T INDIAN INSTITUTE OF MANAGEMENT AHMEDABAD-380 015 INDIA . No. IIMA is committed to maintain academic freedom.P. view(s) and conclusion(s) expressed in the working paper are those of the authors and not that of IIMA. 2011-03-01 March 2011 The main objective of the working paper series of the IIMA is to help faculty members. research staff and doctoral students to speedily share their research findings with professional colleagues and test their research findings at the pre-publication stage.INDIAN INSTITUTE OF MANAGEMENT AHMEDABAD INDIA Efficiency of Indian Banking Industry in the Post-Reform Era Amit Kumar Dwivedi D. Kumara Charyulu W. The opinion(s).

This paper seeks to determine the impact of various market and regulatory initiatives on efficiency improvements of Indian banks. 1 2 Keywords: Efficiency measurement. No. banking industry and DEA.P. Data Envelopment Analysis (DEA) has used to identify banks that are on the output frontier given the various inputs at their disposal.com 2 Former Post-Doctoral Fellow. The second phase of reforms began in 1997 with aim to reorganization measures. India E-mail: dwivedipgdm@gmail. Ahmedabad. efficiency of a firm relative to the efficiencies of firms in a sample. structural development which helped them for achieving universal benchmarks in terms of prudential norms and pre-eminent practices. Variable returns to scale (VRS) assumption for estimating the efficiency was not attempted.IIMA INDIAD Research and Publications Efficiency of Indian Banking Industry in the Post-Reform Era Amit Kumar Dwivedi D. new private banks and foreign banks have showed high efficiency over a period time than remaining banks. Kumara Charyulu Abstract One of the major objectives of Indian banking sector reforms was to encourage operational self-sufficiency. 2 . Indian Institute of Management. human capital development. It was found from the results that national banks. technological up-gradation. flexibility and competition in the system and to increase the banking standards in India to the international best practices. Indian Institute of Management. The present study is confined only to the Constant-Return-to-Scale (CRS) assumption of decision making units (DMUs). 1 Academic Associate. Efficiency of firm is measured in terms of its relative performance that is. Ahmedabad.com W. India Email: kumaracharya@gmail. 2011-03-01 Page No.

The major dominance of commercial banks can be easily found in Indian banking. The overall development has been lucrative with enhancement in banking industry efficiency and productivity. i. Under the first category of W.IIMA INDIAD Research and Publications I. The Indian banking sector has two kinds of scheduled banks i. particularly in the wake of the worldwide economic disasters. Our banking industry is a mixture of public.e. This industry recorded CAGR growth of 18 percent as compared to country’s average GDP growth of 7. has always played a key role in prevention the economic catastrophe from reaching terrible volume in the country. 3 . Indian banks have remained flexible even throughout the height of the subprime catastrophe and the subsequent financial turmoil. It has achieved enormous appreciation for its strength.2 percent during the same period.3 trillion. No. Introduction Indian banking industry. scheduled commercial banks and scheduled co-operative banks. The Indian banking industry is measured as a flourishing and the secure in the banking world. 2011-03-01 Page No. The country’s economy growth rate by over 9 percent since last several years and that has made it regarded as the next economic power in the world. from 68 percent to 135 percent. Indian banking industry has increased its total assets more than five times between March 2000 and March 2010.. It should be underlined here is financial turmoil which hit the western economies in 2008 and the distress effect widened to the majority of the other countries but Indian banking system survived with the distress and showed the stable performance. private and foreign ownerships. which pressed its worldwide counterparts to the edge of fall down. although the co-operative and regional rural banks have little business segment.P. 2008).e. the Indian banking system is among the healthier performers in the This sector is world. The commercial banking assets to GDP ratio has increased to nearly 100 percent while the ratio of bank’s business to GDP has recorded nearly twofold. If we compare the business of top three banks in total assets and in terms of return on assets. tremendously competitive and recorded as growing in the right trend (Ram Mohan. US$250 billion to more than US$1. the backbone of the country’s economy.

