Post Financial Deregulations Era and Efficiency of Pakistan Banking Sector

Post Financial Deregulations Era and Efficiency of Pakistan Banking Sector Rafaqet Ali PhD Student Department of Economics Gomal University, D. I. Khan Muhammad Afzal Professor Department of Management Sciences, COMSAT Institute of Information Technology, Islamabad Abstract This study examines technical, pure technical and scale efficiency of Pakistani banks during post financial reforms period by using data from 2004 to 2009. Non parametric Data Envelopment Analysis is applied for this purpose. The results suggest that technical efficiency decreases during middle period but increases in recent years. Small banks are found more efficient as compare to medium and large banks. Decomposition analysis explains dominance of scale inefficiency over pure technical efficiency for technical efficiency of banks. This study further explores potential determinants of calculated efficiencies. Diversification of income, market share in respect of deposits and issuance of loans are positively associated whereas bad cost management of banks and current depressing economic situation of this country exerts negative impact on banking efficiency. Keywords: Pakistani banks; deregulations; efficiency; determinants 1. Introduction Financial system is a vital part of any modern economy and a well developed financial system is pre-requisite for optimal utilization of financial resources. Role of financial institutions in this system is of paramount importance and underdeveloped financial institutions can be hindrance for financial markets to channelize surplus saving efficiently into productive investments. Dominance of banking sector in overall financial system is one of the major characteristics of developing countries (Limi, 2004; Staub, et al., 2010). Like other developing countries, banking industry is also a vital entity in Pakistan financial system (Hussain, 1999; Zaidi, 2005). Apart from theoretical development on finance and growth nexus, ample empirical evidences are available that financial development through financial intermediary role of banks enhances economic growth (Ataullah & Le, 2006). Considering active role of this sector in any economy, improvement in performance of this sector has become crucial question. Most of countries started financial reforms in their banking systems from 1980s with the aim of enhancing performance and efficiency of this sector (Hardy and Patti, 2001). Banking reforms have been implemented in Pakistan since 1990s in order to improve performance, competitiveness and services quality of this sector. Since the years 1991 to date, a number of developments have occurred in banking industry of this country such as: domestic private banks are now enjoying hefty share, some big nationalized banks have been privatized, automation services of banks are flourishing etc i. Efficiency analysis is one of the ways to examine the performance of banking sector. Moreover, this analysis is helpful to understand which banks are most efficient compared to their counterpart s at point in time. Various studies examined efficiency of banking sector all over the world with the intention to assess the performance of this sector ii (see Berger & Mester, 1997; Isik, & Hassan, 2002; Das & Ghosh (2006). Apart from that, numerous empirical studies have also been carried out to finding out prudent determinants of banking efficiency (e.g. Miller & Noulas 1996; Hao et al., 2001; Dacanay III 2007). 1177

More recently global financial crisis started from United State of America at the end of 2007 with the emergence of sub-prime crisis and through contagion affects this crisis spread all over the world in 2008. The major sources of its infectious spread are globalization and technology and this is probably the largest crisis after the great depression of 1930s (Llanto & Badiola, 2009). There is a possibility that banking sector of developing countries might get contagion effect of this crisis. This paper has three objectives. First objective is to empirically address the efficiency of Pakistan banking industry in post reform period. Second objective deals with examining the important determinants of this sector using more recent data. The last objective is to see whether banking sector of Pakistan was affected by the contagious effect of the recent global financial crisis. The rest of this study consists of the following parts. Section 2 provides brief historical summary of Pakistan banking sector. Section 3 is meant for methodology and data issues. Results discussions are given in section 4 whereas section 5 presents concluding remarks. 