline of returns on assets of a bank, but market share is not significantly correlated, although the sign of market share is negative. With regard to ROAE, the result indicates that both measures of market structure have positive coefficients which has the opposite trend from the measure of ROAA, but not significantly. These results run strongly contrary to the existing literature supporting SCP and RMP paradigm in developed countries and consistent with some of the views related to developing countries (Mirzaei et al. 2013). Among the control variables, capital strength is not significant and unexpected to be negative with both ROAA and ROAE. 15

16 The sign of bank size varies with different measures of profitability. But each is not significant, which may be explained by the reason that in China s banks with larger size are involved with massive and complicated government intervention. The coefficients on bank cost and income diversification are not significantly negative, which means operational efficiency and income diversification in business seems a key factor enabling banks to gain high profits. No significant relationship between market growth and profitability has been observed. The weak results for the effects of loan growth on profitability might be interpreted as that banking loan growth is unable to create a market environment in China s banking to promote the high returns. Foreign capital is highly correlated with ROAA, indicating that corporate governance with foreign acquisition fosters profitability in China s banking from the year that foreign capital enters the particular bank. This may be based on three reasons: first, foreign minority ownership has spillover effects, including introduction of new banking technology and management which leads to higher profitability; secondly, it is relatively easy to make financial innovations to gain higher revenue under foreign minority ownership with new products and services in the local market; thirdly, foreign acquisition profit-seeking conduct will improve the competitive environment (Bonin et al., 1998). There is negative relationship between listed and profitability, especially significantly with ROAA, which suggests listed banks are less profitable than unlisted banks in China. Listed banks tend to have a lower NPL ratio, therefore they are prone to less risky behavior. As capital strength has a negative impact on profitability, unlisted banks perform better than listed banks. For macroeconomic environment variables, inflation is positively associated with bank profits, whereas GDP growth is not significantly correlated. Inflation may influence banks through net interest margin but not economically huge (Demirgüç-Kunt et al., 2003). 16

17 6.2 Testing the structure-performance relationship on different types of banks One objective of the estimates is investigating that the difference of the concentration-performance relationship between the top largest banks and the rest with smaller size of the banking sector in China (Table 3). And we add an interactive term for market share ratio with foreign capital dynamic dummy to investigate whether the effect is heterogeneous across banks with different size. The sign of concentration coefficient for Big Five varies when the dependent variables alternate. For ROAA, as one of profitability measures, the coefficient of market share is negative in the Big Five, independent of bank concentration, indicating that for Big Five, the larger proportion in market, the lower profitability. It may reflect the fact that the top largest banks which are state-owned earned lower returns due to government intervention, slack management and moral hazard. And market concentration has a significantly positive coefficient of ROAE. Despite of government policy, Big Five acts as price-makers for the rest of banking sector, therefore the collusive behavior facilitates an opportunity with higher interest rate spreads and hence higher profits. The sign of diversification is expected to become positive, in the meanwhile loan growth is found to be significantly negative with bank profits. It is likely that noninterest income is gradually replacing traditional interest income in Big Five. Turning to the rest of banks, compared with Big Five, the coefficient of market share is not significant, whereas concentration is statistically significantly negative with ROAA. The results show concentration affects ROAA in banks with smaller size more significantly. On one hand, larger banks extract higher profits through bank concentration, on the contrary, other banks cannot benefit from this. On the other hand, government interference in lending policies has an impact on bank profitability so large banks or local banks may facilitate firms access to credit and the stability of the banking system. Barriers to entry in the banking sector are high, which may result in that banks having little access to government support earn less profits (Ruthenberg, 1994). There is also a difference lying in foreign entry that for the rest of banks, foreign investment will facilitate smaller banks an increase in profitability. Instead, for Big Five, 17