. The Indian Economy Conference. 2011-03-01 Page No. 4 . banks are divided into two types i. ii) Private Banks (25). observance in view these 5 conditions. No. Reserve Bank of India (RBI) has initiated quite a few measures to ensure safety and consistency of the banking system in the country and at the same point in time to support banks to play an effective role in accelerating the economic growth process. turn down in efficiency and erosion in profitability had developed in the system.e. certain weaknesses. Old private banks (17) and New-private banks (8). Reforms and Banking system In the post liberalization-era. Reddy (2002). i. Under Private Bank category. 4Although the Indian banks have contributed much in the Indian economy. type.Y.e. banks are more clearly defined according to nationalization and privatization. The banks under public banks are Nationalized Banks (20) and State Banks of India (with its associates.V. four types of entities have found based on their establishments and legal obligations. Development 5 Narsimham CommitteeI W. II. 3 4 The number in brackets are No. USA.IIMA INDIAD Research and Publications scheduled banks.P. They are: i) Public banks (28)3 . “Monetary and Financial Sector Reforms in India: A Perspective”. the number is come to 8). One of the major objectives of Indian banking sector reforms was to encourage operational self-sufficiency. the Committee on Financial System (CFS) was lay down. of the Banks exists in that Dr. flexibility and competition in the system and to increase the banking standards in India to the international best practices . iii) Foreign Banks working in India (29) and iv) Regional Rural Banks (91) The second category of scheduled cooperative banks consists of: i) Scheduled Urban Co-operative banks (55) and ii) Scheduled State Co-operative Banks (16) Under public & private sector. Programon Comparative Economic Practitioner’s (PCED) at Cornell University.

disclosures and transparency norms etc. While the reform process has resulted in the private sector replacing the government as the source of resources for public sector banks (PSBs).capitalized and become competitive. technological up-gradation. The competitive pressures to improve efficiency in the banking sector has resulted in a switch from traditional paper based banking to electronic banking. The first reform focused on introduction of several prudential norms. The financial reform process has commenced since 1991 which was made the banking sector healthy.P. however. major changes in the policy framework. the infusion of private equity capital has led to shareholders challenges to bureaucratic decision making. 5 . structural development which helped them for achieving universal benchmarks in terms of prudential norms and pre-eminent practices. financial sector reforms focused on improving the competitive efficiency of the banking system. The second phase of reforms began in 1997 with aim to reorganization measures. The committee also argued that ‘economic reforms in thereal sector of economy will. and formation of competitive atmosphere. However. No. This paper seeks to determine the impact of various market and regulatory initiatives on efficiency improvements of Indian banks. The Narsimham Committee has provided the proposal for reforming the financial sector. It focused on several issues like. releasing of more funds to banks. income recognition. well. use information technology and shift of emphasis from brick and mortar banking to use of ATMs. human capital development. Later. 2011-03-01 Page No. fail to realize their full potential without a parallel reform of the financial sector. capital adequacy. PSBs also face increasing competition not only from private and foreign banks but also from growing non-banking financial intermediaries like mutual funds and other capital market entities. The reform process has shifted the focus of public sector dominated banking system from social banking to a more efficient and profit oriented industry. The Financial sector reforms were undertaken in 1992 based on the recommendations of the CFS. sound.IIMA INDIAD Research and Publications Reserve Bank of India has implemented banking sector reforms in two phases. deregulation in interest rates. W.