2. History of Pakistan banking industry At the time of independence in 1947, Pakistan inherited a weak banking structure. All the Indian banks moved their head offices to India and closed most of their branches. There was only one bank which had its head office in Pakistan in August 1947 and this was the only bank which moved its head office from India to Pakistan. In addition to that there were a few foreign banks which were merely attached with the financial business of international trade. There was no separate central bank of this country at that time. State Bank of Pakistan (SBP) was established as a central bank in 1948 and was entrusted the goal to strengthen banking sector of Pakistan. Later on, National Bank of Pakistan (NBP) was established and this bank was marked as agency bank of SBP to handle the banking business smoothly (Zaidi 2005). In earlier period, informal financial intermediary business was prevailed but over the period of time, formal banking system replaced informal sector gradually (Hussain 1999). Expansion in branches of commercial banks had been observed during 1950s. Another important development in Pakistan s banking system was expansion in credit provision by Pakistani banks. The share of overall credit, issued by Pakistani banks increased from 38 percent in 1952 to 59 percent in 1955. During 1960 to 1965, banking sector of Pakistan flourished and branch offices of scheduled banks increased from 430 in 1960 to 1591 in 1965. Several new banks were also established. Exponential rising trend in expansion of banking sector of Pakistan occurred during 1960s and total number of branches reached up to 3133 in June 1970 (Zaidi, 2005; Meenai, 2010). A report of State Bank of Pakistan in 1970 revealed that only eighty-eight account holders in banks had access to almost 25 percent of the total credit and most of these account holders were the directors of banks themselves. In the wake of this and other contemporary issues, banking reforms were introduced in 1972 (Zaidi, 2005). In January 1974, banks were nationalized by the government. Pakistani banks operating before nationalization were merged into five banks namely; National Bank of Pakistan, Habib Bank Ltd., United Bank Ltd., Muslim Commercial Bank Ltd. and Allied Bank of Pakistan (Meenai, 2010). This nationalization policy drastically altered financial system of this country. Besides the other social objectives of nationalization, nationwide branches expansions and allocation of credit to public and agricultural sectors were notable. Nationalization witnessed inefficient role of banking sector due to overstaffing, over-branching and political influence for credit allocations etc. These problems signaled negative repercussions for financial system of Pakistan (Limi, 2004). Moreover, during this era real interest rate was negative in most of the years (Hussain, 1999). In sum, prior to 1990 administrative control on interest rates, government control on the banking system, credit allocation to the priority sectors instead of borrowing firms profitability were the major characteristics of financial sector of Pakistan which impeded efficiency of this sector. The public sector s ownership of commercial banks created lot of problems e.g. political intervention into credit allocation, loan recovery problems and deterioration in services quality etc. Nationalized commercial banks were not operating on commercial principles and consequently the efficiency, market responsiveness and financial strength of the banks were badly affected; therefore, reforms in the banking sector were introduced during 1990s (Khan, 1996). The notable features of Pakistan s banking sector reforms are: (a) determination of interest as per market signals; (b) abolition of subsidized credit allocation; (c) restructuring and re-capitalizing the state-owned commercial banks through autonomy and privatization of nationalized banks; (d) removal of restriction on 1178

