Hiệu quả ngân hàng trong hoạt động kinh doanh vốn. Hiệu quả hoạt động ngân hàng Hiệu quả tín dụng ngân hàng Hiệu quả đầu tư Hiệu quả nhân sự Hiệu quả tổ chức bộ máy, chi nhánh Hiệu quả vận dụng các chính sách vĩ mô Hiệu quả quản lý rủi ro
Margin—The Journal of Applied Economic Research : (2010): 427–461 SAGE Publications Los Angeles/London/New Delhi/Singapore/Washington DC DOI: 10.1177/097380101000400403 Bank-specific, Industry-specific and Macroeconomic Determinants of Bank Efficiency: Empirical Evidence from the Thai Banking Sector Fadzlan Sufian Muzafar Shah Habibullah This paper examines the efficiency of the Thai banking sector from 1999 to 2008 by using the data envelopment analysis (DEA) method The results indicate that inefficiency in the sector stems mainly from scale rather than pure technical efficiencies The findings suggest that small banks are most efficient, while medium-sized banks have been the least efficient banking group Domestic banks have been relatively more efficient than their foreign bank peers, which can be attributed largely to a higher pure technical efficiency (PTE) level The results from the multivariate regression analysis suggest that banks with higher loans intensity and which are relatively better capitalised tend to exhibit higher efficiency levels On the other hand, credit risk is negatively related to bank efficiency The empirical findings suggest that the recent global financial crisis exerts a negative impact on the efficiency of Thai banks Keywords: Banks, Efficiency, Data Envelopment Analysis, Panel Regression Analysis, Thailand JEL Classification: G21 Sufian is at the Department of Economics, Faculty of Economics and Management, Universiti Putra Malaysia; and at Khazanah Research and Investment Strategy, Khazanah Nasional Berhad, Level 35, Tower 2, Petronas Twin Towers, Kuala Lumpur City Centre, 50088 Kuala Lumpur, Malaysia (mailing address); e-mail: fadzlan.sufian@khazanah.com.my; fsufian@gmail.com, Tel: 603-2034-0197, Fax: 603-2034-0035; Habibullah is at the Department of Economics, Faculty of Economics and Management, Universiti Putra Malaysia, 43400, Serdang, Selangor Darul Ehsan; e-mail: muzafar@econ.upm.edu.my The authors would like to thank the anonymous referees for their constructive comments and suggestions; all remaining errors are their own Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 428 Margin—The Journal of Applied Economic Research : (2010): 427–461 INTRODUCTION Since 1997, when Thailand suffered severe economic damage due to the Asian financial crisis, major structural changes have occurred in the Thailand banking sector The sharp decline in the domestic currency had damaging effects on leading banks’ balance sheets and their capital adequacy In response to the depreciating exchange rate, Bank of Thailand (the central bank of Thailand) raised interest rates on deposits This resulted in a decline in bank revenues, as banks could not pass on the higher interest rates to distressed corporate borrowers, resulting in negative interest rate spreads, and subsequently reducing banks’ net income Prior to the Asian financial crisis in 1997, the Thailand banking sector has been sheltered from foreign competition In the aftermath of the crisis the government launched two major strategies to revive the financial sector First, several ailing financial institutions were nationalised or merged with other Thai commercial banks (for example, the assets of Bangkok Bank of Commerce were transferred to Krung Thai Bank and Union Bank was merged with Krung Thai Bank), or acquired by foreign banks (for example, Radanasin Bank was acquired by United Overseas Bank Limited, Nakornthon Bank was acquired by Standard Chartered Bank and Thai Danu Bank was acquired by DBS Bank) Second, which was a key element of Thailand’s plan, the banking sector was re-capitalised by relaxing regulation on foreign shareholding limits in Thai commercial banks Thai authorities allowed foreign investors to hold more than 49 per cent of the share in Thai banking markets for up to 10 years, as against the 25 per cent foreign shareholding limit before the Asian financial crisis.1 At the end of 2008, the Thai banking sector’s total assets, deposits and loans were dominated by the ‘Big 4’ As can be seen from Table 1, the concentration ratio in terms of total assets and loans of the largest four banks (CR4) remains Under the Financial Institutions Business Act (2008), which came into force on August 2008, a foreign bank may own Up to 25 per cent of a Thai bank, without approval by the Bank of Thailand This percentage also applies to its shareholding in any parent company, and also to voting rights Up to 25 per cent of the directors may be foreign Up to 49 per cent of a Thai bank, subject to approval by the Bank of Thailand Up to half of the directors may be foreign More than 49 per cent of a Thai bank, and appoint foreigners to be more than half of the directors, subject to approval by the Ministry of Finance However, these rules are applicable only to foreign ownership of Thai registered banks, and not to the branches of foreign-owned banks operating in the Thai banking sector Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 1,659,844 1,327, 184 1,303,552 1,228,494 742,576 599,389 414,400 368,272 290,638 216,535 1,182,878 994,819 – 685,387 441,594 335,007 270,787 – 63,481 60,512 Panel A: 2008 Bangkok Bank Krung Thai Bank Kasikornbank Siam Commercial Bank Bank of Ayudhya TMB Bank Siam City Bank Thanachart Bank Standard Chartered Bank (Thai) United Overseas Bank (Thai) Panel B: 1999 Bangkok Bank Krung Thai Bank Kasikornbank Siam Commercial Bank Bank of Ayudhya TMB Bank Siam City Bank Thanachart Bank Standard Chartered Bank (Thai) United Overseas Bank (Thai) Bank Total Assets 957,172 802,102 – 562,642 356,098 259,843 215,832 – 48,130 42,051 1,311,477 1,063,532 968,788 913,534 540,747 450,560 343,846 270,832 129,517 161,903 Total Deposits 754,898 708,921 – 468,851 327,192 261,081 186,318 – 46,831 3,341 1,111,948 1,010,687 872,085 854,142 516,717 379,820 263,334 266,540 84,560 152,550 Total Loans Deposits 21.8 18.3 – 12.6 8.1 6.2 5.0 – 1.2 1.1 21.8 18.3 – 12.8 8.1 5.9 4.9 – 1.1 1.0 Domestic Banks 17.70 18.64 14.15 15.11 13.90 13.77 13.10 12.98 7.92 7.68 6.39 6.40 4.42 4.89 3.93 3.85 3.10 1.84 2.31 2.30 Assets Market Share 20.5 19.3 – 12.7 8.9 7.1 5.1 – 1.3 0.1 17.27 15.70 13.54 13.26 8.02 5.90 4.09 4.14 1.31 2.37 Loans Table Structure of the Thai Financial System—Top 10 Domestic and Foreign Banks 0.5 –0.5 – 0.1 –0.6 0.2 12.7 – 1.1 –1.2 1.2 0.9 1.2 1.7 0.7 0.1 0.9 0.4 0.7 0.6 Return on Assets (ROA) (Table continued) 570 610 – 477 403 364 210 – 45 37 868 832 688 949 579 470 407 216 39 152 No of Branches Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 246,141 201,401 196,117 159,193 149,957 119,958 73,149 35,399 25,144 20,722 107,316 89,933 – 59,829 – 48,382 32,338 – 34,085 13,547 Bank Panel C: 2008 Bank of Tokyo Mitsubishi UFJ Citibank Sumitomo Mitsui Banking Corporation The Hongkong and Shanghai Banking Corp Mizuho Corporate Bank Deutsche Bank JP Morgan Chase Bank Calyon Corporate and Investment Bank ABN-AMRO Bank BNP Paribas Panel D: 1999 Bank of Tokyo Mitsubishi UFJ Citibank Sumitomo Mitsui Banking Corporation The Hongkong and Shanghai Banking Corp Mizuho Corporate Bank Deutsche Bank JP Morgan Chase Bank Calyon Corporate and Investment Bank ABN-AMRO Bank BNP Paribas Source: Bank of Thailand