6.benchmarking the

21
Benchmarking the efficiency of the Korean banking sector: a DEA approach Fadzlan Sufian Khazanah Nasional Berhad, Kuala Lumpur, Malaysia and Universiti Putra Malaysia, Serdang, Malaysia Abstract Purpose – The purpose of this paper is to critically examine the sources of inefficiency in the Korean banking sector. The present study focuses on three different approaches: intermediation approach, value-added approach, and operating approach, to differentiate how efficiency scores vary with changes in inputs and outputs. Design/methodology/approach – The paper utilizes the non-parametric data envelopment analysis methodology to measure the efficiency of banks operating in the Korean banking sector. The method allows for the decomposition of technical efficiency (TE) into its mutually exhaustive components of pure technical and scale efficiencies. Findings – The empirical findings suggest that estimates of TE are consistently higher under an operating approach vis-a ` -vis the intermediation and value-added approaches. On the other hand, banks are characterized by a relatively low level of TE under the intermediation approach. Research limitations/implications – Further analysis on the performance of the Korean banking sector performance will examine the efficiency changes over time by employing the parametric stochastic frontier analysis method. Investigations into productivity changes over time, as a result of a technical change or technological progress or regression by employing the Malmquist productivity index could yet be another extension to the paper. Practical implications – The findings from this study are essential not only for the managers of the banks, but for numerous stakeholders such as the central banks, bankers associations, governments, and other financial authorities. Knowledge of these factors would also be helpful to the regulatory authorities and bank managers who formulate going forward policies for improved efficiency of the Korean banking sector. Originality/value – Unlike the previous studies on the efficiency of the Korean banking sector, the paper focuses on three different approaches: intermediation approach, value-added approach, and operating approach to differentiate how efficiency scores vary with changes in inputs and outputs. Keywords Korea, Banking, Organizational performance Paper type Research paper 1. Introduction The banking sector is the backbone of the Korean economy and plays an important financial intermediary role. Therefore, the health of the sector is very critical to the health of the economy at large. Given the relation between the well being of the banking sector and the growth of the economy (Rajan and Zingales, 1998; Levine, 1998; Levine and Zervos, 1998; Cetorelli and Gambera, 2001; Beck and Levine, 2004), knowledge The current issue and full text archive of this journal is available at www.emeraldinsight.com/1463-5771.htm The author would like to thank Angappa Gunasekaran (the Editor) and two anonymous referees for the constructive comments and suggestions, which have significantly improved the contents of the paper. The usual caveats apply. Benchmarking the Korean banking sector 107 Benchmarking: An International Journal Vol. 18 No. 1, 2011 pp. 107-127 q Emerald Group Publishing Limited 1463-5771 DOI 10.1108/14635771111109841

Upload: libfsb

Post on 21-Jan-2015

90 views

Category:

Documents


0 download

DESCRIPTION

 

TRANSCRIPT

Page 1: 6.benchmarking the

Benchmarking the efficiencyof the Korean banking sector:

a DEA approachFadzlan Sufian

Khazanah Nasional Berhad, Kuala Lumpur, Malaysia andUniversiti Putra Malaysia, Serdang, Malaysia

Abstract

Purpose – The purpose of this paper is to critically examine the sources of inefficiency in the Koreanbanking sector. The present study focuses on three different approaches: intermediation approach,value-added approach, and operating approach, to differentiate how efficiency scores vary with changesin inputs and outputs.

Design/methodology/approach – The paper utilizes the non-parametric data envelopment analysismethodology to measure the efficiency of banks operating in the Korean banking sector. The methodallows for the decomposition of technical efficiency (TE) into its mutually exhaustive components ofpure technical and scale efficiencies.

Findings – The empirical findings suggest that estimates of TE are consistently higher underan operating approach vis-a-vis the intermediation and value-added approaches. On the other hand,banks are characterized by a relatively low level of TE under the intermediation approach.

Research limitations/implications – Further analysis on the performance of the Korean bankingsector performance will examine the efficiency changes over time by employing the parametricstochastic frontier analysis method. Investigations into productivity changes over time, as a result of atechnical change or technological progress or regression by employing the Malmquist productivityindex could yet be another extension to the paper.

Practical implications – The findings from this study are essential not only for the managers of thebanks, but for numerous stakeholders such as the central banks, bankers associations, governments,and other financial authorities. Knowledge of these factors would also be helpful to the regulatoryauthorities and bank managers who formulate going forward policies for improved efficiency of theKorean banking sector.

Originality/value – Unlike the previous studies on the efficiency of the Korean banking sector,the paper focuses on three different approaches: intermediation approach, value-added approach,and operating approach to differentiate how efficiency scores vary with changes in inputs and outputs.

Keywords Korea, Banking, Organizational performance

Paper type Research paper

1. IntroductionThe banking sector is the backbone of the Korean economy and plays an importantfinancial intermediary role. Therefore, the health of the sector is very critical to thehealth of the economy at large. Given the relation between the well being of the bankingsector and the growth of the economy (Rajan and Zingales, 1998; Levine, 1998; Levineand Zervos, 1998; Cetorelli and Gambera, 2001; Beck and Levine, 2004), knowledge

The current issue and full text archive of this journal is available at

www.emeraldinsight.com/1463-5771.htm

The author would like to thank Angappa Gunasekaran (the Editor) and two anonymous refereesfor the constructive comments and suggestions, which have significantly improved the contentsof the paper. The usual caveats apply.

Benchmarkingthe Korean

banking sector

107

Benchmarking: An InternationalJournal

Vol. 18 No. 1, 2011pp. 107-127

q Emerald Group Publishing Limited1463-5771

DOI 10.1108/14635771111109841

Page 2: 6.benchmarking the

of the underlying factors that influence the banking sector’s efficiency is thereforeessential, not only for the managers of the banks, but for numerous stakeholders, such asthe central banks, bankers associations, governments, and other financial authorities.Knowledge of these factors would also be helpful to help the regulatory authoritiesand bank managers formulate going forward policies for improved efficiency of theKorean banking sector.

The purpose of the present study is to extend the earlier works on the performanceof the Korean banking sector and examine the efficiency of the Korean bankingsector during the period of 1992-2003. The efficiency estimate of each bank is computedby using the data envelopment analysis (DEA) method. The method allows us todistinguish between three different types of efficiency measures, namely technical,pure technical, and scale. The DEA method has many advantages over traditionalparametric techniques such as regression techniques. While regression analysis focuseson central tendency and approximates the efficiency of banks under investigationrelative to the average performance, DEA in contrast, focuses on the yearlyobservations of individual banks and optimizes the performance measure of each bank.We differentiate this paper from previous ones that focus on the Korean banking sectorand contribute to the present literature in several respects discussed below.

First, the previous studies that examined the efficiency of the Korean banking sectoradopt the intermediation approach. On the other hand, the present study compare theresults obtained from the intermediation approach with value-added or revenueapproach that was recently proposed by Drake et al. (2006) and operating approach.This allows us to observe if different input and output definitions affect the obtainedefficiency levels.

Second, unlike the previous studies that examined the efficiency of the Koreanbanking sector, the present study constructs and analyzes the results derived fromdynamic panels. Constructing a separate frontier for each of the years under study is acritical issue in a dynamic business environment because a bank may be the mostefficient in one year, but may not be in the following year. Within the context of theKorean banking sector, it becomes more crucial, as there is an ongoing liberalization inthe banking sector over the estimation period. A dynamic panel analysis may alsohighlight any significant changes taking place in the Korean banking sector during theperiod under study induced by the Bank of Korea (BOK) supervisory policies.

Finally, the present study attempts to critically analyze the returns to scale in theKorean banking sector, which has not been critically examined in the previous research.Given the initiatives by the BOK to strengthen the banking sector, which among othersinvolved mergers and acquisitions of the domestic banking institutions, the precisenature of scale efficiency (SE) is critically important to both comprehend the economicrationale and to prescribe their going forward policies. Furthermore, it will be ofpractical interest to estimate and compare the returns to scale of the nationwidecommercial banks[1] compared to their regional commercial bank[2] peers given thatthe size of the former are much larger than that of the latter. Furthermore, the regionalcommercial banks only maintain branch banking within their 0 localities and mainlyserve small- and medium-sized enterprises, households, and individual borrowers intheir respective regions. On the other hand, the nationwide commercial banks areallowed to maintain branch banking systems throughout the country and engage in bothshort- and long-term financing.

BIJ18,1

108

Page 3: 6.benchmarking the

This paper is set out as follows: in the next section, we provide reviews of the mainliterature. In Section 3, we outline the approaches to the measurement of efficiencychange. Section 4 discusses the results, and finally we conclude in Section 5.

2. Review of the literaturesSince its introduction by Charnes et al. (1978) and Banker et al. (1984), researchers havewelcomed DEA as a methodology for performance evaluation (Gregoriou and Zhu,2005). However, a large body of literature exists on banking efficiency in the USA(see surveys in Berger et al., 1993; Berger and Humphrey, 1997; Berger, 2007, andreferences therein) and the banking systems in the Western and developed countries(Sathye, 2001; Drake, 2001; Canhoto and Dermine, 2003; Webb, 2003; Fiordelisi, 2007;Pasiouras, 2008; Sturm and Williams, 2008; Siriopoulos and Tziogkidis, 2010).

Fukuyama (1993) who considers the efficiency of 143 Japanese banks in 1990 isamong the earliest to employ frontier estimation technique to examine the performanceof Asian banks. The results suggest that banks of different organizational statusperform differently in respect to all efficiency measures and that SE is positively,but weakly associated with bank size.