opening new private banks and; (e) improvement in the regulation and supervision criteria of the financial institutions and many more (Khan and Khan, 2007). As mentioned earlier, to encourage the private sector, initiatives for opening-up of new banks were also taken. Ten new commercial banks were permitted to start their operations in 1991. In the later years eleven more banks were also allowed to be operational. Moreover, two provincial banks; Bank of Punjab and Bank of Khyber were also announced as scheduled banks in 1994 iii. Considerable autonomous power was also given to State Bank of Pakistan to act more neutrally. Among the other development during the financial reform period, privatization of four state-owned banks was made in order to improve the efficiency and operations of these banks (Zaidi 2005). Even during the recent past some reforms were introduced in the banking sector of Pakistan. These include (i) liberalization of bank branches in order to enhance their share in the market, (ii) merger and acquisition of banks, (iii) notable measures were taken to reduce non-performing loans, (iv) enhancement of automation services of banks i.e. ATM, on-line banking (Khan and Khan, 2007). All these measures strengthened the confidence of private sector on financial sector of Pakistan. Now, Pakistan s banking sector consists of 25 private local banks along with 6 foreign banks. Besides that, there are four public sector commercial banks and three specialized scheduled banks. It is pertinent to note that in the year 1990, prior to financial reforms, there was no private bank in Pakistan but now the situation has completely changed and over the period of time private sector got dominance in the banking sector. The year 2009 witnessed that share of private local bank assets in the total banking sector is more than 80 percent (GOP 2008-09). 3. Methodology and Data Two types of efficiency estimation methodologies have been applied in literature in order to assess banking efficiency. First, econometric based parametric techniques and second, linear programming based non parametric techniques. Data Envelopment Analysis (DEA) is non-parametric technique which has been extensively used in financial literature. Instead of pre-specifying production frontier, DEA technique formulates production frontier according to actual outputs and inputs data used in the analysis (Miller and Noulas, 1996; Coelli, 1996). This approach is more feasible for small sample as compared to parametric technique (Damar, 2006). This study uses small cross section sample for each year, therefore, we apply this approach. Farell (1957) is the pioneered the efficiency analysis concept. However, Charnes et al., (1978) introduced non parametric DEA method to assess efficiency of the firms at micro level. For efficiency analysis through DEA technique, consider each firm, out of total set of firms iv (T), uses K numbers of outputs marked as z i and M numbers of inputs known as y i. Further, K x T is output matrix: Z and M x T is input matrix: Y, shows the data of all T numbers of firms. In DEA framework, efficiency is examined through ratio form, therefore; for single firm (bank), ratio of all outputs to all inputs is measured. Usually multiple outputs and inputs are used in financial literature. Considering this, weights are needed to be assigned to all outputs and inputs, hence K x 1 vector of output weights: u and M x 1 vector of input weights: v are applied. In order to have optimal weights, the following problem is defined. max u,v (uz i /vy i ) (1 Subject to uz j /vy j < 1, j = 1, 2 T (2 u, v 0 (3 This problem mentioned in equations 1 to 3 shows that maximize the efficiency of i th firm subject to the constraint that efficiency of all firms ( from 1 to T) is less than or equal to unity and values of output as well as input weights are non negative. Nevertheless, infinite solution occurs with this problem, therefore, the constraint vy i =1 is to be applied to overcome infinite solution. By applying this constraint, the following new formulation occurs. 1179

max u,v (uz i ). (4 Subject to vy i = 1, uz j /vy j < 1, j = 1, 2 T (5 u, v 0 (6 Duality in linear programming, which is input oriented problem, is presenting in the following form: min θ, λ θ (7 Subject to. z i + Zλ 0 (8 θy i Yλ 0 (9 λ 0 (10 In the above problem θ is a scalar refers to efficiency of i th firm ranging from 0 to 1 and; λ represents vector of constant (n x 1);. This equation has to be solved for each bank up to T numbers of banks. Efficiency can be measured in output oriented and/or input oriented approaches. With given level of inputs, production of maximum outputs is known output oriented technique whereas its opposite is output oriented technique which postulates that considering output is fixed, minimum level of inputs utilization is input oriented approach. Following Das and Ghosh (2006); Sufian (2009), we apply input oriented