Total Assets (Table continued) 42,157 49,166 – 17,985 – 12,505 3,720 – 2,154 1,471 124,854 107,846 73,607 63,173 60,709 24,747 7,123 7,282 2,887 5,515 Total Deposits 83,503 61,794 – 25,628 – 14,322 28,019 – 18,382 10,788 161,397 75,278 117,601 52,585 94,909 14,355 669 11,593 7,578 2,573 Total Loans Deposits 14.0 11.7 – 7.8 – 6.3 4.2 – 4.5 1.8 19.5 22.8 – 8.3 – 5.8 1.7 – 1.0 0.7 Foreign Banks 2.62 1.77 2.15 1.53 2.09 1.05 1.70 0.90 1.60 0.86 1.28 0.35 0.78 0.10 0.38 0.10 0.27 0.04 0.22 0.08 Assets Market Share 16.6 12.3 – 5.1 – 2.9 5.6 – 3.7 2.2 2.51 1.17 1.83 0.82 1.47 0.22 0.01 0.18 0.12 0.04 Loans – – – – – – – – – – – – – – – – – – – – No of Branches 0.1 2.0 – 3.7 – 0.9 1.5 – 0.9 –1.1 1.1 1.7 1.4 1.4 – 1.6 3.1 2.1 –0.9 –1.4 Return on Assets (ROA) Sufian and Habibullah DETERMINANTS OF BANK EFFICIENCY 431 high at 58.9 per cent and 59.8 per cent respectively, compared to 60.8 per cent and 61.4 per cent, respectively, during 1999 Similarly, the ‘Big 4’ also controlled more than 60 per cent of the Thai banking sector’s deposits market in 2008, slightly lower than the 61.0 per cent recorded in 1999 In total, domestic commercial banks controlled 86.9 per cent, 87.5 per cent and 85.6 per cent of the banking sector’s total assets, total deposits and total loans, respectively, in 2008 compared to 74.3 per cent, 73.9 per cent and 75.0 per cent, respectively, during 1999 (Table 1) In terms of branch networks, Siam Commercial Bank has the widest coverage across the country at the end of 2008, with 949 branches, followed by Bangkok Bank and Krung Thai Bank with 868 and 832 branches, respectively On the other hand, during 1998 Krung Thai Bank with 610 branches ranked first in terms of branch networks, followed by the Bangkok Bank and Siam Commercial Bank with 570 and 477 branches, respectively Finally, it can be observed that foreign banks have been relatively more profitable than their domestic bank counterparts: At end 2008, foreign banks have reported a return on average assets (ROAA) of 1.1 per cent (stable since 1999) compared to their domestic commercial bank peers ROAA of 0.8 per cent (a fall from 1.5 per cent in 1999) It is reasonable to assume that these developments posed great challenges to financial institutions in Thailand as the environment in which they operated changed rapidly, and this is turn had an impact on the determinants of efficiency of Thailand banks By analysing all the Thailand commercial banks, this article seeks to examine the efficiency of the Thai banking sector during the post-Asian financial crisis period of 1999–2008, a time of reform in the country’s financial sector While there has been extensive literature examining the efficiency of financial sectors in developed countries, empirical works on factors that influence the performance of financial institutions in developing economies are relatively scarce This article is structured as follows The next section reviews related studies, followed by a section that outlines the approach to the measurement and estimation of efficiency change, the econometric framework, and details on the construction of the dataset Section discusses the results, and finally Section provides some concluding remarks RELATED STUDIES Since its introduction by Charnes et al (1978), researchers have welcomed the data envelopment analysis (DEA) method as a methodology for performance Margin—The Journal of Applied Economic Research : (2010): 427–461 Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 432 Margin—The Journal of Applied Economic Research : (2010): 427–461 evaluation (Gregoriou and Zhu, 2005) The first application of the DEA method by Sherman and Gold (1985) analysed the efficiency of 14 branches of a US savings bank They employed 17 types of transactions processed by the bank branches as output variables, while three variables, namely, labour, office space and supply costs, were used as inputs The DEA results showed that six branches were operating inefficiently compared to others A similar study by Parkan (1987) who analysed 35 bank branches of a major Canadian chartered bank suggested that 11 branches were relatively inefficient In his study, Parkan (1987) employed the DEA method by using six input variables, namely, (i) total authorised full-time employees, (ii) annual rent, (iii) quality of customer service space ranking, (iv) telephone/stationery expenses, (v) number of online terminals and (vi) marketing activity ranking, while (i) number of transactions, (ii) commercial account openings, (iii) retail account openings, (iv) number of loan applications, (v) customer service survey rating and (vi) number of corrections were considered output variables Rangan et al (1988) shifted the unit of assessment from bank branches to consolidated banking institutions They applied the DEA method to a larger sample of 215 US banks during 1986 and attempted to break down inefficiency to that stemming from pure technical inefficiency (PTIE) and from scale inefficiency (SIE) They employed the intermediation approach by using three inputs, namely, labour, capital and purchased funds, and five outputs, namely, three types of loans and two categories of deposits Their results indicated that banks could have produced the same amount of output with only 70 per cent of the inputs actually used On the other hand, they found that scale inefficiencies of the banks were relatively small, implying that the sources of inefficiency were mainly technical Aly et al (1990) extended the work of Rangan et al (1988) and examined whether there were differences in allocative efficiency between unit and branch banking organisational forms of 322 US banks during 1986 They applied the DEA method to five outputs, namely, (i) real-estate loans, (ii) commercial and industrial loans, (iii) consumer loans, (iv) all other loans and (v) demand deposits; and to three inputs, namely, (i) labour, (ii) capital and (iii) loanable funds To compute cost efficiency they employed three input price variables, namely, the price of (i) labour, (ii) capital and (iii) loanable funds They concluded that the main source of inefficiency was technical in nature, rather than allocative or due to scale effects Similar to the findings of Aly et al (1990), Yue’s (1990) findings showed that PTIE is the major source of overall technical inefficiency, rather than SIE He employed the DEA method to four inputs, namely, interest expenses, non-interest expenses, transaction deposits and Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 Sufian and Habibullah DETERMINANTS OF BANK EFFICIENCY 433 non-transaction deposits, and three outputs, namely, interest income, noninterest income and total loans, to examine the efficiency of 60 commercial banks in Missouri, US between 1984 and 1990 The DEA method has also been widely employed to examine the efficiency of European countries’ banking sectors Among earlier studies applying the DEA method to European banking sectors were Berg et al (1993) who examined the efficiency of Nordic countries’ banking sectors during 1990 Their sample comprised 503 Finnish banks, 150 Norwegian banks and 126 Swedish banks, which was practically the entire banking industry of the three countries They used two inputs, namely, labour and capital, and three outputs, namely, total loans to other financial institutions, total deposits from other financial institutions and number of branches When