Single country studies focusing on the Asian banking sectors have mainlyconcentrated on the comparison between the foreign and domestic banks’ performance.Generally, the empirical evidence showed that foreign banks have succeeded incapitalizing on their advantages and exhibit a higher level of efficiency than theirdomestic bank peers. Leightner and Lovell (1998) find that the average domesticThai banks experienced falling total factor productivity (TFP) growth, while theaverage foreign bank experienced increasing TFP. Unite and Sullivan (2003) suggestthat the entry of foreign banks in the Philippines has resulted in the reduction of interestrate spreads and bank profits of the domestic banks that are affiliated with familybusiness groups. In a study on the Malaysian banking sector, Matthews and Ismail(2006) suggest that foreign banks in Malaysia have exhibited a higher level of technicalefficiency (TE). They also suggest that the productivity of the domestic banks is moresusceptible to macroeconomic shocks than their foreign bank counterparts.

The South Asian banking sectors have also been studied extensively. Sathye (2003)and Shanmugam and Das (2004) find that the public and foreign-owned banks in Indiahave exhibited a higher level of TE compared to their privately owned bank peers.Iimi (2004) suggests that privatized banks in Pakistan are the most efficient, followed bythe foreign-owned banks, while the public banks are the least efficient. Hardy and di Patti(2001) investigate the effects of financial reforms on profitability, cost, and revenueefficiency of the Pakistan banking sector during 1981-1998. They show that financialliberalization has positive impact on banks’ performance. Subsequently, di Patti andHardy (2005) examine the cost and profit efficiency of Pakistan commercial banksduring the period 1981-2002. They find that financial liberalization leads to higher bankprofitability, but only during the first round of financial reform of 1991-1992.

Despite substantial studies performed in regard to the efficiency of financial institutionsin the USA, Europe, and other Asia-Pacific banking industries, empirical evidence on theKorean banking sector is relatively scarce. By employing the Malmquist productivityindex (MPI) approach, Gilbert and Wilson (1998) focus on the impact of deregulation inthe 1980s on the productive efficiency of Korean private banks during the post-1980s topre-crisis years, showing that deregulation did improve bank efficiency. Hao et al. (2001)

Benchmarkingthe Korean

banking sector

109

Page 4: 6.benchmarking the

employ the stochastic cost frontier approach and focus on the pre-crisis period of 1985-1995.The results indicate that high-growth institutions with vast geographical networks andbanks funded with retail deposits were the most efficient during that era. Most recently,Park and Weber (2006) examine the efficiency and productivity of the Korean bankingsector during the 1992-2002 period. The empirical findings derived from the directionaltechnology distance function method suggest that the Korean banking sector exhibitsproductivity growth attributed to technical progress, which outweighs declines inefficiency. Other studies like An et al. (2007), Jeon et al. (2006), Choi and Hasan (2005) andChoe and Lee (2003) examine the performance of Korean banking sector by employingcentral tendency regression approach rather than frontier estimation techniques.

3. Methodology and data3.1 Data envelopment analysisThe present study employs the DEA method, first introduced by Charnes et al. (1978)(hereafter the CCR model) to estimate the input-oriented TE of the Korean bankingsector. The DEA method involves constructing a non-parametric production frontierbased on the actual input-output observations in the sample relative to which efficiencyof each bank in the sample is measured (Coelli, 1996). This approach measures theefficiency of a decision-making unit (DMU) relative to other similar DMUs with thesimple restriction that all DMUs lay on, or below the efficiency frontier. If a DMU lies onthe frontier, it is referred to as an efficient unit; otherwise, it is labeled as inefficient.The data are enveloped in such a way that radial distances to the frontier are minimized.

The CCR model can be formulated as follows:

min l0 2 1Xmi¼1

S2i þ

Xs

r¼1

Sþr

" #ð1Þ

subject to :XNf¼1

lf xif ¼ loxif o 2 S2i where i ¼ 1. . .m

XNf¼1

lf yrf ¼ Sþr þ yrf o where r ¼ 1. . .s

lf $ 0; f ¼ 1. . .N ; S2i ;S

þr $ 0 ; i and r

where xif and yrf are levels of the ith input and rth output, respectively, for DMU f.N is the number of DMUs. 1 is a very small positive number (non-Archimedean) usedas a lower bound to inputs and outputs. lf denotes the contribution of DMU f inderiving the efficiency of the rated DMU fo (a point at the envelopment surface). S2

i andSþr are slack variables to proxy extra savings in input i and extra gains in output r. lo is

the radial efficiency factor that shows the possible reduction of inputs for DMU fo. If l*o(optimal solution) is equal to one and the slack values are both equal to zero, then DMUfo is said to be efficient. When S2

i or Sþr take positive values at the optimal solution,

one can conclude that the corresponding input or output of DMU fo can improve furtheronce input levels have been contracted to the proportion l*o .

The CCR model presupposes that there is no significant relationship betweenthe scale of operations and efficiency by assuming constant returns to scale (CRS) and

BIJ18,1

110

Page 5: 6.benchmarking the

it delivers the overall TE. The CRS assumption is only justifiable when all DMUs areoperating at an optimal scale. However, banks in practice may face either economies ordiseconomies of scale. Thus, if one makes the CRS assumption when not all DMUs areoperating at the optimal scale, the computed measures of TE will be contaminatedwith SE.

Banker et al. (1984) extended the CCR model by relaxing the CRS assumption. Theresulting Banker, Charnes, and Cooper (BCC) model is used to assess the efficiency ofDMUs characterized by variable returns to scale (VRS). The VRS assumption providesthe measurement of pure technical efficiency (PTE), which is the measurement ofTE devoid of the SE effects. If there appears to be a difference between the TE andPTE scores of a particular DMU, then it indicates the existence of scale inefficiency,i.e. TE ¼ PTE £ SE. The former relates to the capability of managers to utilize banks’given resources, whereas the latter refers to exploiting scale economies by operating ata point where the production frontier exhibits CRS.

The input-oriented BCC model with VRS assumption can be represented by thefollowing linear programming problem:

min lo 2 1Xmi¼1

S2i þ

Xs

r¼1

Sþr

" #ð2Þ

subject to :XNf¼1

lf xif ¼ loxif o 2 S2i where i ¼ 1. . .m

XNf¼1

lf yrf ¼ Sþr þ yrf o where r ¼ 1. . .s

XNf¼1

lf ¼ 1

lf $ 0; f ¼ 1. . .N ; S2i ; S

þr $ 0 ; i and r

The BCC model differs from the CCR model in that it includes the so-called convexity

constraint,PN

f¼1lf ¼ 1, which prevents any interpolation point constructed from the

observed DMUs from being scaled up or down to form a referent point. In this model, theset ofl values minimize lo to l*o and identify a point within the VRS assumption, of whichthe input levels reflect the lowest proportion of l*o . At l*o , the input levels of DMU fo can beuniformly contracted without detriment its output levels. Therefore, DMU fo hasefficiency equal to l*o . The solution to model (2) is summarized in the following fashion:

DMU fo is pareto-efficient if l*o ¼ 1 and Sþ*

r ¼ 0; r ¼ 1. . .s; S2*

i ¼ 0; i ¼ 1. . .m.If the convexity constraint in model (2) is dropped, one obtains model (1), which

generates TE under the CRS assumption. This implies that PTE of a DMU is alwaysgreater or equal to its TE. Under the VRS assumption, the resulting SE can be measured,since in most cases, the scale of operation of the firm may not be optimal. The firminvolved may be too small in its scale of operation, which might fall within the increasing

Benchmarkingthe Korean

banking sector

111

Page 6: 6.benchmarking the

returns to scale (IRS) part of the production function. Similarly, a firm may be toolarge and operate within the decreasing returns to scale (DRS) part of the productionfunction. In both cases, efficiency of the firms may be improved by changing theirscale of operation. If the underlying production technology follows CRS, then the firmis automatically scale efficient. The resulting ratio illustrates SE, which is the impactof scale size on the efficiency of a DMU. Formally, the SE of DMU fo is given as TE/PTE.Where, TE and PTE are technical efficiency and pure technical efficiency of DMU fo,respectively.

Since PTE is always greater or equal to TE, it means that SE (TE/PTE) is less or equalto unity. If TE and PTE of a DMU are equal, then SE is equal to one. This means thatirrespective of scale, size has no impact on efficiency. If the TE scores derived from theCRS assumption is less than the TE scores derived from the VRS assumption, thenSE will be below unity, meaning that the scale of operation does impact the efficiencyof the DMU.

3.2 Specification of bank inputs, outputs, and dataIt is commonly acknowledged that the choice of variables in efficiency studiessignificantly affects the results. The problem is compounded by the fact that variableselection is often constrained by the paucity of data on relevant variables. The cost andoutput measurements in banking are especially difficult because many of the financialservices are jointly produced and prices are typically assigned to a bundle of financialservices. The role of the commercial banks is generally defined as collecting the savings ofhouseholds and other agents to finance the investment needs of firms and consumptionneeds of individuals. Four main approaches dominate the literature: the productionapproach, the intermediation approach, the operating approach, and more recently,the revenue or profit-oriented approach. The first two approaches apply the traditionalmicroeconomic theory of the firm to banking and differ only in the specification of bankingactivities. The final two approaches go a step further and incorporate some specificactivities of banking into the classical theory and thereby modify it.