approach. The above mentioned DEA based efficiency analysis is based on constant return to scale (CRS) assumption and is generally marked as technical efficiency (TE) which means; for given level of outputs, minimum use of inputs by the banks. Banker et al. (1984) introduced return to scale (RTS) by relaxing CRS, therefore, with equations 7 to 10, convexity constraint N1 λ = 1 has to be applied in order to have pure technical efficiency (PTE) which refers to managerial efficiency and shows minimum level of inputs used avoiding wastage of inputs by management. Once, TE and PTE results are given, scale efficiency (SE) can be calculated as TE / PTE. Sale efficiency shows whether banks operates on the right scale or not but its results do not help find out which bank is operating at decreasing or increasing return to scale (Coelli 2005). For this purpose, NIRS v constraint: (N1 λ = 1) has to be included with equations 7 to 10. 3.1 Input and Output Variables There are a few approaches which guide to select input and output variables for banking efficiency analysis and it is upheaval task that according to which approach, these variables should be selected. Among them, production and intermediation approaches are most commonly used. According to production approach, banks are service providers to their users whereas as per intermediation approach, banks dominantly play financial intermediary role. Chen et al. (2005) explicate that the former approach is more appropriate for branch appraisal whereas for overall efficiency analysis of financial institutions, it is pertinent to apply latter approach. Efficiency analysis of overall banking sector of Pakistan is a subject matter of this study, therefore, following Miller and Noulas, 1996; Isik and Hassan, 2003; Das and Ghosh, 2006; Sufian, 2009, we opt intermediation approach for selection of inputs and outputs. Number of variables to be used for analysis is also equally important question. Utilization of more input and output variables may loss the power of DEA to discriminate between efficient and inefficient banks. Mostafa (2009) argues that the number of banks must be greater than three times to the number of selected variables. Considering these points, we have selected two outputs and three inputs which constitute five variables. Following Mostafa (2009) five variables based study should take more than 15 firms / banks. This study uses the data of 26 schedule banks operating in Pakistan, hence fulfill this condition. Another debatable point is the selection of deposits variable. Some researchers used this variable as input, considering this as interest earnings source where some selected this as output with mark as final product (Miller & Noulas, 1996). Following Miller & Noulas 1180

(1996); Hsiao et al. (2010), this study selected deposits as input variable. Finally, following intermediation approach, numbers of employees, operating fixed assets and deposits plus other accounts are selected as inputs whereas investment and loans are used as output variables. Table 1 depicts statistical summary of these variables. Variables Inputs Mean Table 1: Input and Output Variables ( Million Pak Rs.) Standard Deviation Minimum Maximum Median Labor* 4,205 5,020 55 18,625 1,682 Fixed assets 3,995.91 5,663.76 34.90 25,922.98 1,354.56 Deposits 114,520.59 161,930.57 232.66 726,464.83 37,259.90 Outputs Investment 34,072.91 48,999.81 15.37 217,642.82 11,019.00 Advances 80,802.60 109,044.26 57.62 475,243.43 31,569.96 Note: * Labor is defined as numbers of employees 3.2 Potential Determinants of Efficiency After ascertaining efficiency of banking sector, at second stage, we examine the determinants of efficiency through multivariate panel data analysis. In this case, dependent variable is efficiency having values 0 to 1. There is a general view in banking efficiency literature that Tobit model is more appropriate technique to manage the characteristics of distribution of efficiency and gives better results for policy prescriptions (Das and Ghosh, 2006). Following Das and Ghosh, 2006; Ariff and Can, 2008; Sufian, 2009), this study uses Tobit model for the following model: φ jt = β o + β 1 Dep jt + β 2 lnasst jt + β 3 NIE jt + β 4 Loans jt + β 5 NII jt + Β 6 Debt jt + β 7 GDP-Gr jt + β 8 Crisis jt + β 9 Largebanks jt + β 10 Smallbanks jt + ε jt (11 Where φ = Following, Damar (2006), this study uses TE, PTE & SE as dependent variables separately. These efficiency measures have been discussed earlier in details. Dep = Loan-net/ Deposits