analysing individual countries, they found that efficiency spreads between banks were most important in Finland and Norway, and least important in Sweden A comparison of the best-practice frontiers of the three countries yields the result that the highest share of banks on the frontier belongs to Sweden and the lowest share belongs to Finnish banks They also found that the average Swedish banks were much more efficient than the average Finnish or Norwegian banks Apart from their heavy application in the US and European banking sectors, the DEA method has fast become a popular method in assessing financial institutions’ efficiency among banking researchers in other nations Fukuyama (1993) was among the earlier researchers in Asia to examine bank efficiency by employing the DEA method By employing labour, capital and funds from customers as input variables and revenue from loans and other business activities as outputs, Fukuyama (1993) examined the efficiency of 143 Japanese banks during 1990 He found that pure technical efficiency (PTE) averaged around 86 per cent and scale efficiency (SE) around 98 per cent The SIE is found to be mainly due to increasing returns to scale He also found that banks of different organisational status perform differently with respect to all efficiency measures (that is, overall, pure technical and scale) Scale efficiency is found to be positively but weakly associated with bank size Among the notable studies on the Thai banking sector is the one by Leightner and Lovell (1998) who examined the efficiency of 31 banks operating in the Thai banking sector from 1989 to 1994 They applied the DEA method to three input variables, namely, personnel expenses, premise and equipment expense, while the output variables consist of net interest income, non-interest income, credits granted and investment in securities They found that the large domestic banks were most efficient, while foreign-owned banks have slightly higher efficiency levels compared to medium-sized domestic banks Smaller Margin—The Journal of Applied Economic Research : (2010): 427–461 Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 434 Margin—The Journal of Applied Economic Research : (2010): 427–461 domestic banks were found to be the most inefficient A more recent study by Williams and Intarachote (2003) conclude that the efficiency of domestic and foreign commercial banks operating in the Thai banking sector during 1990–97 were comparable They employ the stochastic frontier analysis (SFA) method to a balanced panel data of 29 Thai commercial banks and suggest that Japaneseowned banks operating in the Thailand banking sector are significantly more efficient compared to their domestic counterparts and foreign banks Their results support the view that foreign banks from strong home environments may carry efficiency advantages overseas To the best of our knowledge, apart from the few studies discussed earlier, virtually nothing has been published on the efficiency of the Thailand banking sector in the post-Asian financial crisis period In the light of these knowledge gaps, this article seeks to provide new empirical evidence on the sources and determinants of Thailand’s banking sector production efficiency DATA AND METHODOLOGY To measure the input-oriented technical efficiency of the Thailand banking sector, this article employs the non-parametric DEA method, first introduced by Charnes et al (1978) (hereafter the CCR model) The CCR model presupposes there is no significant relationship between the scale of operations and efficiency by assuming constant returns to scale (CRS) and it delivers overall technical efficiency (TE) Therefore, the CRS assumption is only justifiable when all decision making units (DMUs) are operating at an optimal scale However, banks in practice may face either economies or diseconomies of scale Thus, if one makes the CRS assumption when not all DMUs are operating at the optimal scale, the computed measures of TE will be contaminated with SE Banker et al (1984) extended the CCR model by relaxing the CRS assumption The resulting BCC model is used to assess the efficiency of DMUs characterised by variable returns to scale (VRS) The VRS assumption provides the measurement of PTE, which is the measurement of TE devoid of SE effects An apparent difference between the TE and PTE scores of a particular DMU indicates the existence of SIE, that is, TE PTE SE The former relates to the capability of managers to use banks’ given resources, whereas the latter refers to exploiting scale economies by operating at a point where the production frontier exhibits CRS To make a detailed analysis of the inefficient units, the present article considers the VRS assumption The DEA method involves constructing a non-parametric production frontier based on actual input–output observations in the sample relative to which the Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 Sufian and Habibullah DETERMINANTS OF BANK EFFICIENCY 435 efficiency of each bank in the sample is measured (Coelli, 1996) Let us give a short description of the DEA method Assume that there is data on K inputs and M outputs for each N bank For the ith bank these are represented by the vectors xi and yi, respectively Let us call the K N input matrix, X, and the M N output matrix, Y To measure the efficiency for each bank, the ratio of all inputs is computed, such as (u’yi/v’xi) where u is an M vector of output weights and v is a K vector of input weights To select optimal weights we specify the following mathematical programming problem: (u’yi /v’xi), u,v u’yi /v’xi 1, j 1, 2, …, N, u,v (1) The formulation has a problem of infinite solutions and therefore we impose the constraint v’xi 1, which leads to: (’yi), , ’xi ’yi – ’xj j 1, 2, …, N, , (2) where we change the notation from u and v to and , respectively, in order to reflect transformations Using the duality in linear programming, an equivalent envelopment form of this problem can be derived: , , yi Y xi – X 0 (3) where is a scalar representing the value of the efficiency score for the ith bank which will range between and is a vector of N constants The linear programming has to be solved N times, once for each bank in the sample In order to calculate efficiency under the assumption of VRS, the convexity constraint (N1 = 1) will be added to ensure that an inefficient bank is only compared against banks of similar size, and therefore provides the basis for measuring economies of scale within the DEA method Margin—The Journal of Applied Economic Research : (2010): 427–461 Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 436 Margin—The Journal of Applied Economic Research : (2010): 427–461 3.1 Multivariate Regression Analysis The DEA method has the disadvantage that it interprets random errors as inefficient, making it sensitive to outliers and degrees of freedom However, Banker (1993) and Banker and Natarajan (2004) provide proof that the efficiency i is a consistent estimator Furthermore, in an influential development, Banker and Natarajan (2008) provide further proof that the use of a two-stage procedure involving DEA followed by an ordinary least square (OLS) regression yields consistent estimators of the regression coefficients Furthermore, in an important development, McDonald (2009) provided the statistical foundation that the use of DEA and OLS is a consistent estimator, and if White’s (1980) heteroskedastic consistent standard errors are calculated, large sample tests can be performed which are robust to heteroskedasticity and the distribution of the disturbances.2 Thus, following Banker and Natarajan (2008) among others, as a robustness check, equation (4) is also re-estimated by using the OLS method Furthermore, as suggested by McDonald (2009) we estimate equation (4) by using White’s (1980) transformation which is robust to heteroskedasticity and the distribution of the disturbances in the second-stage regression analysis involving DEA scores as the dependent variable Following De Bandt and Davis (2000) and Staikouras et al (2008) among others, the log-linear form is chosen as it typically improves the regression’s goodness-of-fit and may reduce simultaneity bias By using the TE score as the dependent variable, the following regression model is estimated: jt 0 1LN(LOANS/TA)jt 2LN(TA)jt 3LN(LLP/TL)jt 4LN(NII/TA)jt 5LN (NIE/TA)jt 6(LN(EQASS)jt 7LN(ROA)jt 1LN(GDP)t 2LN(INFL)t 3LN(CR3)t 4LN(MKTCAP/GDP)t 5DUMCRISt 1DUMFORBjt jt (4) where ‘i’ denotes the bank, ‘t’ the examined time period, and is the disturbance term LOANS/TA is a measure of a bank’s loans intensity calculated as the ratio of total loans to total assets; TA is used as a proxy measure of bank size in terms of The findings by McDonald (2009) are further supported by Estelle et al (2010) who suggest that regression models involving the DEA method in the first stage are robust with respect to the selection of OLS, flogit or non-parametric regression models in the second-stage analysis Furthermore, Hoff (2007) also suggests that the Tobit regression is adequate to represent the second-stage regression models, while the OLS regression models may sufficiently replace the Tobit regression analysis as a second-stage DEA model Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 Sufian and Habibullah DETERMINANTS OF BANK EFFICIENCY 447 have reported a higher mean technical efficiency compared to their large bank counterparts The empirical findings concur with the earlier studies by, among others, Hauner (2005) and Sufian (2007), which have found that small-sized banks are relatively more efficient compared to their large- and medium-sized bank counterparts It is observed from Panel A of Table that during the period under study, foreign banks exhibit a mean technical efficiency of 83.5 per cent, and that PTIE outweighs SIE in determining total technical inefficiency of foreign banks It is interesting to note from Panel B of Table that domestic banks have a higher Table Summary of Efficiency Scores by Ownership Bank Panel A: Foreign Banks Mega International Commercial Bank PCL United Overseas Bank (Thailand) Standard Chartered Nakornthon Bank DBS Thai Danu Bank PCL UOB Radanasin Bank PCL Mean Panel B: Domestic Banks Land and Houses Retail Bank PCL Bangkok Bank PCL Thai Farmers Bank PCL Krung Thai Bank PCL Thai Credit Retail Bank PCL Kasikornbank PCL Siam Commercial Bank PCL Government Savings Bank Bank of Ayudhya PCL Thai Military Bank PCL Government Housing Bank Siam City Bank PCL Thanachart Bank PCL Bankthai PCL Bangkok Metropolitan Bank PCL Tisco Bank PCL Kiatnakin Bank PCL Small and Medium Enterprise Development Bank of Thailand ACL Bank PCL Mean Mean TE Mean PTE Mean SE 1.000 0.707 0.744 0.730 0.994 0.835 1.000 0.771 0.777 0.737 1.000 0.857 1.000 0.919 0.947 0.992 0.994 0.970 0.983 0.854 0.777 0.829 1.000 0.804 0.846 0.892 0.801 0.816 1.000 0.876 0.990 0.839 0.856 0.909 0.976 0.848 1.000 0.996 0.986 0.945 1.000 0.962 0.991 1.000 0.925 0.891 1.000 0.927 1.000 0.924 0.907 0.956 0.994 1.000 0.983 0.858 0.788 0.882 1.000 0.840 0.853 0.892 0.868 0.914 1.000 0.946 0.990 0.909 0.945 0.951 0.981 0.848 0.973 0.888 1.000 0.969 0.973 0.917 Source: Authors’ own calculation Margin—The Journal of Applied Economic Research : (2010): 427–461 Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 448 Margin—The Journal of Applied Economic Research : (2010): 427–461 level of technical efficiency than their foreign bank peers It is worth noting that while foreign banks’ technical inefficiency was mainly due to PTIE, domestic banks’ technical inefficiency was found to be mainly due to SIE The empirical findings from this study should come as a surprise since earlier studies by, among others, Isik and Hassan (2002), Ataullah and Le (2006), and Havrylchyk (2006), have found that foreign banks in developing and transition countries are relatively more efficient than their domestic bank peers A plausible reason could be that although foreign banks have the ability to capitalise on better risk management and operational techniques provided by their parent banks abroad, the ‘liability of foreignness’ is difficult to overcome in certain regulatory and economic environment Berger et al (2000) pointed out that under the ‘limited form’ of the global advantage hypothesis, only the efficient institutions in one or a limited number of nations with specific favourable market or regulatory conditions in their home countries can operate more efficiently than domestic banks in other nations 4.2 Composition of the Efficiency Frontiers We now turn to a discussion on the sources of the SIE of Thai banks Since the dominant source of total technical (in)efficiency in the Thailand banking sector seems to be scale related, it is worth examining further the trend in the returns to scale of the Thailand banks It is worth noting that a bank can operate at CRS or VRS where CRS signifies that an increase in inputs results in a proportionate increase in outputs and VRS means a rise in inputs results in a disproportionate rise in outputs Further, a bank operating at VRS can be at increasing returns to scale (IRS) or declining returns to scale (DRS) To recap, IRS means that an increase in inputs results in a higher increase in outputs, while DRS indicate that an increase in inputs results in lesser output increases To identify the nature of returns to scale, CRS scores (obtained with the CCR model) are first compared with VRS (by using the BCC model) scores For a given bank, if the VRS score equals its CRS score, the bank is operating at CRS On the other hand, if the scores are not equal, a further step is needed to establish whether the bank is operating at IRS or DRS To this, the DEA model is used under the non-increasing returns to scale (NIRS) assumption If the score under VRS equals the NIRS score, the bank is said to be operating at DRS Alternatively, if the score under VRS is different from the NIRS score, then the bank is said to be operating at IRS (Coelli et al., 1998) Table shows the composition of banks that lie on the efficiency frontiers The composition of the efficiency frontier suggests that the number of banks Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 CRS DRS IRS DRS DRS IRS CRS DRS CRS CRS IRS DRS DRS DRS DRS CRS CRS DRS CRS DRS ACL Bank PCL Bangkok Bank PCL Bangkok Metropolitan Bank PCL Bank of Ayudhya PCL Bankthai PCL DBS Thai Danu Bank PCL Government Housing Bank Government Savings Bank Kasikornbank PCL Kiatnakin Bank PCL Krung Thai Bank PCL Land and Houses Retail Bank PCL 2000 1999 Bank DRS CRS CRS CRS DRS DRS DRS DRS CRS CRS 2001 DRS CRS DRS DRS CRS IRS CRS DRS CRS DRS CRS DRS CRS CRS CRS DRS 2003 CRS DRS 2002 DRS CRS DRS CRS CRS DRS CRS DRS CRS IRS CRS DRS 2005 DRS DRS CRS DRS 2004 CRS CRS DRS IRS IRS IRS DRS IRS IRS CRS 2006 CRS DRS DRS IRS DRS