data established with respect to both technical and scale banking: iquidasset s and other investments . Georgiou. 7 W. appeared inefficiency years. and Scale Efficiencies in i) net-interest income.Q Outputs: density enhances efficiency. No. Greek banking industry improved remarkably mainly since 1999. 6 . in the 2006. of output and productivity the Greek banking industry for the period 1990– intermediation. input-oriented efficiency of Indian public have sector banking industry.Capital Cost THE ZAMBIAN BANKING 3 employees. Capital and total factor productivity have also banking industry 1990– “other” services. 2011-03-01 Page No. approach Deposits 4.IIMA IIMA INDIAD INDIAD Research and Publications Research and Publications Sl. Inputs vary from purely financial such 1. stochastic frontier provisions or product/service related – loans. 2008 3.Labour. investments and non-interest income. i) physical capital. The efficiency Allocative Efficiency of the intermediation credit output Misra dynamics has been the post reform period for the majority Indian Banking System in studied for threerelatively of the Statesto the specification in terms of broad the Post-Reform scores were found to be sensitive sectors of each State viz.L 1 as interest non-interest expenses to purely physical like number of branches andEFFICIENCY IN Cost Anthony and frontier cost function foreign banks remains wide. 2009 6.P. banks during the post-reform period 1997–2003.Ea 6 Panayiotis P. and outputs as evident from Table 1. Year 2008 2. Pure Technical. industry.In 5 rning advances. 1. Nonparametric DEA Relatively high cost efficiency levels for Indian Distribution of cost and Ray . Recent studies on efficiency of Banking system in India W. . inefficiencyare either income relatedexemplifiedor non-interest incomeSECTOR(1998– 2006) Outputs model techniques as interest by higher loan loss 3. Other income . capital (mainly IT) investments andbanks.P. Leigh Drake. at least in the case of Indian (1997–2003). 3.L 2 Authors Sunil KUMAR & Rachita GULATI Table 1. methodology. industry Analysis(1981-1992) and and services (1993. profit efficiency: Evidence. Brief review of literature Assessment Parameters/ Scope Brief Recommendations Drivers Outputs: 1. No. due to the 2006. he effect of labor quality on 2 . Analysis provides that the factors like market share.Fu 2 Musonda 2.translog stochastic the efficiency gap between domestic banks and DETERMINANTS OF Cost abour . Period: A State Level inputs agriculture. ii) non-interest Setting Analysis have any significant impact on the overall technical (1992-2005) contains three while input vector A few DEA-based studies of efficiency in the Indian banking system Inputs: 3. for the period 1990– structural changes that took place within the productivity measurement of Greek Staikouras 2006. They have used a variety of specifications for recent scores ii) labour. 2011-03-01 Page No. 2001 Methodology III. and (c) the user-cost approach.Fu nds.Slacks and Targets profitability.T Maximilian approach. and 2. and asset quality do not Indian Public Sectorincome. 2010 5. inputs’ and TFP growth to output approach. conditional mean (CM) Banks continue to exhibit poor risk assessment . Non-parametric frontier Powerful size-efficiency relationships are otalEfficiency in Japanese loans and bills discounted.1 Abhiman Das Nonparametric kernel There is no definite evidence that privatization from Indian uasi-fixed inputs4 banking .Capital. COST nding . 4.data envelopment appear as inputs or outputs depending upon whether the authors work with analysis (DEA) the Biswa Swarup Theor production interpretation of in the overall allocative efficiency in improvement banking business.2001) Inputs: Subhash C. 1 . 2.Logistic Regression The results of logistic regression analysis Technical. payment services and productivity growth in the Evangelia A. (b) the intermediation approach. and inputs iii) loanable funds. 2003 and outputs.T Athanasoglou. estimation banks vestments.5 as well as the contribution of (a) the production improvement in the quality of human capital. Assesses the evolution Assesses the evolution of output and productivity in Assessing output and hree categories: financial1 . growth.t Christos C.