18 the structure-performance relationship in a bank with foreign capital entry shows an inverse trend. This result may be based on the following reasons: Firstly, the effects on banks earning ability by foreign minority ownership should be recognized; Secondly, foreign investors are not only seeking profits in China s banking but also attempting to step into the financial market in China through shares of bank equity by different entry modes, so foreign investors are targeting at banks with larger size to get more support from the government; Thirdly, although foreign acquisition seems to undermine profitability in Big Five, bigger size in the five banks with foreign entry has more profits; At last, the obvious difference between two types of banks may be a result of that Big Five is more involved with government intervention and restricted with financial policies, so that foreign minority ownership is unable to play an important role in boosting profits. 6.3 Testing the structure-performance relationship during the financial crisis Since the financial crisis during the period over , world economy experienced a recession. The global crisis was brought on by the disruptions of systemically important financial institutions. Banking and financial sector across various countries collapsed and many banks failed during the crisis. Turning to the financial crisis period, Table 4 shows that a positive sign on the concentration coefficient, when both of ROAA and ROAE are employed as the dependent variable, suggesting that banks with smaller size can also benefit from the high bank concentration during the crisis through redistribution of profits or trickle-down effects within the whole banking sector. Market share is also significantly positive with bank profitability, indicating that market power is able to facilitate banks an access to gain high profits. Another important finding is that the coefficients on listed and foreign capital are inverse of previous regressions. The significant relationship between listed and ROAA indicates that listed banks outperform unlisted banks. As seen from Table 4, foreign investors are capable of making profits while banking is confronted with non-financial crisis. But unexpectedly banks with foreign minority ownership earn much less profits than banks without foreign minority ownership. 18

19 This may arise from the fact that these foreign companies were taking a heavy toll abroad during crisis period. 7. ROBUSTNESS CHECK The empirical results are verified by applying an alternative measure of dependent variables and focal variables. Herfindahl-Hirschman index, the summation of each bank s square of market share, is substituted for 5-firm concentration ratio, and share of a bank s loans to total market loans is employed instead of assets share. The estimates of this model specification are consistent with previous construction. By using net interest margin and non-interest returns (NIR)(Goldberg& Rai, 1996), instead of ROAA and ROAE, the alternative approach cannot have distinct results. 8. CONCLUSION The objective of this study is to examine the impact of market structure on profitability in banking sector incorporating two industrial organization theories. We find bank concentration is negatively associated with ROAA in bank and no significant relationship between market power and banking profitability, which provides little evidence to support SCP and RMP hypotheses in banking of China. Different form the result based on developed countries, the structure-performance relationship is negative in China, which is consistent with the viewpoint of developing countries (Mirzaei et al., 2013). But for banks with different size, we find that market power has more significantly negative impacts on profitability in top largest banks, independent of concentration, whereas there is a more significantly positive relationship between concentration and profitability. For the 19

20 rest of banking, both concentration and market power have no significant relationship with banking profits. The signs of other crucial determinants(i.e. foreign capital, listed), also varies between these two types of banks. However, concentration and market power do account for bank profitability when the banking system is confronted with financial crisis. The coefficients on market share remains negative almost all the time. The results have important implication: firstly, SCP hypothesis does not dominate in the banking system in China, therefore deconcentration is conducive to improving bank profitability. Policy makers should focus on antitrust regulation and facilitate banks with a competitive environment; secondly, since a decrease in profitability when market share increases, profits are not derived from market power in China; thirdly, for banking excluding largest banks, deconcentration is crucial for profits, but for five largest banks, eliminating the diseconomy of scale is necessary; fourthly, during the crisis, an increase in bank concentration ratio may contribute to soundness of banking in China; Fifthly, since foreign investment entering into China s banking improves profitability to some extent, implementing measures to open up the financial market, boost competition and establish various types of ownership should be encouraged, i.e. desterilization of foreign minority ownership, income diversification promotion, entry barrier removal. 20