DRS IRS IRS DRS 2008 1 2 10 Count Bank (Table continued) CRS DRS DRS DRS DRS DRS IRS CRS DRS 2007 The table shows the evolution of returns to scale in the Thai banking sector during the period 1999–2008 CRS, DRS and IRS denote constant returns to scale, decreasing returns to scale and increasing returns to scale respectively ‘Count Year’ denotes the number of banks appearing on the efficiency frontier during the year ‘Count Bank’ denotes the number of times a bank has appeared on the efficiency frontier during the period of study Banks which correspond to the shaded regions have not been efficient in any year in the sample period compared to the other banks in the sample Table Thai Banks—Composition of Efficiency Frontiers Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 Source: Authors’ own calculation Mega International Commercial Bank PCL Siam City Bank PCL Siam Commercial Bank PCL Small and Medium Enterprise Development Bank of Thailand Standard Chartered Nakornthon Bank Thai Credit Retail Bank PCL Thai Farmers Bank PCL Thai Military Bank PCL Thanachart Bank PCL Tisco Bank PCL United Overseas Bank (Thailand) UOB Radanasin Bank PCL Count Year Bank (Table continued) DRS DRS IRS DRS CRS CRS IRS DRS CRS IRS DRS DRS CRS DRS DRS CRS 2000 DRS DRS 1999 DRS CRS DRS DRS CRS DRS CRS DRS 2001 DRS CRS CRS DRS CRS IRS CRS DRS 2002 DRS CRS CRS DRS DRS IRS CRS DRS 2003 DRS CRS CRS DRS CRS IRS DRS DRS 2004 IRS CRS CRS DRS 10 10 CRS CRS CRS CRS CRS IRS 2006 DRS CRS CRS IRS CRS CRS CRS CRS 2005 DRS DRS DRS DRS CRS CRS DRS DRS CRS 2007 DRS CRS IRS DRS CRS CRS DRS DRS IRS 2008 1 5 71 Count Bank Sufian and Habibullah DETERMINANTS OF BANK EFFICIENCY 451 that span the efficiency frontier varies between and 10 The Government Housing Bank and Mega International Bank appear to be the global leaders (from Table 8), as they appeared most often on the efficiency frontier Interestingly, the empirical findings suggest that only (20.8 per cent) banks have never made it to the efficiency frontier throughout the period of study In general, the empirical findings presented in Table clearly indicate that while small banks tend to operate at CRS or IRS, large banks tend to operate at CRS or DRS, findings which are similar to earlier studies by among others Miller and Noulas (1996), McAllister and McManus (1993) and Noulas et al (1990) To recap, McAllister and McManus (1993) suggest that while small banks have generally exhibited IRS, large banks, on the other hand, tend to exhibit DRS and at best CRS As it appears, the small Thailand banks experience IRS in their operations during the period of study One implication is that for small Thailand banks, a proportionate increase in inputs would result in more than a proportional increase in outputs Hence, the small Thailand banks, which have been operating at IRS, could achieve significant cost savings and efficiency gains by increasing the scale of operations In other words, substantial gains can be obtained by altering the scale via internal growth or further consolidation in the sector In fact, in a perfectly competitive and contestable market, the efficient banks should absorb the scale-inefficient banks to exploit cost advantages Thus, banks that experience IRS should either eliminate their SIE or prepare to become a prime target for acquiring banks, which can create value from underperforming banks by streamlining their operations and eliminating their redundancies and inefficiencies (Evanoff and Israelvich, 1991) On the other hand, the results seem to suggest that a further increase in size would only result in a smaller increase in outputs for every proportionate increase in inputs for large banks, as they have been operating at DRS and CRS Hence, decision-makers ought to be more cautious in promoting mergers among large banks as a means to enjoying efficiency gains Overall, the empirical findings from this study seem to suggest that in the Thailand banking sector, technical inefficiency has much more to with the scale of production rather than the inefficient utilisation of resources The dominant effect of the SIE indicates that most of Thailand banks have been operating at the ‘incorrect’ or non-optimal scale of operations They either experience economies of scale (IRS) due to being at less than optimum size, or diseconomies of scale (DRS) from being larger than the optimum size Thus, decreasing or increasing the scale of production could result in cost savings or efficiencies Margin—The Journal of Applied Economic Research : (2010): 427–461 Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 452 Margin—The Journal of Applied Economic Research : (2010): 427–461 4.3 Determinants of Thai Banks’ Efficiency The regression results focussing on the relationship between bank efficiency and the explanatory variables are presented in Table The equations are based on 178 bank year observations during 1999–2008 As pointed out by Saxonhouse (1976), heteroscedasticity can emerge when estimated parameters are used as dependent variables in the second-stage analysis Thus, following among others Hauner (2005), Pasiouras (2008) and Sufian (2009), QML (Huber/White), standard errors and covariates are calculated Several general comments regarding the test results are warranted First, the model performs reasonably well with most variables remaining stable across the various regressions tested And second, the explanatory power of the models is reasonably high, while the F-statistics is significant at the per cent level for all regression models Referring to the impact of bank liquidity, LOANS/TA is positively related to the efficiency of Thai banks indicating a negative relationship between bank efficiency and the level of liquid assets held by banks As higher figures of the ratio denote lower liquidity, the results imply that less (more) liquid banks tend to exhibit higher (lower) efficiency levels The findings imply that banks with higher loans-to-asset ratios tend to exhibit higher technical efficiency scores Thus, bank loans seem to be more highly valued than alternative bank outputs, that is, investments and securities Concerning the impact of bank size, LNTA is negatively related to the efficiency of Thai banks and is statistically significant at the 10 per cent level in the baseline regression model The negative coefficient indicates that larger (smaller) banks tend to exhibit lower (higher) efficiency levels and provides support to earlier studies that found economies of scale and scope for smaller banks or diseconomies of scale for larger banks (Pasiouras and Kosmidou, 2007; Staikouras et al., 2008) This is particularly pronounced among foreign banks; column of Table shows that the coefficient of the variable is relatively strong and becomes significant at the per cent level when we control for foreign banks Thus, assuming that the average cost curve for Thailand banks is U-shaped, recent growth policies of medium and large Thai banks seem inconsistent with cost minimisation As expected, the coefficient of the LLP/TL variable entered all the regression models with a negative sign and is statistically significant at the per cent level The results suggest that Thai banks with higher credit risks tend to exhibit lower efficiency levels The empirical findings imply that the banks should focus more on credit risk management, which has been proven to be problematic in the recent past Serious banking