Under the production approach, pioneered by Benston (1965), a financial institution isdefined as a producer of services for account holders, that is, they perform transactionson deposit accounts and process documents such as loans. Hence, according to thisapproach, the number of accounts or its related transactions is the best measure foroutput, while the number of employees and physical capital are considered as inputs.However, the production approach might be more suitable for branch efficiency studies,as at most times bank branches basically process customer documents and bankfunding, while investment decisions are mostly not under the control of branches(Berger and Humphrey, 1997).

The intermediation approach on the other hand, assumes that financial firms act as anintermediary between savers and borrowers and posits total loans and securitiesas outputs, whereas deposits along with labour and physical capital are defined asinputs (Sealey and Lindley, 1977). The operating approach (or income-based approach)views banks as business units with the final objective of generating revenue from thetotal cost incurred for running the business (Leightner and Lovell, 1998). Accordingly,it defines banks’ output as total revenue (interest and non-interest income) and inputs asthe total expenses (interest and non-interest expenses). More recently, Drake et al. (2006)proposed the revenue approach in DEA. The revenue approach (or income-based

BIJ18,1

112

Page 7: 6.benchmarking the

approach) views banks as business units with the final objective of generating revenuefrom the total cost incurred for running the business (Leightner and Lovell, 1998).Accordingly, it defines banks’ output as total revenue (interest and non-interest income)and inputs as the total expenses (interest and non-interest expenses).

The appropriateness of each approach varies according to the circumstances.It is apparent that banks undertake simultaneous functions. However, based on practicalconsiderations and to examine the robustness of the estimated efficiency scores undervarious alternatives, the present study focuses on three major approaches: intermediationapproach, operating approach, and value-added approach. Under the intermediationapproach, we assume deposits (X1), labour (X4), and capital (X2) as inputs for producingloans (Y1) and investments (Y2). Under the operating approach, two types of inputs areconsidered namely, interest expenses (X3) and labour (X4). The relevant outputs areinterest income (Y4) and non-interest income (Y5) emanating mostly from commission,exchange, brokerage, etc. Under the value-added approach, labour (X4), capital (X2),and interest expenses (X3) are used as inputs producing outputs like deposits (X1), loans(Y1), and investments (Y2).

We use the annual bank level data of Korean commercial banks over the period1992-2003. The variables are obtained from published balance sheet information inannual reports of each individual bank. The final sample consists of 31 banks, whichaccount for more than 80 percent of the Korean banking sector’s total assets. Table Ipresents summary statistics of the output and input variables used to construct theDEA model.

4. Empirical findingsIn this section, we will discuss the TE change of the Korean banking sector, measured bythe DEA method and its decomposition into PTE and SE components. The efficiency ofthe Korean banking sector is first examined by applying the DEA method for each yearunder investigation. To allow efficiency to vary over time, the efficiency frontiers areconstructed for each year by solving the linear programming (LP) problems rather thanconstructing a single multi-year frontier.

Loans(Y1)

Investments(Y2)

Interestincome

(Y3)

Non-interestincome

(Y4)

Totaldeposits

(X1)Capital

(X2)

Interestexpenses

(X3)

Non-interestexpense

(X4)

Min. 304.65 82.36 16.15 11.74 82.00 0.27 11.09 10.50Mean 36,336.16 4,793.75 2,931.93 622.26 37,596.46 552.26 1,485.15 873.72Max. 126,005.20 29,846.61 9,616.51 2,424.91 131,068.00 1,528.13 6,148.47 2,938.84SD 28,568.93 5,320.97 2,211.16 521.67 29,328.75 402.07 1,148.98 634.82

Notes: Y1 – loans (includes loans to customers and other banks); Y2 – investments (includes dealingand investment securities); Y3 – interest income; Y4 – non-interest income (defined as fee income andother non-interest income, which among others consist of commission, service charges and fees,guarantee fees, and foreign exchange profits); X1 – total deposits (includes deposits from customers andother banks); X2 – capital (measured by the book value of property, plant, and equipment); X3 – interestexpenses; X4 – non-interest expenseSource: Banks annual reports and authors own calculations

Table I.Descriptive statistics

for inputs and outputs

Benchmarkingthe Korean

banking sector

113

Page 8: 6.benchmarking the

4.1 Efficiency of the Korean banking sectorThe summary results of technical, pure technical, and SE estimates under the threeapproaches are presented in Tables II-IV, respectively. The average TE estimate (M)represents the average of all optimal values obtained from the CCR model for eachbank (Table II). The empirical results suggest a large asymmetry between banksregarding their TE scores. In particular, the different approaches of classifying inputsand outputs of banks produced divergent sets of efficiency estimates. The estimates ofTE were observed to be consistently higher under operating approach vis-a-vis the

IntervalYear

No. ofbanks

No. ofefficientbanks

Averageefficiency

(M)SD(s) (I ¼ M2s) (I ¼ M þ s)

Percentageof banks in

I

Percentageof banks I SDbelow mean

Intermediation approach1992 11 2 0.736 0.218 0.518 0.954 72.73 9.091993 14 2 0.787 0.154 0.633 0.941 71.43 7.141994 15 3 0.719 0.183 0.536 0.902 60.00 20.001995 16 3 0.867 0.100 0.767 0.967 68.75 12.501996 18 4 0.845 0.131 0.714 0.976 55.56 11.111997 23 2 0.622 0.158 0.464 0.780 82.61 4.351998 25 3 0.606 0.193 0.413 0.799 68.00 16.001999 24 3 0.714 0.177 0.537 0.891 66.67 16.672000 29 4 0.719 0.169 0.550 0.888 65.52 13.792001 26 3 0.741 0.155 0.586 0.896 61.54 19.232002 23 3 0.653 0.175 0.478 0.828 69.57 13.042003 23 3 0.678 0.166 0.512 0.844 73.91 8.70Value-added approach1992 11 5 0.942 0.078 0.864 1.020 81.82 18.181993 14 5 0.897 0.111 0.786 1.008 78.57 21.431994 16 6 0.893 0.107 0.786 1.000 75.00 25.001995 17 5 0.866 0.119 0.747 0.985 52.94 17.651996 19 7 0.902 0.111 0.791 1.013 84.21 15.791997 24 5 0.882 0.097 0.785 0.979 62.50 16.671998 25 5 0.682 0.205 0.477 0.887 68.00 8.001999 24 10 0.908 0.125 0.783 1.033 91.67 8.332000 29 2 0.546 0.159 0.387 0.705 82.76 6.902001 26 2 0.750 0.152 0.598 0.902 76.92 11.542002 23 1 0.727 0.163 0.564 0.890 73.91 17.392003 23 7 0.890 0.130 0.760 1.020 82.61 17.39Operating approach1992 11 6 0.972 0.035 0.937 1.007 63.64 36.361993 14 7 0.962 0.052 0.910 1.014 78.57 21.431994 16 8 0.935 0.070 0.865 1.005 68.75 31.251995 17 6 0.965 0.033 0.932 0.998 47.06 17.651996 19 5 0.932 0.090 0.842 1.022 89.47 10.531997 24 6 0.926 0.068 0.858 0.994 62.50 12.501998 25 9 0.871 0.124 0.747 0.995 44.00 20.001999 24 6 0.913 0.091 0.822 1.004 75.00 25.002000 29 5 0.878 0.119 0.759 0.997 68.97 13.792001 26 6 0.846 0.199 0.647 1.045 80.77 19.232002 24 5 0.858 0.118 0.740 0.976 62.50 16.672003 23 5 0.833 0.133 0.700 0.966 60.87 17.39

Table II.Average TEof Korean banks

BIJ18,1

114

Page 9: 6.benchmarking the

intermediation and value-added approaches. On the other hand, under the intermediationapproach, banks are characterized by relatively low level of TE. Illustratively, in the year1999 only three (12.5 percent) banks were found to be efficient and the average TE for allbanks stood at 71.4 percent under the intermediation approach. The number of efficientbanks during the sample period ranged from two banks in 1992 and 1993 to four banks in1996 under the intermediation approach and six banks in 1992 to nine banks in 1999under the operating approach. On the other hand, the number of efficiency banks rangedfrom a high of ten banks in 1999 to a low of one banks in 2002 under the value-added

Interval

YearNo. ofbanks

No. ofefficientbanks

Averageefficiency

(M)SD(s) (I ¼ M2s) (I ¼ M þ s)