2011-03-01 Page No. higher under value-added approach than under and period dequacy . Results suggest a large asymmetry between maximum deregulation and reflects the ability of a firm to obtain & Saibal Envelopment Analysis banks regarding their technical efficiency over the efficiency: An empirical size. intermediation approach. Efficiency (panel estimation).1 . Lovell (1994).c 3 Output: efficient in frontier to the public sector and the derived relative to banksbest-practice bank on thecomparisonthat uses a comparable 2007)of the and on the basis mix of empirical investigation foreign banks tal loans. reflecting the importance of and Profit Efficiency:eposits.3 of firms in a sample.b 1 Ghosh (DEA).1 nterest . Variables: 9. intermediation approach.B.e 4 frontier”.c 3 approaches viz.Univariate analysis and inefficiencies reference side of “best umber of employees. 3 7. Hall and economic efficiency measures however are completely different in nature and are not likely to coincide for other reasons than by chance.i and RBI inputs Efficiency: Theabor cost andl . using Japanese banking Specification a recent cross-section model sample An empirical analysis in securities. More detailed reviews of the DEA methodology were also presented by Seiford and Thrall (1990). .o 2.P..fixed. 2011-03-01 Page No.IIMA INDIAD Research and Publications envelopment analysis. Best practice frontier. No. on-performing loans and . analysis wnership. 2 and soundness as determined by bank’s capital of two .P. nvestments. . a termefficiency appear to suggest that big state-owned originally coined by Farrell (1957) denotes Output: three measuresbanks performed reasonably well and are more likely to operate at higher levels of profit efficiency.n 5 anagement qualityproduction technology. Retail and wholesale deposits. which reflects the ability of aof Indian banks . 2006 Abhiman Das 1. These two measuresbanks thenmarked with relatively low level measure of value-added approach Indian are were combined to provide a .Non-parametric DEA & Saibal methodology levels in profits.O 3 General and 1 of administrative expenses.d 1 Ghosh 2.employed to but apparentlyperspective in technical efficiency differentiate how efficiency scores vary with under measuring technical efficiencychanges Constant-Return-to-Scale (CRS).i inputs. Data Envelopment Analysis (DEA) involves the use of linear programming methods to construct a non-parametric piecewise surface (or frontier) over the data.. He also proposed that the efficiency of a firm consists dvances and other non-interesta adequacy ratio. to efficiency. 2 frontier given the variousdata for at their public sector Efficiency of each institution is then disposal. fee based incomes.Fi equipment). the of firm is measured in terms of its relative performance that is. which1. the domestic private banks are becoming more Indian Experienceapital cost (1980. A 2 approach for estimating relative efficiency is withon the revenueto the banking practice Non-parametric 3. Under the latter approach. and allocative efficiency. ank . 8 able to calculate efficiencies relative to this surface. W. Lovell (1993).n determinants of activity.m 6 and operating approach of technical efficiency and there was a persistent have been The output and input declining trend will coincide when total economic efficiency. banks 60 Indian commercial 2. 2010 Santosh Stochastic Frontier 1. No. . years. xed assets (premises and 2 . Nonparametric Data Financial components: technical efficiency.Three different 2.a 4 firm to use the inputs in optimal proportions. The proximate determinants of profit Analysis of Indian Banksand assets inefficiency (1992-2004) quity . so as to be W.after financial liberalization there has been no Financial LiberalizationInput: Data Envelopment Analysis (DEA) was used to identify banks that are on the output Kumar Das Approach (cost frontier) significant change in the cost efficiency of the and Banking Sector cost. 2 output from a given set of inputs.o 3 8. given their respective prices the the (1992-2002) ratio. ther investments . Technical efficiency estimates were found to be during the post reform apital . 2009 Abhiman a firm relative to the efficiencies of firms in a sample. (1997) ther income.i maximum output that can be obtained with a given set of input quantities for a given set A close relationship is observed between efficiency oans andl . costs and lower indicate high levels of efficiency in A formal econometric Deregulation Input: Financial efficiency of Das 1.to 1 nvestment in securities and 2 . Charnes et al (1995) and Seiford (1996). The allocative in inputs and outputs J. Ali and Seiford (1993). analyse the technical and scale efficiency in IV. 9 .