problems have arisen from the failure of banks to recognise impaired assets and create reserves for writing off these assets Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 jt 0 1LOANS/TAjt 2LNTAjt 3LLP/TLjt 4LNNII/TAjt 5NIE/TAjt 6EQASSjt 7ROAjt 1LNGDPt 2INFLt 3CR3t 4MKTCAP/GDPt 5DUMCRISt 1DUMFORBjt jt Panel Regression Analysis Bank Characteristics LOANS/TA CONSTANT Explanatory Variables /– Expected Relationship 0.034 (1.968) 0.659 (2.222) 0.041 (2.373) 0.040 (2.273) TOBIT –1.474 2.008 (–0.814) (2.287) 0.007 (0.392) 2.143 (2.681) 0.025 (2.282) 0.808 (4.077) 0.028 (2.535) 0.000 (0.024) 1.481 (2.697) (Table continued) 0.027 (2.459) OLS –0.738 1.399 (–0.642) (2.376) The dependent variable is bank’s technical efficiency scores derived from the DEA; LOANS/TA is a measure of bank’s loans intensity calculated as the ratio of total loans to bank total assets; LNTA is the size of the bank’s total asset measured as the natural logarithm of total bank assets; LLP/TL is a measure of banks risk calculated as the ratio of total loan loss provisions divided by total loans; NII/TA is a measure of bank diversification towards non-interest income, calculated as total non-interest income divided by total assets; NIE/TA is a measure of bank management quality calculated as total non-interest expenses divided by total assets; EQASS is a measure of banks leverage intensity measured by banks total shareholders equity divided by total assets; ROA is a proxy measure for bank’s profitability calculated as bank’s profit after tax divided by total assets; LNGDP is natural logarithm of gross domestic product; INFL is the inflation rate; CR3 is the three bank concentration ratio; MKTCAP/GDP is the ratio of stock market capitalisation divided by GDP; DUMCRIS is a dummy variable that takes a value of for the global financial crisis period, otherwise; DUMFORB is a dummy variable that takes a value of if a bank is foreign owned bank, otherwise Values in parentheses are standard errors , and indicate significance at 1, and 10 per cent levels Table Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 /– – – LLP/TL NII/TA NIE/TA EQASS ROA Macroeconomic Conditions LNGDP Expected Relationship LNTA Explanatory Variables (Table continued) –0.165 (–1.516) 0.264 (1.185) –0.150 (–1.495) TOBIT –0.017 –0.020 –0.019 –0.040 –0.007 (–1.544) (–1.125) (–1.776) (–1.709) (–3.445) –0.050 –0.057 –0.058 –0.069 –0.031 (–2.743) (–3.026) (–3.177) (–4.165) (–2.905) –0.012 –0.028 –0.027 0.005 –0.094 (–0.207) –0.466) (–0.452) (0.141) (–1.681) –0.035 –0.026 –0.031 0.001 –0.043 (–0.768) (–0.573) (–0.707) (0.014) (–1.386) 0.121 0.126 0.118 0.103 0.070 (2.616) (2.433) (2.294) (0.030) (2.230) 0.039 0.039 0.042 0.062 0.000 (0.724) (0.687) (0.727) (1.557) (0.009) –0.063 (–0.887) –0.006 (–1.042) –0.033 (–3.057) 0.001 (0.023) –0.042 (–1.273) 0.068 (2.388) –0.008 (–0.200) 0.200 (1.422) –0.008 (–1.263) –0.033 (–3.116) 0.003 (0.083) –0.045 (–1.387) 0.063 (2.242) –0.005 (–0.128) OLS –0.050 (–0.737) –0.021 (–3.402) –0.043 (–4.363) –0.041 (–1.282) –0.026 (–0.928) 0.051 (2.167) 0.005 (0.188) Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 – DUMCRIS Bank Ownership DUMFORB Source: Authors’ own calculation R2 Adj R2 Log Likelihood F-statistic No of Obs – /– CR3 MKTCAP/GDP /– INFL 0.111 0.046 –13.402 178 0.111 0.069 –14.847 178 0.019 (0.760) 0.139 (0.422) 0.035 (0.879) 178 0.141 0.073 –10.801 –0.026 (–0.811) 0.032 (0.097) –0.029 (–0.593) –0.169 (–2.332) 178 –0.207 (–5.466) 0.293 0.236 0.491 0.015 (0.665) 0.114 (0.386) 0.043 (1.214) 0.179 0.125 0.202 0.144 –0.017 (–0.807) 0.058 (0.287) –0.019 (–0.595) –0.105 (–2.315) –0.150 (–5.303) 0.325 0.276 0.008 (0.550) 0.099 (0.527) 0.027 (1.101) 5.000 3.295 3.482 6.615 178 178 178 178 0.171 0.137 0.011 (0.648) 0.126 (0.615) 0.020 (0.764) 456 Margin—The Journal of Applied Economic Research : (2010): 427–461 Smoothing these anomalies could take place by improving transparency in the banking system, which in turn will assist banks to evaluate credit risk more effectively and avoid problems associated with hazardous exposure The impact of NII/TA is negative (statistically significant at the 10 per cent level) when foreign banks are controlled for in the sample (Table 9) The results imply that banks which derive a higher proportion of income from non-interest sources such as foreign exchange trading tend to report lower efficiency levels The empirical findings provide support to earlier studies by among others Stiroh (2006a, 2006b) and Stiroh and Rumble (2006) To recap, Stiroh and Rumble (2006) suggest that diversification benefits of US financial holding companies are offset by the increased exposure to non-interest activities, which are much more volatile but not necessarily more profitable than interest-generating activities During the period under study, the empirical findings seem to suggest that the impact of NIE/TA is negative The results imply that an increase (decrease) in these expenses reduces (increases) the efficiency of banks operating in Thailand Pasiouras and Kosmidou (2007) and Kosmidou (2008) among others have also found poor expense management to be among the main contributors to poor bank performance Clearly, efficient cost management is a prerequisite for the improved efficiency of Thailand’s banking sector Furthermore, Thailand’s banking sector has not reached the maturity level required to link quality effects from increased spending to higher bank efficiency However, the results need to be interpreted with caution since the coefficient of the variable is not statistically significant at any conventional levels Capital strength as measured by EQASS is positively related to Thai banks’ efficiency and is statistically significant at the per cent level or better in both the Tobit and OLS regression models The empirical finding is consistent with Berger (1995), Demirguc-Kunt and Huizinga (1999), Staikouras and Wood (2003), Goddard et al (2004), Pasiouras and Kosmidou (2007) and Kosmidou (2008) providing support to the argument that well-capitalised banks face lower costs of going bankrupt Furthermore, a strong capital structure is essential for banks in developing economies, since it provides them additional strength to withstand financial crises and increased safety for depositors during unstable macroeconomic conditions However, it is worth noting that the coefficient loses its explanatory power when we control for macroeconomic and market conditions Gross domestic product has a negative relationship with Thai bank efficiency (Table 9) Demand for financial services tends to grow as economies expand and societies become wealthier However, volatile economic growth during the period under study could have resulted in banks suffering from lower Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 Sufian and Habibullah DETERMINANTS OF BANK EFFICIENCY 457 demand for their financial services and higher loan defaults On the other hand, inflation (INFL) exhibits a positive sign, implying that during the period under study, levels of inflation were anticipated by these banks This gave them the opportunity to adjust interest rates accordingly and consequently to earn higher profits The result is consistent with findings by Pasiouras and Kosmidou (2007) among others During the period under study, the empirical findings suggest that the impact of concentration (CR3) is positively related to Thai banks’ efficiency They also appear