Percentageof banks in

I

Percentageof banks I SDbelow mean

Intermediation approach1992 11 5 0.868 0.208 0.660 1.076 90.91 9.091993 14 4 0.852 0.150 0.702 1.002 92.86 7.141994 15 6 0.811 0.187 0.624 0.998 46.67 13.331995 16 8 0.921 0.088 0.833 1.009 75.00 25.001996 18 9 0.888 0.127 0.761 1.015 83.33 16.671997 23 7 0.727 0.204 0.523 0.931 56.52 13.041998 25 7 0.678 0.243 0.435 0.921 52.00 16.001999 25 7 0.785 0.188 0.597 0.973 44.00 24.002000 29 8 0.817 0.160 0.657 0.977 55.17 17.242001 26 6 0.797 0.159 0.638 0.956 57.69 15.382002 23 6 0.727 0.186 0.541 0.913 56.52 17.392003 23 6 0.727 0.182 0.545 0.909 65.22 8.70Value-added approach1992 11 7 0.955 0.078 0.877 1.033 81.82 18.181993 14 8 0.922 0.106 0.816 1.028 85.71 14.291994 16 8 0.921 0.101 0.820 1.022 81.25 18.751995 17 11 0.917 0.123 0.794 1.040 76.47 23.531996 19 12 0.947 0.085 0.862 1.032 78.95 21.051997 24 12 0.932 0.097 0.835 1.029 75.00 25.001998 25 7 0.712 0.211 0.501 0.923 60.00 8.001999 24 14 0.943 0.114 0.829 1.057 87.50 12.502000 29 9 0.798 0.197 0.601 0.995 58.62 10.342001 26 4 0.796 0.159 0.637 0.955 69.23 11.542002 23 4 0.773 0.163 0.610 0.936 65.22 17.392003 23 10 0.909 0.125 0.784 1.034 78.26 21.74Operating approach1992 11 8 0.989 0.022 0.967 1.011 81.82 18.181993 14 12 0.998 0.005 0.993 1.003 92.86 7.141994 16 11 0.981 0.033 0.948 1.014 81.25 18.751995 17 13 0.991 0.020 0.971 1.011 88.24 11.761996 19 13 0.989 0.024 0.965 1.013 89.47 10.531997 24 16 0.974 0.046 0.928 1.020 87.50 12.501998 25 12 0.925 0.104 0.821 1.029 80.00 20.001999 24 16 0.962 0.073 0.889 1.035 87.50 12.502000 29 11 0.937 0.102 0.835 1.039 89.66 10.342001 26 11 0.887 0.170 0.717 1.057 80.77 19.232002 24 11 0.928 0.097 0.831 1.025 83.33 16.672003 23 10 0.917 0.106 0.811 1.023 78.26 21.74

Table III.Average PTE

of Korean banks

Benchmarkingthe Korean

banking sector

115

Page 10: 6.benchmarking the

approach. In sum, the empirical findings seem to suggest that there was no perceptiblechange in number of efficient banks under the intermediation and operating approaches,although a noticeable increase was observed under the value-added approach duringthe year 1999.

Under the operating approach, the dispersion of TE scores as measured by itsstandard deviation depicts an increasing trend during the years 1996-1999. On the otherhand, the percentage of banks wherein TE lies within the interval of one standarddeviation around the mean hovered around 89.5 percent in 1996 to 75 percent in 1999

Interval

YearNo. ofbanks

No. ofefficientbanks

Averageefficiency

(M)SD(s) (I ¼ M2s) (I ¼ M þ s)

Percentageof banks in

I

Percentageof banks I SDbelow mean

Intermediation approach1992 11 2 0.857 0.161 0.696 1.018 81.82 18.181993 14 2 0.928 0.099 0.829 1.027 85.71 14.291994 15 3 0.897 0.148 0.749 1.045 86.67 13.331995 16 3 0.944 0.092 0.852 1.036 93.75 6.251996 18 4 0.952 0.066 0.886 1.018 83.33 16.671997 23 3 0.875 0.141 0.734 1.016 78.26 21.741998 25 3 0.919 0.121 0.798 1.040 84.00 16.001999 25 4 0.915 0.110 0.805 1.025 88.00 8.002000 29 4 0.879 0.087 0.792 0.966 75.86 3.452001 26 4 0.935 0.106 0.829 1.041 92.31 7.692002 23 3 0.909 0.123 0.786 1.032 86.96 13.042003 23 3 0.944 0.120 0.824 1.064 91.30 8.70Value-added approach1992 11 5 0.986 0.019 0.967 1.005 72.73 27.271993 14 6 0.973 0.040 0.933 1.013 78.57 21.431994 16 11 0.970 0.059 0.911 1.029 81.25 18.751995 17 5 0.946 0.066 0.880 1.012 88.24 11.761996 19 7 0.954 0.090 0.864 1.044 89.47 10.531997 24 5 0.948 0.052 0.896 1.000 83.33 16.671998 25 5 0.959 0.056 0.903 1.015 88.00 12.001999 24 11 0.963 0.055 0.908 1.018 87.50 12.502000 29 2 0.703 0.173 0.530 0.876 65.52 10.342001 26 3 0.944 0.077 0.867 1.021 92.31 7.692002 23 4 0.939 0.076 0.863 1.015 82.61 17.392003 23 8 0.978 0.041 0.937 1.019 91.30 8.70Operating approach1992 11 6 0.983 0.027 0.956 1.010 81.82 18.181993 14 7 0.963 0.049 0.914 1.012 78.57 21.431994 16 8 0.953 0.062 0.891 1.015 81.25 18.751995 17 6 0.973 0.027 0.946 1.000 82.35 17.651996 19 5 0.943 0.086 0.857 1.029 94.74 5.261997 24 6 0.951 0.048 0.903 0.999 54.17 20.831998 25 9 0.941 0.063 0.878 1.004 88.00 12.001999 24 10 0.951 0.082 0.869 1.033 87.50 12.502000 29 5 0.939 0.093 0.846 1.032 89.66 10.342001 26 8 0.950 0.105 0.845 1.055 88.46 11.542002 24 5 0.925 0.080 0.845 1.005 79.17 20.832003 23 6 0.912 0.117 0.795 1.029 82.61 17.39

Table IV.Average SEof Korean banks

BIJ18,1

116

Page 11: 6.benchmarking the

under the operating approach and 84.2 percent in 1996 to 91.7 percent in 1999 underthe value-added approach. These numbers were slightly lower under the intermediationapproach. As the TE estimates itself is time varying, these proportions do not necessarilycorroborate the degree of efficiency of the banking sector. For example, under theintermediation approach, around 82.6 percent in 1997 and around 61.5 percent in 2001 ofbanks recorded TE within the interval of one standard deviation around the mean. Yet,banks were much more efficient in 2001 than in 1997.

As against the changing benchmark of comparison, these proportions quantify thenumber of banks that are close to the average over time and thus merely capture thekurtosis of the efficiency distribution depending on the approach. For instance, underthe intermediation approach the efficiency scores displays a leptokurtic distribution,i.e. the efficiency scores has a high peak with a small variance, suggesting that a lotof scores fall in the center of the distribution. On the other hand, under the operatingapproach the efficiency of the Korean banking sector seem to follow a mesokurticdistribution, i.e. the efficiency scores display a moderate peak with gradual curvessuggesting a normal number of scores in the middle range of the distribution.

Overall, the empirical findings presented in Table II clearly bring forth low degree ofefficiency in the Korean banking sector, particularly a year after the Asian financialcrisis stemming from the under utilization of resources (inputs). Finally, consideringthe evolution of efficiency over time, a clear temporal pattern does not emerge fromthese different approaches. It is also worth noting from Table II that although in generalTE level seems to deteriorate a year after the Asian financial crisis under all approaches,the deterioration seems to be more pronounced under the intermediation approachmodel.

Table III presents the PTE estimates, while SE estimates under the three approachesare presented in Table IV. It is observed that over the sample period, both pure technicaland SE measures, especially under intermediation and value-added approaches displaysignificant variation and the Korean banking sector did not achieve sustained efficiencygains. Estimates of PTE under the intermediation approach vary from a high of92.1 percent in 1995 to a low of 67.8 percent in 1998. In most of the years, only about20-30 percent of banks were found to be pure technically efficient under theintermediation approach. Interestingly, the percentage of banks whose PTE falls withinthe interval of one standard deviation around the mean displayed a large asymmetry,particularly during the period 1997-1999 under the intermediation and value-addedapproaches. It is observed from Table III that under the intermediation approachthe percentage stood at around 56.5 percent in 1997 to 44 percent in 1999, while undervalue-added approach the figures stood at around 75 percent in 1997 to 87.5 percentin 1999.

It is interesting to note that the number of efficient banks under CRS (TE) assumptionand VRS (PTE) assumption differs markedly, irrespective of the choice of various inputsand outputs. The findings clearly demonstrate low degree of SE among Koreancommercial banks. Under the intermediation approach for example, Table III revealsthat eight banks were found to be efficient under VRS in 2000, whereas only four bankswere found to be efficient under CRS. This indicates that the remaining four banks failedto reach the CRS frontier owing solely to scale.

It is observed from Tables III and IV that under the operating approach, SE seems tooutweigh PTE in determining the total TE of the Korean banking sector. On the other

Benchmarkingthe Korean

banking sector

117

Page 12: 6.benchmarking the

hand, under the intermediation approach, the empirical findings seem to suggestthat PTE outweighs SE in determining the total TE of the Korean banking sector. Finally,under the operating approach, although SE is generally lower during the pre-Asianfinancial crisis, the trend is less clear during the post-Asian financial crisis period.

4.2 Composition of the efficiency frontiersThe results in the preceding analysis highlight the sources of TE of the Korean bankingsector. Since the dominant source of total TE in the Korean banking sector seems to bescale related, it is worth to examine further the trend in the returns to scale of the Koreanbanks. As pointed out in the previous section, a bank can operate at CRS or VRS whereCRS signifies that an increase in inputs results in a proportionate increase in outputs andVRS means a rise in inputs results in a disproportionate rise in outputs. Further, a bankoperating at VRS can be at IRS or DRS. IRS means that an increase in inputs results in ahigher increase in outputs, while DRS indicate that an increase in inputs results in lesseroutput increases.