and accounts for market share of the banks. lnasst = Assets of banks in real form vi. NIE = Ratio of non-interest expenses to total assets, representing cost management of banks Loans = Ratios of net loans to total assets, accounts for liquidity position NII = Ratio of non-interest income to total assets is an indicator of diversifications of income Debt = Ratio of provision and bed debt written off directly to total assets and accounts asset quality GDP-Gr = Real GDP growth rate of Pakistan account for the impact of prevailing macroeconomic conditions on banking efficiency Crisis = Dummy = 1 for the year 2008 otherwise 0, representing recent global financial crisis 2008. 1181

Largebanks = Dummy =1 for the banks, possess assets more than Rs.250,000 (millions) Smallbanks = Dummy =1 for the banks, possess assets less than Rs.50,000 (millions) We expect positive signs for Dep, NII and Loans whereas signs of LnAsst, are priori undecided. Moreover, sign of NIE is expected to be negative. GDP-Gr variable is included to find out the impact of prevailing macroeconomic conditions on banking efficiency. The effect of this variable may vary from country to country; therefore, we are uncertain about its expected sign. Recent global financial crisis sparked after mid of 2007, got worst momentum in 2008 and affected all over the world, therefore, this study included Crisis variable to find its impact on banking sector of Pakistan. We can t predict its expected sign priori. For deep understanding about efficiency trend in banking sector of Pakistan, we further decomposed all banks into large, medium and small banks groups according to asset accumulations of banks. In order to avoid dummy variable trap, we only included two categories of banks; large and small banks and sign of these dummy variables are also priori undecided. This study used annual data of 26 schedule banks operating in Pakistan during the period from 2004 to 2009. The sample of 26 schedule banks consists of 16 private commercial, 3 public sector commercial, 3 specialized and 4 foreign banks. Banks are categorized into large, medium and small banks as per their assets accumulations. The banks possessed assets more than Pak. Rs.250,000 (in millions), between Pak Rs. 50,000 to 249,000 (in millions) and less than Pak Rs. 49,000 (in millions) are marked as large, medium and small banks respectively. Numbers of large, medium and small banks varied year by year throughout the analysis period. Data on Input and output variables are collected from Banking Statistics of Pakistan 2009, published by State Bank of Pakistan (SBP). For equation 11, data on banks related variables are also taken from Banking Statistics of Pakistan 2009 whereas data on real GDP growth rate is taken from annual reports of SBP. 4. Empirical Results 4.1 Efficiency The results of technical efficiency are presented in Table 2. Further decompositions of this efficiency pure technical and sale efficiency have also been exercised and their results are also given in this table. The first part of this table provides annual average efficiency score of all banks. This table reveals that average technical efficiency enhances in the initial period, however, deteriorated in the middle period. Once again in last two years of the selected period this efficiency continuously increases. The results of pure technical and scale efficiencies illuminate that scale inefficiency is the major cause of deterioration in technical efficiency during the year 2006 and 2007 whereas pure technical efficiency which is also generally called managerial efficiency almost stagnated during the analysis period. For deep understanding about efficiency in banking sector of Pakistan, efficiency trends of large, medium and small banks groups according to asset accumulations of banks are also presented in the same table. It is evident from this table that large banks are least technical efficient as compare to their other counterparts; however, technical efficiency of large bank groups only decreases in 2006 whereas in the next three years this efficiency increases continuously. This group is the most pure technical efficient whereas, also least scale efficient among the three groups of banks. The role of scale (in)efficiency dominated over technical efficiency of large banks. Efficiency results of medium and small banks explain that technical efficiency decreases during 2006 and 2007 as it is evident from the efficiency results of all banks but after that this efficiency