to support the Structure-Conduct-Performance (SCP) hypothesis To recap, the SCP hypothesis states that banks in highly concentrated markets tend to collude and therefore earn monopoly profits (Gilbert, 1984; Molyneux et al 1996; Short, 1979) The impact of stock market capitalisation (MKTCAP) on bank efficiency is positive, implying that during the period under study the Thai stock market offered substitution possibilities to potential borrowers However, it is interesting to note that the coefficient of the variable becomes negative when we control for the recent global financial crisis period (DUMCRIS), suggesting that the stock market exerted a negative influence on the efficiency of the Thai banking sector during the recent period of economic turbulence It is interesting to note that the coefficient of DUMFORB entered both the Tobit and OLS regression models with a negative sign and is statistically significant at the per cent level The empirical findings clearly support the ‘liability of foreignness’ hypothesis Although foreign banks have the ability to capitalise on better risk management and operational techniques provided by their parent banks abroad, the ‘liability of foreignness’ is difficult to overcome in certain regulatory and economic environments Berger et al (2000) pointed out that under the ‘limited form’ of the global advantage hypothesis, only efficient institutions in one or a limited number of nations with specific favourable market or regulatory conditions in their home countries can operate more efficiently than domestic banks in other nations CONCLUSIONS AND AVENUES FOR FUTURE RESEARCH This article attempts to empirically analyse the efficiency of the Thai banking sector during the post-Asian financial crisis period, which is a period of significant reform in the country’s banking sector The efficiency estimates of individual banks are evaluated using the DEA method The analysis is confined to the universe of domestic and foreign commercial banks that have been operating in Thailand during 1999–2008 Margin—The Journal of Applied Economic Research : (2010): 427–461 Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 458 Margin—The Journal of Applied Economic Research : (2010): 427–461 Empirical findings indicate that the efficiency of the Thai banking sector reached a peak of 92.8 per cent during 2006 before declining to 84.1 per cent during 2008 Despite the decline, the results suggest that the Thai banking sector was relatively more efficient during 2008 compared to 1999 During the period under study, the empirical findings suggest that SIE outweighs PTIE in determining the banking sector’s technical efficiency The Thai banking sector exhibited mean technical inefficiency of 13.3 per cent mainly attributed to scale rather than PTIE Further, small banks appear to be most efficient, while medium-sized banks are the least efficient banking group As expected, results from the multivariate regression analysis indicate that the relatively better capitalised banks with higher loans intensity tend to exhibit higher efficiency levels On the other hand, credit risk is negatively related to bank efficiency During the period under study, domestic banks have been relatively more efficient compared to their foreign bank peers supporting the ‘liability of foreignness’ hypothesis Although foreign banks have the ability to capitalise on better risk management and operational techniques provided by their parent banks abroad, the ‘liability of foreignness’ is difficult to overcome in certain regulatory and economic environment Berger et al (2000) pointed out that under the ‘limited form’ of the global advantage hypothesis, only efficient institutions in one or a limited number of nations with specific favourable markets or regulatory conditions in their home countries can operate more efficiently than domestic banks in other nations The empirical findings suggest that the recent global financial crisis exerts a negative impact on the efficiency of Thailand’s banking sector The findings have considerable policy relevance First, the empirical findings imply that banks operating in the Thai banking sector are either too small to benefit from economies of scale or too large to be scale efficient Thus, from the policy-making perspective, relatively smaller banks could raise their efficiency levels by expanding, while larger banks would need to scale down their operations to be scale efficient Second, in terms of SE, larger banks are lagging behind their smaller counterparts The optimal size for a firm would be when it achieves CRS Larger banks may be underperforming in comparison to their smaller peers because their size has become more of a burden than an advantage arising from mergers and acquisitions There are also considerable costs associated with the management of a large organisation and it is important to make sure these costs not outweigh the size benefits Finally, the finding that Thailand banks have been operating at a non-optimal scale of operations is in line with findings by among others Sufian (2007) on Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 Sufian and Habibullah DETERMINANTS OF BANK EFFICIENCY 459 the Malaysian banking sector To recap, Sufian (2007) found that during the post-merger period the inefficiency of the Malaysian banking sector was largely due to scale rather than pure technical inefficiencies, suggesting that mergers were particularly successful for small and medium-sized banks, which have benefited most from expansion Due to its limitations, the article could be extended in a variety of ways First, the scope of this study could be further extended to investigate changes in cost, allocative and technical efficiencies over time Second, future research into the efficiency of the Thai banking sector could consider the production function along with the intermediation function Finally, investigation of changes in productivity over time as a result of technical change or technological progress by employing the Malmquist Total Factor Productivity Index (MPI) could yet be another extension to the article References Aly, H.Y., Grabowski, R., Pasurka, C., & Rangan, N (1990) Technical, scale and allocative efficiencies in U.S banking: An empirical investigation, Review of Economics and Statistics, 72(2), 211–18 Ariff, M., & Can, L (2007) Cost and profit efficiency of Chinese banks: A non-parametric analysis, China Economic Review, 19(2), 260–73 Ataullah, Ali, & Le, Hang (2006) Economic reforms and bank efficiency in developing countries: The case of the Indian banking industry, Applied Financial Economics, 16(9), 653–63 Banker, R (1993) Maximum likelihood, consistency, and data envelopment analysis: A statistical foundation, Management Science, 39(10), 1265–73 Banker, R.D., & Natarajan, R (2004) Productivity change, technical progress, and relative efficiency change in the public accounting industry, Management Science, 51(2), 291–304 ——— (2008) Evaluating contextual variables affecting productivity using data envelopment analysis, Operations Research, 56(1), 48–58 Banker, R.D., Charnes, A., & Cooper, W.W (1984) Some models for estimating technical and scale inefficiencies in data envelopment analysis, Management Science, 30(9), 1078–92 Benston, G.J (1965) Branch banking and economies of scale, Journal of Finance, 20(2), 312–31 Berg, S.A., Førsund, F.R., Hjalmarsson, L., & Suominen, M (1993) Banking efficiency in the Nordic countries, Journal of Banking and Finance, 17(2–3), 371–88 Berger, A.N (1995) The relationship between capital and earnings in banking, Journal of Money, Credit and Banking, 27(2), 432–56 Berger, A.N., & Humphrey, D.B (1997) Efficiency of financial institutions: International survey and directions for future research, European Journal of Operational Research, 98(2), 175–212 Berger, A.N., DeYoung, R., Genay, H., & Udell, G.F (2000) Globalization, of financial institutions: Evidence from cross-border banking performance, Brookings-Wharton Papers on Financial Services, 3(1), 23–158 Margin—The Journal of Applied Economic Research : (2010): 427–461 Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 460 Margin—The Journal of Applied Economic Research : (2010): 427–461 Charnes, A., Cooper, W.W., & Rhodes, E (1978) Measuring the efficiency of decision making units, European Journal of Operational Research, 2(4), 429–44 Charnes, A., Clark, C.T., Cooper, W.W., & Golany, B (1985) A developmental study of data envelopment analysis in measuring the efficiency of maintenance units in the US air forces, Annals of Operations Research, 2(1), 95–112 Coelli, T (1996) A guide to DEAP version 2.1, CEPA Working Paper 8/96, University of New England, Armidale Coelli, T., Prasada-Rao, D.S., & Battese, G.E (1998) An introduction to efficiency and productivity analysis Boston: Kluwer Academic Publishers De Bandt, O., & Davis, P (2000) Competition, contestability and market structure in European banking sectors on the eve of EMU, Journal of Banking and Finance, 24(6), 1045–66 Demirguc-Kunt, A., & Huizinga, H (1999) Determinants of commercial bank interest margins and profitability: Some international evidence, World Bank Economic Review, 13(2), 379–408 Estelle, S.M., Johnson, A.L., & Ruggiero, J (2010) Three-stage DEA models for incorporating exogenous inputs, Computers and Operations Research, 37(6), 1087–90 Evanoff, D.D., & Israelvich, P.R (1991) Productive efficiency in banking, Economic Perspectives, Federal Reserve Bank of Chicago (July/August), 11–32 Fukuyama, H (1993) Technical and scale efficiency of Japanese commercial banks: A nonparametric approach, Applied Economics, 25(8), 1101–12 Gilbert, R (1984) Bank market structure and competition—A survey, Journal of Money, Credit and Banking, 16(4), 617–45 Goddard, J., Molyneux, P., & Wilson, J.O.S (2004) Dynamic of growth and profitability in banking, Journal of Money, Credit and Banking, 36(6), 1069–90 Gregoriou, G.N., & Zhu, J (2005) Evaluating hedge funds and CTA performance: Data envelopment analysis approach New York: John Wiley and Sons Hauner, D (2005) Explaining efficiency differences among large German and Austrian banks, Applied Economics, 37(9), 969–80 Havrylchyk, O (2006) Efficiency of the Polish banking industry: Foreign versus domestic banks, Journal of Banking and Finance, 30(7), 1975–96 Hoff, A (2007) Second stage DEA: Comparison of approaches for modelling the DEA score, European Journal of Operational Research, 181(1), 425–35 Isik, I., & Hassan, M.K (2002) Technical, scale and allocative efficiencies of the Turkish banking industry, Journal of Banking and Finance, 26(4), 719–66 Kosmidou, K (2008) The determinants of banks’ profits in Greece during the period of EU financial integration, Managerial Finance, 34(3), 146–59 Leightner, J.E., & Lovell, C.A.K (1998) The impact of financial liberalization on the performance of Thai banks, Journal of Economics and Business, 50(2), 115–31 McAllister, P.H., & McManus, D.A (1993) Resolving the scale efficiencies puzzle in banking, Journal of Banking and Finance, 17(2–3), 389–405 McDonald, J (2009) Using least squares and Tobit in second stage DEA efficiency analyses, European Journal of Operational Research, 197(2), 792–98 Miller, S.M., & Noulas, A.G (1996) The technical efficiency of large bank production, Journal of Banking and Finance, 20(3), 495–509 Molyneux, P., Altunbas, Y., & Gardener, E.P.M (1996) Efficiency in European banking Chichester: John Wiley & Sons Downloaded from mar.sagepub.com at OhioLink on March 15, 2015 Sufian and Habibullah DETERMINANTS OF BANK EFFICIENCY 461 Noulas, A.G Ray, S.C., & Miller, S.M (1990) Returns to scale and input substitution for large U.S banks, Journal of Money, Credit and Banking, 22(1), 94–108 Parkan, C (1987) Measuring the efficiency of service operations: An application to bank branches, Engineering Costs and Production Economics, 12(1–4), 237–42 Pasiouras, F (2008) Estimating the technical and scale efficiency of Greek commercial banks: The impact of credit risk, off-balance sheet activities, and international operations, Research in International Business and Finance, 22(3), 301–18 Pasiouras, F., & Kosmidou, K (2007) Factors influencing the profitability of domestic and foreign commercial banks in the European Union, Research in International Business and Finance, 21(2), 222–37 Rangan, N., Grabowski, R., Aly, H.Y., & Pasurka, C (1988) The technical efficiency of U.S banks, Economics Letters, 28(2), 169–75 Saxonhouse, G.R (1976) Estimated parameters as dependent variables, American Economic Review, 66(1), 178–83 Sealey, C., & Lindley, J.T (1977) Inputs, outputs and a theory of production and cost at depository financial institutions, Journal of Finance, 32(4), 1251–66 Sherman, H.D., & Gold, F (1985) Bank branch operating efficiency: Evaluation with data envelopment analysis, Journal of Banking and Finance, 9(2), 297–315 Short, B.K (1979) The relation between commercial bank profit rate and banking concentration in Canada, Western Europe and Japan, Journal of Banking and Finance, 3(3), 209–19 Staikouras, C., & Wood, G (2003) The determinants of bank profitability in Europe, Paper presented at the European Applied Business Research Conference, Venice, 9–13 June Staikouras, C., Mamatzakis, E., & Koutsomanoli-Filippaki, A (2008), An empirical investigation of operating performance in the new European banking landscape, Global Finance Journal, 19(1) 32–45 Stiroh, K.J (2006a), New evidence on the determinants of bank risk, Journal of Financial Services Research, 30(3), 237–63 ——— (2006b) A portfolio view of banking with interest and non-interest activities, Journal of Money, Credit and Banking, 38(5), 1351–61 Stiroh, K.J., & Rumble, A (2006) The dark side of diversification: The case of US financial holding companies, Journal of Banking and Finance, 30(8), 2131–61 Sufian, F (2007) Mergers and acquisitions in the Malaysian banking industry: Technical and scale efficiency effects, International Journal of Financial Services Management, 2(4), 304–26 ——— (2009) Determinants of bank efficiency during unstable macroeconomic environment: Empirical evidence from Malaysia, Research in International Business and Finance, 23(1), 54–77 White, H.J (1980) A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity, Econometrica, 48(4), 817–38 Williams, J., & Intarachote, T (2003) Financial liberalisation and profit efficiency in the Thai banking system, 1990–1997: The case of domestic and foreign banks, Working Paper, University of Wales, Bangor Yue, P (1990) Data envelopment analysis and commercial bank performance: A primer with applications to Missouri banks, Federal Reserve Bank of St Louis Review, 74(January), 31–45 Margin—The Journal of Applied Economic Research : (2010): 427–461 Downloaded from mar.sagepub.com at OhioLink on March 15, 2015