To identify the nature of returns to scale, first the CRS scores (obtained with the CCRmodel) is compared with VRS (derived from the BCC model) scores. For a given bank, ifthe VRS score equals to its CRS score, the bank is said to be operating at CRS. On theother hand, if the scores are not equal, a further step is needed to establish whetherthe bank is operating at IRS or DRS. To do this, the DEA model is used under thenon-increasing returns to scale (NIRS) assumption[3]. If the score under VRS equals theNIRS score, then the bank is said to be operating at DRS. Alternatively, if the score underVRS is different from the NIRS score, then the bank is said to be operating at IRS(Coelli et al., 1998).

Table V shows the composition of banks that lie on the efficiency frontier under theintermediation approach. The composition of the efficiency frontier suggests that thenumber of banks that span the frontier varies between two and four banks underthe intermediation approach. It is observed from Table V that Korean French Bank andHousing and Commercial Bank appeared to be the global leaders, i.e. have appeared themost times on the efficiency frontier under the intermediation approach. Under theintermediation approach, the empirical findings seem to suggest that 15 (48.4 percent)banks have managed to appear on the frontier, while 16 other banks have never made itto the efficiency frontier throughout the period of study.

On the other hand, the number of banks which span the efficiency frontier under thevalue-added approach displays large variations ranging from one bank in the year 2002to ten banks in the year 1999. It is observed from Table VI that three banks namely,Housing and Commercial Bank, Cheju Bank, and Woori Bank have appeared the mosttimes on the efficiency frontier. Unlike the intermediation approach, the empiricalfindings seem to suggest that only three (9.68 percent) banks have failed to appear on theefficiency frontier under the value-added approach.

Under the operating approach, the composition of the efficiency frontier is relativelymore stable with small variations. It is observed from Table VII that the number ofbanks which span the efficiency frontier ranged from five to nine banks. The empiricalfindings seem to suggest that 22 (70.97 percent) banks have managed to reach theefficiency frontier. Interestingly, we find that six banks have managed to appear as theglobal leader banks under the operating approach, compared to two banks underthe intermediation approach and three banks under the value-added approach.

BIJ18,1

118

Page 13: 6.benchmarking the

Ban

k19

9219

9319

9419

9519

9619

9719

9819

9920

0020

0120

0220

03C

oun

tb

ank

Bu

san

Mu

tual

Sav

ing

sB

ank

DR

SIR

SIR

SIR

S0

Ch

eju

Ban

kIR

SIR

SIR

SD

RS

IRS

IRS

IRS

DR

SIR

SIR

SIR

S0

Ch

ohu

ng

Ban

kD

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

DR

S0

Cit

iban

kK

orea

IRS

IRS

IRS

IRS

DR

SIR

SIR

SD

RS

DR

SC

RS

CR

SC

RS

3D

aeg

uB

ank

IRS

IRS

CR

SC

RS

CR

SC

RS

IRS

DR

SD

RS

DR

SIR

SIR

S4

Deb

ecM

utu

alS

avin

gs

Ban

kIR

SIR

SIR

SIR

SIR

SIR

S0

Eu

tteu

mM

utu

alS

avin

gs

Ban

kC

RS

IRS

IRS

1H

ana

Ban

kIR

SIR

SIR

SIR

SD

RS

IRS

DR

SD

RS

DR

SD

RS

DR

SD

RS

0H

anm

aum

Mu

tual

Sav

ing

sB

ank

DR

SIR

SIR

SIR

S0

Han

sol

Mu

tual

Sav

ing

sB

ank

DR

SIR

SIR

SIR

S0

Hou

sin

gan

dC

omm

erci

alB

ank

CR

SC

RS

CR

SC

RS

DR

SN

AD

RS

DR

SC

RS

5In

du

stri

alB

ank

ofK

orea

CR

SC

RS

DR

SD

RS

DR

SC

RS

CR

S4

Jeil

Mu

tual

Sav

ing

sB

ank

IRS

IRS

IRS

DR

SIR

SIR

SIR

S0

Jeon

bu

kB

ank

IRS

IRS

IRS

IRS

DR

SIR

SIR

SIR

SD

RS

IRS

IRS

DR

S0

Jin

heu

ng

Mu

tual

Sav

ing

sB

ank

CR

SIR

SIR

SIR

SIR

SIR

SIR

SIR

S1

Kan

gw

onB

ank

IRS

IRS

IRS

IRS

DR

SIR

SIR

S0

Koo

km

inB

ank

DR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SC

RS

DR

SD

RS

1K

orea

Ex

chan

ge

Ban

kC

RS

CR

SIR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

2K

orea

Fir

stB

ank

DR

SD

RS

CR

SD

RS

DR

SN

AD

RS

DR

SD

RS

DR

SD

RS

DR

S1

Kor

eaM

utu

alS

avin

gs

Ban

kIR

SIR

SIR

SD

RS

IRS

IRS

IRS

0K

orea

nF

ren

chB

ank

ing

Cor

p.

CR

SC

RS

CR

SC

RS

CR

SC

RS

6K

wan

gju

Ban

kIR

SIR

SIR

SD

RS

IRS

IRS

NA

DR

SD

RS

0K

yon

gn

amB

ank

CR

SD

RS

IRS

IRS

DR

SD

RS

DR

S1

Pea

ceB

ank

ofK

orea

Ban

kIR

SIR

SC

RS

1P

ure

un

Mu

tual

Sav

ing

sB

ank

DR

SIR

SIR

SIR

S0

Pu

san

Ban

kIR

SIR

SIR

SIR

SD

RS

IRS

IRS

NA

DR

SN

AIR

SIR

S0

Seo

ul

Mu

tual

Sav

ing

sB

ank

CR

SIR

SIR

SIR

SD

RS

IRS

IRS

IRS

1S

hin

han

Ban

kIR

SIR

SIR

SC

RS

DR

SC

RS

CR

SD

RS

IRS

3S

hin

min

Mu

tual

Sav

ing

sB

ank

CR

SIR

SIR

S1

Sol

omon

Mu

tual

Sav

ing

sB

ank

IRS

IRS

IRS

IRS

IRS

IRS

0W

oori

Ban

kIR

SIR

SIR

SIR

SD

RS

DR

SD

RS

DR

SD

RS

0C

oun

ty

ear

22

33

42

33

43

33

35

Notes:

Th

eta

ble

show

sth

eev

olu

tion

ofre

turn

sto

scal

ein

the

Kor

ean

ban

kin

gse

ctor

du

rin

gth

ep

erio

d19

92-2

003;

CR

S,D

RS

,an

dIR

Sd

enot

eco

nst

ant

retu

rns

tosc

ale,

dec

reas

ing

retu

rns

tosc

ale,

and

incr

easi

ng

retu

rns

tosc

ale,

resp

ecti

vel

y;

“cou

nt

yea

r”d

enot

esth

en

um

ber

ofb

ank

sap

pea

rin

gon

the

effi

cien

cyfr

onti

erd

uri

ng

the

yea

r;“c

oun

tb

ank

”d

enot

esth

en

um

ber

ofti

mes

ab

ank

has

app

eare

don

the

effi

cien

cyfr

onti

erd

uri

ng

the

per

iod

ofst

ud

y;

ban

ks

wh

ich

corr

esp

ond

toth

esh

aded

reg

ion

sh

ave

not

bee

nef

fici

ent

inan

yy

ear

inth

esa

mp

lep

erio

dco

mp

ared

toth

eot

her

ban

ks

inth

esa

mp

le

Table V.Composition of

production frontiers –intermediation approach

Benchmarkingthe Korean

banking sector

119

Page 14: 6.benchmarking the

Ban

k19

9219

9319

9419

9519

9619

9719

9819

9920

0020

0120

0220

03C

oun

tb

ank

Bu

san

Mu

tual

Sav

ing

sB

ank

IRS

DR

SIR

SC

RS

1C

hej

uB

ank

CR

SC

RS

CR

SC

RS

CR

SIR

SIR

SD

RS

IRS

IRS

IRS

5C

hoh

un

gB

ank

DR

SD

RS

DR

SD

RS

CR

SD

RS

DR

SC

RS

DR

SD

RS

NA

DR

S2

Cit

iban

kK

orea

IRS

IRS

IRS

CR

SC

RS

DR

SIR

SD

RS

DR

SD

RS

IRS

NA

2D

aeg

uB

ank

IRS

DR

SN

AD

RS

DR

SD

RS

IRS

CR

SD

RS

DR

SIR

SC

RS

2D

ebec

Mu

tual

Sav

ing

sB

ank

CR

SIR

SIR

SIR

SIR

SIR

S1

Eu

tteu

mM

utu

alS

avin

gs

Ban

kIR

SIR

SIR

S0

Han

aB

ank

CR

SD

RS

DR

SD

RS

DR

SD

RS

NA

NA

DR

SIR

SC

RS

CR

S3

Han

mau

mM

utu

alS

avin

gs

Ban

kC

RS

IRS

IRS

IRS

1H

anso

lM

utu

alS

avin

gs

Ban

kC

RS

DR

SIR

SD

RS

1H

ousi

ng

and

Com

mer

cial

Ban

kC

RS

CR

SC

RS

CR

SC

RS

CR

SD

RS

CR

SD

RS

7In

du

stri

alB

ank

ofK

orea

DR

SD

RS

CR

SD

RS

IRS

NA

CR

S2

Jeil

Mu

tual

Sav

ing

sB

ank

DR

SC

RS

IRS

IRS

DR

SIR

SN

A1

Jeon

bu

kB

ank

IRS

IRS

NA

IRS

DR

SIR

SIR

SIR

SD

RS

IRS

IRS

IRS

0Ji

nh

eun

gM

utu

alS

avin

gs

Ban

kC

RS

DR

SC

RS

CR

SIR

SIR

SIR

SIR

S3

Kan

gw

onB

ank

CR

SIR

SC

RS

IRS

IRS

DR

SIR

S2

Koo

km

inB

ank

DR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SC

RS

DR

SD

RS

1K

orea

Ex

chan

ge

Ban

kC

RS

CR

SC

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SN

AD

RS

3K

orea

Fir

stB

ank

DR

SD

RS

CR

SD

RS

DR

SD

RS

IRS

DR

SD

RS

IRS

IRS

CR

S2

Kor

eaM

utu

alS

avin

gs

Ban

kC

RS

CR

SC

RS

IRS

IRS

IRS

IRS

3K

orea

nF

ren

chB

ank

ing

Cor

p.