21 Reference Athanasoglou, P.P., Brissimis, S.N., Delis, M.D., Bank-specific industry-specific and macroeconomics determinants of bank profitability. Journal of International Financial Markets, Institutions and Money, 18 (2): Beck, T., Demirguc-Kunt, A., Levine, R., Bank concentration, competition, and crises: first results. Journal of Banking and Finance, 30: Berger, A.N., The profit-structure relationship in banking: tests of market power and efficientstructure hypotheses. Journal of Money, Credit and Banking,27 (2): Berger, A.N., Demirguc_-Kunt, A., Levine, R., Haubrich, J.G., Bank concentration and competition: an evolution in the making. Journal of Money, Credit, and Banking, 36 (3): (June, Part 2). Berger, A.N., Hannan, T.H., Using efficiency measures to distinguish among alternative explanations of the structure-performance relationship in banking. Managerial Finance, 23 (1): Berger A., Hasan I., Zhou M., Bank ownership and efficiency in China: What will happen in the world s largest nation? Journal of Banking & Finance, 33: Bonin, J.P., Hasan, I., Wachtel, P., Bank performance, efficiency and ownershipin transition countries. Journal of Banking and Finance, 29: Bonin, J.P Banking in Transition Economies: Developing Market Oriented Banking Sectors in Eastern Europe. Brookfield, Vermont: Edward Elgar Publishing. Bourke, P., Concentration and other determinants of bank profitability in Europe, North America, and Australia. Journal of Banking and Finance, 13 (1):

22 Boyd, J.H., Levine, R., Smith, B.D., The impact of inflation on financial sector performance. Journal of Monetary Economics, 47: Casu, B., Girardone, C An analysis of the relevance of off-balance sheet items in explaining productivity change in European banking. Applied Financial Economics, 15: Claessens, S., Demirguc-Kunt, A., Huizinga, H., How does foreign entry affect domestic banking markets? Journal of Banking and Finance, 25 (5): Claeys, S., Vander-Vennet, R., Determinants of bank interest margins in Central and Eastern Europe: a comparison with the West. Economic Systems, 32(2): Demirguc-Kunt, A., Huizinga, H., 1999.Determinants of commercial bank interest margins and profitability: Some international evidence. World Bank Economic Review, 13 (2): Demirguc-Kunt, A., Laeven, L., Levine, R., The impact of bank regulations, concentration, and institutions on bank margins. The World Bank. Policy Research Working Paper Series DeYoung, R., Roland, K., Product mix and earnings volatility at commercial banks: Evidence from a degree of total leverage model. Journal of Financial Intermediation, 10: Fu, X., Heffernan, S., The effects of reform on China s bank structure and performance. Journal of Banking and Finance, 33: Garcia-Herrero, A., Gavila S., Santa barbara D.,2009. What explains the low profitability of Chinese banks? Journal of Banking & Finance, 33: Garcia-Herrero, A., Vazquez, F., International diversification gains and home bias in banking. IMF Working Paper No. WP/07/

23 Gilbert, A., Studies of bank market structure and competition: a review and evaluation. Journal of Money, Credit, and Banking, 16: Goddard, J., Molyneux, P., Wilson, J.O.S., The profitability of European banks: across-sectional and dynamic panel analysis. Manchester School, 72 (3): Goldberg, L.G., Rai, A., The structure-performance relationship for European banking. Journal of Banking and Finance, 20 (4): Jeon, Y., Miller. M Performance of Domestic and Foreign Banks: The Case of Korea and the Asian Financial Crisis. Global Economic Review,34(2): La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Government ownership of banks. Journal of Finance, 57: Lin, X., Zhang, Y., Bank ownership reform and bank performance in China. Journal of Banking and Finance, 33: Lloyd-Williams, D.M., Ph. Molyneux and J. Thornton, 1994, Market structure and performance in Spanish banking, Journal of Banking and Finance, 18: Mirzaei, A., Moore T., Liu, G., 2013,Does market structure matter on banks profitability and stability? Emerging vs. advanced economies, Journal of Banking and Finance, 37: Molyneux, P., Thornton, J., Determinants of European bank profitability: a note. Journal of Banking and Finance, 16 (6): Pasiouras, F., Kosmidou, K., Factors influencing the profitability of domestic and foreign commercial banks in the European Union. Research in International Business and Finance, 21 (2):