1] W. No. However. the assumption of CRS is correct only as long as firms are operating at an optimal scale (Coelli et al. 0 λ 0 θε(0. A measure of =1 firm i’s inputs can be proportionally reduced without any loss in output. Y is an m*n matrix of outputs and X is an h*n matrix of inputs. isi firm i’s index of technical efficiency relative to the other firms in the sample. Various constraints on inputs like financing and the goals of the farmer may cause the firm to operate at a non-optimal scale. Subject to –yi + Y 0. Thus. 2011-03-01 Page No.IIMA INDIAD Research and Publications Consider ‘n’ firms producing ‘m’ different outputs using ‘h’ different inputs. y and x i i represents the output and input of firm ‘i’ i respectively. θx – i Xλ0. 2002).P. The technical efficiency (TE) measure under the assumption of constant returns to scale (CRS) can be formulated as follows: Min . Thus. Using the indicates that the firm is completely technically efficient.measures how much i CRS-DEA model when firms are not operating at their optimal scale will cause the TEmeasures to be influenced by scale efficiencies and thus the measure of technical efficiency will be incorrect. Y and X are the efficient projections on the frontier.1] and solved for each firm in the sample. Subject to –yi + Y 0. Both matrices contain data for all ‘n’ firms. 10 . θx – i Xλ  . N1’ =1 λ 0 θε(0. By adding a convexity constraint to the model above variable returns to scale (VRS) is instead assumed: Min . 1.

11 . labor and capital to make loans and earn interest income.P. In the input category. one due to scale inefficiency and one due to “pure” technical efficiency. total operating expenses and deposits as inputs and loans (advances) and non-interest income as outputs. higher investments simply imply that banks are not pushing loans adequately. A typical financial intermediation role for banks involves the use of deposits together with physical inputs of land. If there is a difference in the two TE scores for a particular DMU. the present study is confined only to the CRS assumption of DMUs.. The ). three variables were included. bank investments (in addition to loans) are also considered as a legitimate output. This constraint makes the comparison of firms of similar size possible. In most banking systems. 2 per unit per annum. At the heart is the question of whether deposits are input or output. loans/advances (Y ) per unit per annum and non-interest income (Y ) 1 no. 2006). 2 DEA model was estimated using DEAP 2. 1996). TECRS = TEVRS x SE However. so that the data is enveloped more tightly. But such investments in India are mostly in government securities which are often thought of as reflections of “lazy” banking.0 algorithm (Coelli.e. They were W.of branches (X 1 total operating expenses per annum (X ) and deposits (X3). 2011-03-01 Page No. then this indicates that the DMU has scale inefficiency and that the scale inefficiency can be calculated from the differences between the TEVRS score and the TECRS score (Coelli.IIMA INDIAD Research and Publications The new constraint is N1’ =1 where N1 is a n*1 vector of ones. deposits are not coveted as an independent output. We did not use the VRS assumption for estimating the efficiency. According to this line of thinking. No. The technical efficiency measures under VRS (Pure TE) will always be at least as great as under the CRS-assumption. instead they are treated only as a conduit to generating loans. Banks also recognize the importance of generating non-interest income as an anti-dote to the variability in interest income. that is hardly the case in banking. Many studies have decomposed the TE scores obtained from a CRS DEA into two components. This approach suggests that we should treat the number of bank branches. While inputs and outputs are easily identified in most businesses. Each observation included two outputs i. In this formulation. by forming a convex hull of intersecting planes.