IRS

CR

SC

RS

CR

SD

RS

IRS

DR

SIR

SIR

SIR

S3

Kw

ang

juB

ank

CR

SN

AIR

SD

RS

DR

SIR

SIR

SD

RS

DR

S1

Ky

ong

nam

Ban

kD

RS

DR

SD

RS

DR

SC

RS

DR

SD

RS

1P

eace

Ban

kof

Kor

eaB

ank

DR

SIR

SC

RS

1P

ure

un

Mu

tual

Sav

ing

sB

ank

IRS

CR

SIR

SIR

S1

Pu

san

Ban

kIR

SD

RS

NA

DR

SD

RS

DR

SIR

SC

RS

DR

SN

AIR

SC

RS

2S

eou

lM

utu

alS

avin

gs

Ban

kC

RS

CR

SIR

SIR

SIR

SIR

SIR

SIR

S2

Sh

inh

anB

ank

NA

NA

DR

SD

RS

DR

SD

RS

CR

SD

RS

1S

hin

min

Mu

tual

Sav

ing

sB

ank

IRS

IRS

CR

S1

Sol

omon

Mu

tual

Sav

ing

sB

ank

IRS

IRS

IRS

IRS

IRS

IRS

IRS

0W

oori

Ban

kC

RS

CR

SC

RS

CR

SD

RS

DR

SC

RS

DR

SD

RS

5C

oun

ty

ear

55

65

75

510

22

17

60

Notes:

Th

eta

ble

show

sth

eev

olu

tion

ofre

turn

sto

scal

ein

the

Kor

ean

ban

kin

gse

ctor

du

rin

gth

ep

erio

d19

92-2

003;

CR

S,D

RS

,an

dIR

Sd

enot

eco

nst

ant

retu

rns

tosc

ale,

dec

reas

ing

retu

rns

tosc

ale,

and

incr

easi

ng

retu

rns

tosc

ale,

resp

ecti

vel

y;

“cou

nt

yea

r”d

enot

esth

en

um

ber

ofb

ank

sap

pea

rin

gon

the

effi

cien

cyfr

onti

erd

uri

ng

the

yea

r;“c

oun

tb

ank

”d

enot

esth

en

um

ber

ofti

mes

ab

ank

has

app

eare

don

the

effi

cien

cyfr

onti

erd

uri

ng

the

per

iod

ofst

ud

y;

ban

ks

wh

ich

corr

esp

ond

toth

esh

aded

reg

ion

sh

ave

not

bee

nef

fici

ent

inan

yy

ear

inth

esa

mp

lep

erio

dco

mp

ared

toth

eot

her

ban

ks

inth

esa

mp

le

Table VI.Composition ofproduction frontiers –value-added approach

BIJ18,1

120

Page 15: 6.benchmarking the

Ban

k19

9219

9319

9419

9519

9619

9719

9819

9920

0020

0120

0220

03C

oun

tb

ank

Bu

san

Mu

tual

Sav

ing

sB

ank

CR

SD

RS

CR

SC

RS

3C

hej

uB

ank

CR

SC

RS

CR

SC

RS

CR

SD

RS

DR

SD

RS

IRS

DR

SD

RS

5C

hoh

un

gB

ank

DR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

0C

itib

ank

Kor

eaD

RS

CR

SIR

SC

RS

DR

SD

RS

CR

SD

RS

DR

SD

RS

DR

SD

RS

3D

aeg

uB

ank

CR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SC

RS

DR

SC

RS

DR

SD

RS

3D

ebec

Mu

tual

Sav

ing

sB

ank

CR

SC

RS

CR

SC

RS

CR

SIR

S5

Eu

tteu

mM

utu

alS

avin

gs

Ban

kIR

SD

RS

IRS

0H

ana

Ban

kC

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

NA

DR

SD

RS

1H

anm

aum

Mu

tual

Sav

ing

sB

ank

DR

SD

RS

DR

SIR

S0

Han

sol

Mu

tual

Sav

ing

sB

ank

DR

SD

RS

DR

SD

RS

0H

ousi

ng

and

Com

mer

cial

Ban

kC

RS

CR

SC

RS

DR

SD

RS

DR

SD

RS

CR

SD

RS

4In

du

stri

alB

ank

ofK

orea

DR

SD

RS

DR

SD

RS

CR

SC

RS

DR

S2

Jeil

Mu

tual

Sav

ing

sB

ank

DR

SC

RS

DR

SD

RS

DR

SIR

SN

A1

Jeon

bu

kB

ank

IRS

DR

SC

RS

CR

SC

RS

CR

SC

RS

NA

DR

SIR

SD

RS

DR

S5

Jin

heu

ng

Mu

tual

Sav

ing

sB

ank

CR

SD

RS

CR

SD

RS

IRS

IRS

IRS

CR

S3

Kan

gw

onB

ank

CR

SC

RS

CR

SC

RS

DR

SD

RS

DR

S4

Koo

km

inB

ank

DR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

0K

orea

Ex

chan

ge

Ban

kC

RS

DR

SC

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

CR

SD

RS

DR

S3

Kor

eaF

irst

Ban

kC

RS

CR

SC

RS

DR

SD

RS

DR

SD

RS

IRS

DR

SN

AD

RS

DR

S3

Kor

eaM

utu

alS

avin

gs

Ban

kC

RS

CR

SC

RS

CR

SD

RS

CR

SC

RS

6K

orea

nF

ren

chB

ank

ing

Cor

p.

CR

SC

RS

CR

SC

RS

CR

SD

RS

DR

SIR

SD

RS

CR

S6

Kw

ang

juB

ank

DR

SD

RS

DR

SD

RS

DR

SD

RS

NA

DR

SD

RS

0K

yon

gn

amB

ank

DR

SD

RS

DR

SD

RS

DR

SD

RS

DR

S0

Pea

ceB

ank

ofK

orea

Ban

kD

RS

IRS

DR

S0

Pu

reu

nM

utu

alS

avin

gs

Ban

kC

RS

CR

SC

RS

CR

S4

Pu

san

Ban

kD

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

NA

DR

SIR

SD

RS

DR

S0

Seo

ul

Mu

tual

Sav

ing

sB

ank

CR

SC

RS

CR

SC

RS

IRS

DR

SD

RS

IRS

4S

hin

han

Ban

kC

RS

CR

SC

RS

DR

SC

RS

CR

SC

RS

DR

S6

Sh

inm

inM

utu

alS

avin

gs

Ban

kC

RS

CR

SIR

SIR

S2

Sol

omon

Mu

tual

Sav

ing

sB

ank

DR

SIR

SN

AD

RS

IRS

IRS

IRS

0W

oori

Ban

kD

RS

CR

SD

RS

DR

SD

RS

DR

SD

RS

DR

SD

RS

1C

oun

ty

ear

67

86

56

96

56

55

74

Notes:

Th

eta

ble

show

sth

eev

olu

tion

ofre

turn

sto

scal

ein

the

Kor

ean

ban

kin

gse

ctor

du

rin

gth

ep

erio

d19

92-2

003;

CR

S,D

RS

,an

dIR

Sd

enot

eco

nst

ant

retu

rns

tosc

ale,

dec

reas

ing

retu

rns

tosc

ale,

and

incr

easi

ng

retu

rns

tosc

ale,

resp

ecti

vel

y;

“cou

nt

yea

r”d

enot

esth

en

um

ber

ofb

ank

sap

pea

rin

gon

the

effi

cien

cyfr

onti

erd

uri

ng

the

yea

r;“c

oun

tb

ank

”d

enot

esth

en

um

ber

ofti

mes

ab

ank

has

app

eare

don

the

effi

cien

cyfr

onti

erd

uri

ng

the

per

iod

ofst

ud

y;

ban

ks

wh

ich

corr

esp

ond

toth

esh

aded

reg

ion

sh

ave

not

bee

nef

fici

ent

inan

yy

ear

inth

esa

mp

lep

erio

dco

mp

ared

toth

eot

her

ban

ks

inth

esa

mp

le

Table VII.Composition of

production frontiers –operating approach

Benchmarkingthe Korean

banking sector

121

Page 16: 6.benchmarking the

In general, the empirical findings presented in Tables V-VII clearly indicates that whilethe small banks tend to operate at CRS or IRS, the large banks tend to operate at CRS orDRS, findings which are similar to the earlier studies by among others Miller and Noulas(1996), McAllister and McManus (1993) and Noulas et al. (1990). To recap, McAllister andMcManus (1993) suggest that while the small banks have generally exhibited IRS, thelarge banks on the other hand tend to exhibit DRS, and at best CRS. As it appears, thesmall Korean banks have experienced IRS in their operations during the period of study.