24 Ruthenberg, D Structure-performance and economies of scale in banking in a unified Europe, Bank of Israel Banking Review, 4: Shepherd W., 1982, Economies of scale and monopoly profits, in: J.V. Craven (Ed.): Industrial Organization, Antitrust, and Public Policy. Boston: Kluwer Nihoff Smirlock, M., Evidence of the non-relationship between concentration and profitability in banking. Journal of Money, Credit and Banking, 17 (1): Stiroh, K., Diversification in banking: Is non-interest income the answer? Journal of Money, Credit and Banking, 36 (5): Stiroh, K.J., Rumble, A., The dark side of diversification: The case of US financial holding companies. Journal of Banking and Finance, 30: Tregenna, F., The fat years: the structure and profitability of the US banking sector in the pre-crisis period. Cambridge Journal of Economics, 33 (4): Vander Vennet, R., 1993, Concentration, efficiency and entry barriers as determinants of EC bank profitability, Journal of International Financial Markets, Institutions, and Money, 4 (3-4):

25 Table 1 Variable definition Variables Definition Unit Source Profitability measures ROAA Net profits before tax divided by average total Ratio Bankscope assets ROAE Net profits before tax divided by average Ratio Bankscope equity Market structure variables Market share Share of a bank s assets to total assets in the Ratio Bankscope market 5-firm concentration ratio Share of five largest bank assets to total assets Ratio Bankscope in the market Bank-specific variables Capital strength Equity divided by total assets Ratio Bankscope Bank cost Overhead costs divided by total assets Ratio Bankscope Income diversification 1- [(Net interest income - Other operating Ratio Bankscope income) /Total operating income] Market growth Log loans t -log loans t-1 Ratio Bankscope Foreign capital Defined as 1 if foreign minority ownership Dummy Bankscope exists in bank ith in year t Listed Defined as 1 if bank ith is listed in year t Dummy Bankscope Macroeconomics indicators Inflation Based on consumer prices Ratio Worldbank GDP growth Inflation-adjusted growth rate of GDP Ratio Worldbank 25

26 Table 2 Results for baseline specification Dependent variables Independent variables ROAA ROAE 5-firm concentration ** (0.926) (40.420) Market share (6.544) ( ) Capital strength (0.026) (1.076) Bank cost , (29.151) (1, ) Bank size (0.067) (1.984) Income diversification (0.323) (7.094) Market growth (0.013) (0.725) Foreign Capital 0.337** (0.133) (5.529) Listed ** (0.122) (11.278) Inflation 0.049*** 1.274** (0.016) (0.564) GDP growth * (0.021) (0.890) Constant (1.406) (44.033) Fixed Effect No. of observations R-squared No. of banks Bank individual fixed effects are included in the regression. Standard errors are clustered at bank level. Robust standard errors in parentheses. ***, **, * indicate statistical significance at the 1%, 5%, and 10% (two-tail) test levels, respectively. 26