total-public sector banks and old private banks have improved their efficiency in the recent times when compared to 2005 status.000 0. These results are based on data from the Performance of Indian Banks by the Association of Indian Banks and the Banking Statistics and Annual Reports of the Reserve Bank of India (RBI).000 0.972 1.934 1. V.936 1.000 0.984 1.000 0.986 2007-08 1. All banks estimated mean technical under DEA-CRS model were 95.000 1. No.000 0.994 2009-1 0 1.986 1.972 1.878 1.000 0.000 0.889 1.956 2006-07 1.97 9 It is concluded from the table that the national banks.000 0.966 0.000 0.000 1. They are all performed equally in relative terms.000 0. 12 .000 1.904 1.000 0. W.996 2008-09 1. Results and Discussions The technical efficiencies of sample banking units under CRS models of DEA approach is presented in table 2. The units under state-owned banks.P.000 1.000 0.924 1. Table 2 Efficiency of different banking units Type State-owned banks National Banks Total – public sector banks Old-private banks New-private banks Total private banks Foreign banks All-banks 2005-06 0. new private banks and foreign banks have showed high efficiency over a period time than remaining banks. For the present study data from 2005 to 2010 was used for the analysis. we do not use investments as banks’ output.000 1.000 0.000 1.000 0.9 per cent in 2010.6 per cent in 2005 and move up to 97.982 1. 2011-03-01 Page No.IIMA INDIAD Research and Publications In view of this.000 1. most of the sample units showed more than 90 per cent efficiency.000 1. In terms of technical efficiency. The scheduled commercial banks together improved their efficiency continuously up to 2007-08 and after there was a slight decline in the last two subsequent years.000 0.

IIMA INDIAD Research and Publications In general. Similarly. No. 2011-03-01 Page No. the efficiency scores were improved among all the scheduled commercial bank types when compared with earlier research studies. The variables influencing the bank efficiency negatively could also be identified.P. Further scrutiny of these results is needed by bank-wise to get better understanding of them. 13 . identification of factors for enhancing the efficiency could also be analyzed. W. This comprehensive assessment will bring better clarity about the efficiency in Indian banking system.

(New York: Oxford University Press). & Hall. Pure Technical and Scale Efficiency in Indian Publick Sector Banks using Data Envelopment Analysis”. Farrell. Pp: 1-49. Pp: 294-315. Eurasian Journal of Business and Economics.. C. Abhiman & Gosh. C.8/96. th March IGIDR. Das. 12th Money and Finance Conference.39.K Lovell and S. Pp: 219-245 Coelli. 11-12 (India). 27.B.A. Mumbai W. S. Australian Journal of Agricultural Economics. Pp: 3-67. (1996): “A Guide to DEAP 2. Vol.1 (2). No.E Battese (2002): “An Introduction to Efficiency and Productivity Analysis”. 14 . (2008): “An Examination of Technical. CEPA working paper No. Journal of Banking & Finance. Vol. J. T. R. Gulati. Pp:193-221. Vol no.. P. Coelli.D.(2003): “Efficiency in Japanese Banking: An Empirical Analysis”. Leigh. S. University of New England. ISBN 1863894969. (1995): “Recent developments in frontier modeling and efficiency measurement.S Rao and G.) The Measurement of Productive Efficiency. Kumar. M. 15.P. T.A. D.J (1957): “Generalized Farrell Measures of Efficiency: An application to Milk processing in Swedish Dairy plants”.89. Pp:33-69.J.O Fried.(2010): “Financial Liberalization and Banking Sector Efficiency: The Indian Experience”.K. (London: Kluwer Academic Publishers).. Department of Econometrics.S Schmidt (ed. Das. Abhiman & Gosh. Review of Financial Economics. Economic Journal..IIMA INDIAD Research and Publications References Coelli T. 2011-03-01 Page No. Das. Munich Personal RePEc Archive.1: A Data Envelopment Analysis Computer Program”.K (1993): “Production Frontiers and Productive Efficiency” in H. Saibal (2006): “ Financial Deregulation and Efficiency: An empirical analysis of Indian Banking during the Post-Reform Period”. Lovell. Saibal (2009): “Financial Deregulation and Profit Efficiency: A Non-Parametric Analysis of Indian Banks”. Pp:891-917.