One implication is that for the small Korean banks, a proportionate increase in inputswould result in more than a proportionate increase in outputs. Hence, the small Koreanbanks, which have been operating at IRS, could achieve significant cost savingsand efficiency gains by increasing its scale of operations. In other words, substantialgains can be obtained by altering the scale via internal growth or further consolidation inthe sector. In fact, in a perfectly competitive and contestable market, the efficientbanks should absorb the scale inefficient banks in order to exploit cost advantages.Thus, the banks that experience IRS should either eliminate their scale inefficiency,or be ready to become a prime target for acquiring banks, which can create value fromunderperforming banks by streamlining their operations and eliminating theirredundancies and inefficiencies (Evanoff and Israelvich, 1991).

On the other hand, the results seem to suggest that further increase in size would onlyresult in a smaller increase of outputs for every proportionate increase in inputs for thelarge banks, resulting from the fact that the large banks have been operating at DRS andCRS. If anything could be delved from the results, decision makers ought to be morecautious in promoting mergers, particularly among the large banks as means to enjoyefficiency gains.

The composition of the efficiency frontier shows that under the intermediationapproach the majority of Korean banks have experienced economies of scale (operatingat IRS) ranging from 63.64 percent in 1992 to 60.87 percent in 2003. On the other hand,the share of banks experiencing diseconomies of scale (operating at DRS) acceleratesfrom 18.18 percent in 1992 to 26.09 percent in 2003, suggesting the extra production costsfaced by the rapidly growing domestic banks. On the other hand, the share of scaleefficient banks (operating at CRS) decelerates from 18.18 percent in 1992 to 13.04 percentin 2003, signaling worsening SE over time.

Under the operating approach, we find that the majority of Korean banks haveexperienced diseconomies of scale (operating at DRS). During the period under study,the empirical findings suggest that the share of banks experiencing diseconomies ofscale (operating at DRS) accelerates sharply from 36.36 percent in 1992 to 52.17 percentin 2003. Likewise, the empirical findings seem to suggest that Korean banks whichexperienced economies of scale (operating at IRS) accelerates gradually from 9.09 percentin 1992 to 21.74 percent in 2003. On the other hand, the share of scale efficientbanks (operating at CRS) decelerates sharply from 54.55 percent in 1992 to 21.74 percentin 2003.

Similar to the intermediation approach, the empirical findings seem to suggest thatthe majority of Korean banks have experienced economies of scale (operating at IRS)under the value-added approach, ranging from 36.36 percent in 1992 before increasing to43.48 percent in 2003. On the other hand, the share of banks experiencing diseconomiesof scale (operating at DRS) decelerates from 45.45 percent in 1992 to 30.43 percent in

BIJ18,1

122

Page 17: 6.benchmarking the

2003, while the share of scale efficient banks (operating at CRS) also decelerates from45.45 percent in 1992 to 30.43 percent in 2003.

Overall, the empirical findings from this study seem to suggest that in thecase of the Korean banking sector, technical inefficiency has much to do with the scaleof production rather than the inefficient utilization of resources. The dominant effectof scale indicates that most of Korean banks have been operating at the “incorrect” ornon-optimal scale of operations. They either experience economies of scale (i.e. IRS)due to being at less than the optimum size, or diseconomies of scale (i.e. DRS) due to beingat more than the optimum size. Thus, decreasing or increasing the scale of productioncould result in cost savings, or efficiencies.

5. Concluding remarksThe present study examines the efficiency of the Korean banking sector during theperiod 1992-2003. The DEA method is employed to three different approaches todemonstrate how efficiency scores vary with changes in inputs and outputs. Duringthe period under study, Korean banks’ have exhibited a higher level of TE under theoperating approach compared to the intermediation and value-added approaches, whilebanks are characterized by relatively low level of TE under the intermediation approach.The empirical findings suggest that the inefficiency of the Korean banking sector waslargely due to scale rather than pure technical under the operating approach, while scaleinefficiency seems to outweighs pure technical inefficiency under the value-addedapproach. We find that under the intermediation approach, the Korean banking sector’sinefficiency stems largely from pure technical, rather than scale.

The empirical findings from this study present important ramifications. From thepolicy-making perspectives, the empirical findings clearly demonstrate the sensitivity ofthe efficiency scores derived from the DEA method to the choice of inputs and outputs.If anything could be delved, the policy makers ought to be more cautious before makingdecisions from the results derived from a single approach. The results clearly highlightsthat a bank may be the most efficient under certain approach, but may not be underanother approach.

The empirical findings from this study clearly suggest that the decline in the efficiencyof the Korean banks were mainly due to scale. The results imply that banks operating inthe Korean banking sector are either too small to benefit from the economies of scale,or too large to be scale efficient. Thus, from the policy-making perspective, the resultsimply that the relatively smaller banks could raise their efficiency levels by expanding,while the larger banks would need to scale down their operations to be scale efficient.

From the economies of scale perspectives, mergers among the small bankinginstitutions should be encouraged. This should entail the small banks to reap thebenefits of economies of scale. The larger institutions will also have better capacity toinvest in the state-of-the-art technologies, which could enhance the rate of TFP growthof the Korean banking sector. Furthermore, consolidation among the small bankinginstitutions may also enable them to better withstand macroeconomic shocks like theAsian financial crisis.

During the period under study, it is observed that in terms of SE, the larger bankshave lagged behind their smaller counterparts. The optimal size for a firm would be at apoint where it reaches a constant return to scale (CRS). To recap, a DMU operatingunder IRS needs to expand its operations, while a DMU, which is operating at DRS would

Benchmarkingthe Korean

banking sector

123

Page 18: 6.benchmarking the

on the contrary lead to downsizing. Perhaps, the reason why larger banks areunderperforming in comparison to their smaller peers could be that their size has becomemore of a burden than an advantage arising from the mergers and acquisitions activities.There are considerable costs associated with the management of a large organizationand making sure that these costs do not outweigh the size benefits is of great importance.The findings above could be a reflection to the belief that scope economies, rather thaneconomies of scale, are often seen as the main benefit banks derive by merging.

The empirical results which suggest that the Korean banks have been operating ata non-optimal scale of operations are in line with the findings by among others Sufian(2007) on the Malaysian banking sector. To recap, Sufian (2007) found that during thepost-merger period the inefficiency of the Malaysian banking sector was largely due toscale rather than pure technical. He also suggests that the mergers were particularlysuccessful for the small- and medium-sized banks, which have benefited the most fromexpansion and via economies of scale. If anything could be delved from the results,policy makers ought to be more cautious in promoting mergers as a mean to achievegreater efficiency in the banking sector. However, as mentioned earlier, the findingsabove could be a reflection to the belief that scope economies, rather than economies ofscale, are often viewed as the main benefit banks derive by merging, particularly withinthe context of the Korean banking sector.

However, over the long term, improvements may arise arising from a moreprogressive banks developing and introducing new technologies. Such innovativebanks may acquire improved status and benefit from scale operations. Although sizealone is not sufficient to guarantee efficiency, nonetheless being a large bank is animportant aspect to achieving sufficient scale to be able to invest in the identificationand development of cutting-edge technology and management systems. This certainlyapplies where there has been significant progress in enhancing the network of deliverychannels, including optimizing the number of branches within the bank’s network.

Owing to its limitations, the paper could be extended in a variety of ways. First, thescope of this study could be further extended to investigate changes in cost, allocative,and technical efficiencies over time. Second, the non-parametric frontier analysis used inthis paper could be combined with the stochastic frontier analysis method of estimatingthe frontier. This should testify to the robustness of the results against alternativeestimation methods. Finally, investigation of changes in productivity over time as aresult of technical change or technological progress or regress by employing theMPI could yet be another extension to the current paper.

Despite these limitations, the findings of this study are expected to contributesignificantly to the existing knowledge on the operating performance of the Koreanbanking sector. Nevertheless, the study have also provide further insight to bank-specificmanagement as well as the policymakers with regard to attaining optimal utilization ofcapacities, improvement in managerial expertise, efficient allocation of scarce resources,and most productive scale of operation of the banks in the industry. This may alsofacilitate directions for sustainable competitiveness of future banking operations in Korea.

Notes

1. During the period under study, there were 17 nationwide banks namely Cho Hung Bank,Commercial Bank of Korea (merged to form Hanvit Bank in 1999), Korea First Bank(nationalized in 1999), Hanil Bank (merged to form Hanvit Bank in 1999), Bank of Seoul

BIJ18,1

124

Page 19: 6.benchmarking the

(nationalized in 1998), Korea Exchange Bank, Shinhan Bank, Hanmi Bank (KorAm Bank),Donghwa Bank (acquired by Housing and Commercial Bank in 1998), Daedong Bank(acquired by Kookmin Bank in 1998), Hana Bank, Boram Bank (merged with Hana Bank in1999), Peace Bank of Korea (merged into Woori Holding Co. in 2001), Kookmin Bank,Housing and Commercial Bank (merged into Kookmin Bank in 2001), and Woori Holding Co.(formerly known as Hanvit Bank prior to be renamed in 2002).

2. There were ten regional commercial banks during the period under study. The regionalcommercial banks consist of Daegu Bank, Pusan Bank, Chung Chong Bank (acquired byHana Bank in 1998), Kwangju Bank, Bank of Cheju, Kyungki Bank (acquired by Hanmu Bankin 1998), Jeonbuk Bank, Kangwon Bank (merged into Cho Hung Bank in 1999),Kyungnam Bank, and Choongbuk Bank (merged into Cho Hung Bank in 1999).