27 Table 3 Results for the effects of market structure on Big Five and the banks excluding Big Five Big Five Non-Big Five Dependent variables Independent variables ROAA ROAE ROAA ROAE ROAA ROAE ROAA ROAE 5-firm concentration *** *** * (1.762) ( ) (1.389) ( ) (1.077) (39.644) (1.061) (40.520) Market share ** * (3.065) ( ) (4.756) ( ) (24.545) ( ) (23.852) ( ) Capital strength (0.053) (8.790) (0.035) (4.479) (0.026) (1.087) (0.025) (1.085) Bank cost , , (38.076) (7, ) (41.789) (8, ) (29.810) ( ) (30.230) ( ) Bank size (0.346) (40.805) (0.237) (38.486) (0.081) (2.061) (0.081) (1.994) Income diversification (0.774) (80.794) (0.763) (69.238) (0.322) (7.064) (0.327) (7.209) Market growth ** *** (0.017) (1.347) (0.023) (0.625) (0.014) (0.726) (0.014) (0.729) Foreign Capital ** 0.327** (0.135) (0.393) (29.157) (0.135) (5.323) (0.155) (5.613) Market share* Foreign Capital *** ** 1, ** (2.217) ( ) (25.170) ( ) Listed ** *** (7.511) (0.210) (6.351) (0.427) (8.055) Inflation *** 1.312** 0.054*** 1.346** (0.017) (0.699) (0.018) (0.704) (0.018) (0.626) (0.018) (0.636) GDP growth * * * * (0.025) (2.151) (0.023) (1.727) (0.023) (0.915) (0.023) (0.932) Constant , (7.716) ( ) (5.052) ( ) (1.692) (44.470) (1.697) (43.331) Fixed Effect No. of observations R-squared No. of banks Bank individual fixed effects are included in the regression. Standard errors are clustered at bank level. Robust standard errors in parentheses. ***, **, * indicate statistical significance at the 1%, 5%, and 10% (two-tail) test levels, respectively. 27

28 Table 4 Results for different periods Financial crisis period Non-crisis period Dependent Variables Independent variables ROAA ROAE ROAA ROAE 5-firm concentration ** *** (8.703) ( ) (1.092) (25.461) Market share 8.242*** (1.963) ( ) (8.005) ( ) Capital strength *** ** (0.013) (1.785) (0.030) (0.220) Bank cost , (19.083) (1, ) (36.481) ( ) Bank size *** (0.049) (10.303) (0.093) (1.747) Income diversification (0.202) (7.435) (0.328) (2.007) Market growth (0.048) (4.024) (0.014) (0.281) Foreign Capital ** 6.607* (0.139) (7.446) (0.127) (3.867) Listed 0.595** (0.223) (22.636) (0.213) (14.594) Constant (4.582) ( ) (1.987) (41.330) Fixed effects No. of observations R-squared No. of banks Bank individual fixed effects are included in the regression. Standard errors are clustered at bank level. Robust standard errors in parentheses. ***, **, * indicate statistical significance at the 1%, 5%, and 10% (two-tail) test levels, respectively. 28

29 Table 5 Results for robustness checks Dependent Variables Independent variables NIR NIR Net interest margin Net interest margin HHI * *** *** *** (2.921) (2.751) (4.772) (5.140) Market share (asset) (6.233) (10.162) Market share (loan) * ** (9.524) (10.017) Capital strength ** (0.026) (0.019) (0.023) (0.021) Bank cost (28.452) (27.633) (24.753) (24.406) Bank size (0.062) (0.065) (0.078) (0.086) Income diversification *** *** (0.337) (0.317) (0.242) (0.252) Market growth (0.013) (0.013) (0.020) (0.018) Foreign Capital 0.348** 0.401*** (0.133) (0.112) (0.223) (0.218) Listed ** *** * (0.123) (0.096) (0.242) (0.237) Inflation 0.050*** 0.052*** 0.086*** 0.087*** (0.016) (0.016) (0.013) (0.013) GDP growth * * (0.021) (0.019) (0.025) (0.024) Constant ** 5.529*** (1.252) (1.217) (1.655) (1.780) Fixed effects No. of observations R-squared No. of banks Note: Market share (loan) is total loans in each bank as a percentage of the market. The formulation of HHI is HHI= (Market Share) 2. The formulation of NIR is NIR= (1 + ROAA)/(1 + net interest margin/total assets).bank individual fixed effects are included in the regression. Standard errors are clustered at bank level. Robust standard errors in parentheses. ***, **, * indicate statistical significance at the 1%, 5%, and 10% (two-tail) test levels, respectively. 29