3. Interested readers are referred to excellent books by Coelli et al. (1998), Thanassoulis (2001),and Cooper et al. (2000) for detailed discussions on the NIRS within the DEA method.

References

An, J., Bae, S.K. and Ratti, R.A. (2007), “Political influence and the banking sector: evidence fromKorea”, Oxford Bulletin of Economics and Statistics, Vol. 69 No. 1, pp. 75-98.

Banker, R.D., Charnes, A. and Cooper, W.W. (1984), “Some models for estimating technical andscale inefficiencies in data envelopment analysis”, Management Science, Vol. 30 No. 9,pp. 1078-92.

Beck, T. and Levine, R. (2004), “Stock markets, banks, and growth: panel evidence”, Journal ofBanking & Finance, Vol. 28 No. 3, pp. 423-42.

Benston, G.J. (1965), “Branch banking and economies of scale”, Journal of Finance, Vol. 20 No. 2,pp. 312-31.

Berger, A.N. (2007), “International comparisons of banking efficiency”, Financial Markets,Institutions and Instruments, Vol. 16 No. 3, pp. 119-44.

Berger, A.N. and Humphrey, D.B. (1997), “Efficiency of financial institutions: internationalsurvey and directions for future research”, European Journal of Operational Research,Vol. 98 No. 2, pp. 175-212.

Berger, A.N., Hunter, W.C. and Timme, S.G. (1993), “The efficiency of financial institutions:a review and preview of research past, present and future”, Journal of Banking andFinance, Vol. 17 Nos 2/3, pp. 221-49.

Canhoto, A. and Dermine, J. (2003), “A note on banking efficiency in Portugal, new vs. old banks”,Journal of Banking and Finance, Vol. 27 No. 11, pp. 2087-98.

Cetorelli, N. and Gambera, M. (2001), “Banking market structure, financial dependence andgrowth: international evidence from industry data”, Journal of Finance, Vol. 56 No. 2,pp. 617-48.

Charnes, A., Cooper, W.W. and Rhodes, E. (1978), “Measuring the efficiency of decision makingunits”, European Journal of Operational Research, Vol. 2 No. 6, pp. 429-44.

Choe, H. and Lee, B.S. (2003), “Korean bank governance reform after the Asian financial crisis”,Pacific-Basin Finance Journal, Vol. 11 No. 4, pp. 483-508.

Choi, S. and Hasan, I. (2005), “Ownership, governance, and bank performance: Koreanexperience”, Financial Markets, Institutions and Instruments, Vol. 14 No. 4, pp. 245-52.

Coelli, T. (1996), “A guide to DEAP version 2.1”, CEPA Working Paper 8/96, University ofNew England, Armidale.

Coelli, T., Prasada-Rao, D.S. and Battese, G.E. (1998), An Introduction to Efficiency andProductivity Analysis, Kluwer Academic Publishers, Boston, MA.

Benchmarkingthe Korean

banking sector

125

Page 20: 6.benchmarking the

Cooper, W.W., Seiford, L.M. and Tone, K. (2000), Data Envelopment Analysis, Kluwer AcademicPublishers, Boston, MA.

di Patti, E.B. and Hardy, D.C. (2005), “Financial sector liberalization, bank privatization,and efficiency: evidence from Pakistan”, Journal of Banking and Finance, Vol. 29 Nos 8/9,pp. 2381-406.

Drake, L. (2001), “Efficiency in UK banking”, Applied Financial Economics, Vol. 11, pp. 557-71.

Drake, L., Hall, M.J.B. and Simper, R. (2006), “The impact of macroeconomic and regulatoryfactors on bank efficiency: a non-parametric analysis of Hong Kong’s banking system”,Journal of Banking and Finance, Vol. 30 No. 5, pp. 1443-66.

Evanoff, D.D. and Israelvich, P.R. (1991), “Productive efficiency in banking”, EconomicPerspectives, Federal Reserve Bank of Chicago, July/August, pp. 11-32.

Fiordelisi, F. (2007), “Shareholder value efficiency in European banking”, Journal of Banking andFinance, Vol. 31 No. 7, pp. 2151-71.

Fukuyama, H. (1993), “Technical and scale efficiency of Japanese commercial banks:a non-parametric approach”, Applied Economics, Vol. 25 No. 8, pp. 1101-12.

Gilbert, R.A. and Wilson, P.W. (1998), “Effects of regulation on productivity of Korean banks”,Journal of Economics and Business, Vol. 50 No. 2, pp. 133-55.

Gregoriou, G.N. and Zhu, J. (2005), Evaluating Hedge Funds and CTA Performance:Data Envelopment Analysis Approach, Wiley, New York, NY.

Hao, J., Hunter, W.C. and Yang, W. (2001), “Deregulation and efficiency: the case of privateKorean banks”, Journal of Economics and Business, Vol. 53 Nos 2/3, pp. 237-54.

Hardy, D.C. and di Patti, E.B. (2001), “Bank reform and bank efficiency in Pakistan”, workingpaper, International Monetary Fund (IMF), Washington, DC.

Iimi, A. (2004), “Banking sector reforms in Pakistan: economies of scale and scope, and costcomplementarities”, Journal of Asian Economics, Vol. 15 No. 3, pp. 507-27.

Jeon, Y., Miller, S.M. and Natke, P.A. (2006), “Do foreign bank operations provide a stabilizinginfluence in Korea”, The Quarterly Review of Economics and Finance, Vol. 46 No. 1,pp. 82-109.

Leightner, J.E. and Lovell, C.A.K. (1998), “The impact of financial liberalization on theperformance of Thai banks”, Journal of Economics and Business, Vol. 50 No. 2, pp. 115-31.

Levine, R. (1998), “The legal environment, banks, and long run economic growth”, Journal ofMoney, Credit and Banking, Vol. 30 No. 3, pp. 596-613.

Levine, R. and Zervos, S. (1998), “Stock markets, banks, and economic growth”, AmericanEconomic Review, Vol. 88 No. 3, pp. 537-58.

McAllister, P.H. and McManus, D.A. (1993), “Resolving the scale efficiencies puzzle in banking”,Journal of Banking and Finance, Vol. 17 Nos 2/3, pp. 389-405.

Matthews, K. and Ismail, M. (2006), “Efficiency and productivity growth of domestic and foreigncommercial banks in Malaysia”, Cardiff Working Paper E2006/2.

Miller, S.M. and Noulas, A.G. (1996), “The technical efficiency of large bank production”, Journalof Banking and Finance, Vol. 20 No. 3, pp. 495-509.

Noulas, A.G., Ray, S.C. and Miller, S.M. (1990), “Returns to scale and input substitution for largeUS banks”, Journal of Money, Credit and Banking, Vol. 22 No. 1, pp. 94-108.

Park, K.H. and Weber, W.L. (2006), “A note on efficiency and productivity growth in the Koreanbanking industry, 1992-2002”, Journal of Banking and Finance, Vol. 30 No. 8, pp. 2371-86.

BIJ18,1

126

Page 21: 6.benchmarking the

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, Vol. 22 No. 3, pp. 301-18.

Rajan, R.G. and Zingales, L. (1998), “Financial dependence and growth”, American EconomicReview, Vol. 88 No. 3, pp. 559-86.

Sathye, M. (2001), “X-efficiency in Australian banking: an empirical investigation”, Journal ofBanking and Finance, Vol. 25 No. 3, pp. 613-30.

Sathye, M. (2003), “Efficiency of banks in a developing economy: the case of India”, EuropeanJournal of Operational Research, Vol. 148 No. 3, pp. 662-71.

Sealey, C. and Lindley, J.T. (1977), “Inputs, outputs and a theory of production and cost atdepository financial institutions”, Journal of Finance, Vol. 32, pp. 1251-66.

Shanmugam, K.R. and Das, A. (2004), “Efficiency of Indian commercial banks during the reformperiod”, Applied Financial Economics, Vol. 14 No. 9, pp. 681-6.

Siriopoulos, C. and Tziogkidis, P. (2010), “How do Greek banking institutions react aftersignificant events? – A DEA approach”, Omega, Vol. 38 No. 10, pp. 294-308.

Sturm, J.E. and Williams, B. (2008), “Characteristics determining the efficiency of foreign banksin Australia”, Journal of Banking and Finance, Vol. 32 No. 11, pp. 2346-60.

Sufian, F. (2007), “Mergers and acquisitons in the Malaysian banking industry: techncial andscale efficiency effects”, International Journal of Financial Services Management, Vol. 2No. 4, pp. 304-26.

Thanassoulis, E. (2001), Introduction to the Theory and Application of Data Envelopment Analysis:A Foundation Text with Integrated Software, Kluwer Academic Publishers, Boston, MA.

Unite, A.A. and Sullivan, M.J. (2003), “The effect of foreign entry and ownership on the Philippinedomestic banking market”, Journal of Banking and Finance, Vol. 27 No. 12, pp. 2323-45.

Webb, R.W. (2003), “Levels of efficiency in UK retail banks: a DEA window analysis”,International Journal of the Economics of Business, Vol. 10 No. 3, pp. 305-22.

Corresponding authorFadzlan Sufian can be contacted at: [email protected]

Benchmarkingthe Korean

banking sector

127

To purchase reprints of this article please e-mail: [email protected] visit our web site for further details: www.emeraldinsight.com/reprints