constantly increase in these groups. Pure technical efficiency of medium banks continuously decreases throughout the analysis period whereas this efficiency of small banks increases during first two years, decreases in middle and again increases during the last period as is the case of all banks analysis. Average scale efficiency of these groups contains almost same trends. The results of return to scale (RTS), both in numbers and percentage, are presented in Table 3. This table depicts most of the banks operate at decreasing return to scale (DRS) during the years 2006 and 2007. These results are consistent with efficiency analysis results which also explain that efficiency of the banking sector decreased in these two years. It is also unveiled from this table that no any large banks could operate at increasing return to scale (IRS) throughout the analysis period and even not any single enjoyed constant return to scale(crs) from 2004 to 2008. This shows that all large banks are in trap of DRS throughout the period except in the year 2009 1182

where only one bank could attain CRS mark. Most of medium banks face diseconomies of scale and there is no any medium banks which can touch IRS hallmarks from 2004 to 2008. Only two banks operated at IRS while 60 percent banks were in the trap of DRS in 2009. Most of the small banks reap the benefits of economies of scale and even during the turmoil year 2006, as is evident from efficiency analysis, less than 50 percent banks operated at DRS. In 2009, two third small banks operated at IRS and remaining one third small banks were at CRS. All this analysis illuminate that small banks are most scale efficient whereas large banks are the most sale inefficient banks in Pakistan banking sector. All Banks Table 2: Average Efficiency Results Nos. of TE 1 PTE 1 SE 1 Banks Mean S.D. 2 Mean S.D. Mean S.D. 2004 26 0.808 0.185 0.908 0.177 0.896 0.129 2005 26 0.828 0.176 0.910 0.134 0.911 0.134 2006 26 0.697 0.191 0.903 0.145 0.774 0.164 2007 26 0.692 0.192 0.885 0.157 0.788 0.171 2008 26 0.804 0.178 0.900 0.138 0.899 0.156 2009 26 0.851 0.165 0.904 0.132 0.944 0.121 Large Banks 2004 4 0.654 0.023 0.982 0.037 0.667 0.014 2005 5 0.741 0.055 0.965 0.076 0.769 0.043 2006 6 0.583 0.044 0.920 0.094 0.636 0.046 2007 6 0.625 0.059 0.965 0.053 0.647 0.045 2008 6 0.843 0.077 0.972 0.053 0.867 0.057 2009 7 0.847 0.106 0.919 0.108 0.922 0.046 Medium Banks 2004 7 0.872 0.135 0.987 0.035 0.884 0.133 2005 7 0.867 0.144 0.912 0.096 0.947 0.074 2006 7 0.729 0.144 0.925 0.118 0.787 0.107 2007 9 0.681 0.155 0.878 0.123 0.777 0.133 2008 10 0.802 0.143 0.855 0.146 0.943 0.093 2009 10 0.880 0.171 0.887 0.169 0.991 0.012 Small Banks 2004 15 0.819 0.211 0.851 0.218 0.963 0.048 2005 14 0.839 0.212 0.888 0.164 0.944 0.148 2006 13 0.733 0.237 0.884 0.180 0.831 0.190 2007 11 0.738 0.257 0.847 0.206 0.875 0.195 2008 10 0.782 0.250 0.902 0.156 0.873 0.231 2009 9 0.823 0.204 0.910 0.111 0.908 0.198 Note: 1 TE, PTE & SE represent technical, pure technical efficiency. 2 SD= standard deviation 1183

Years Table 3: Return to Scale Analysis All Banks Large Banks Medium Banks Small Banks IRS CRS DRS IRS CRS DRS IRS CRS DRS IRS CRS DRS 2004 Nos. 5 8 13 0 0 4 0 3 4 5 5 5 % 19.2 30.8 50.0 0.0 0.0 100.0 0.0 42.9 57.1 33.3 19.2 19.2 2005 Nos. 5 10 11 0 0 5 0 3 4 5 7 2 % 19.2 38.5 42.3 0.0 0.0 100.0 0.0 42.9 57.1 35.7 50.0 14.3 2006 Nos. 3 5 18 0 0 6 0 1 6 3 4 6 % 11.5 19.2 69.2 0.0 0.0 100.0 0.0 14.3 85.7 23.1 30.8 46.2 2007 Nos. 3 5 18 0 0 6 0 1 8 3 4 4 % 11.5 19.2 69.2 0.0 0.0 100.0 0.0 11.1 88.9 27.3 36.4 36.4 2008 Nos. 6 6 14 0 0 6 2 2 6 4 4 2 % 23.1 23.1 53.8 0.0 0.0 100.0 20.0 20.0 60.0 40.0 40.0 20.0 2009 Nos. 10 9 7 0 1 6 4 5 1 6 3 0 % 38.5 34.6 26.9 0.0 14.3 85.7 40.0 50.0 10.0 66.7 33.3 0.0 Table 4: TOBIT Model Results Regarding Determinants of Efficiency Variable TE PTE SE Constant 0.782774 0.279091 1.491152* Dep 0.001683** -0.001519** 0.003210* LnAST -0.002548 0.042084** -0.044104*** NIE -0.021343* -0.017769* -0.005367 Loan 0.000138** -1.98E-05 0.000164* NII 0.042224* 0.024793* 0.019576** Debt -0.010685-0.006922-0.004371 GDP-Gr -0.015702** -0.000974-0.016124* Crisis -0.021522-0.001993-0.02287 Largebanks -0.085225 0.009402-0.094512** Smallbanks 0.036917 0.042077-0.002884 Note: *, **, & *** represent significant at 1%,5% and 10% level respectively. 4.2 Determinants of efficiency The findings of factors which have potential to influence efficiency of banks are vital for policy guidelines in order to enhance the performance of banking sector. Considering this, we also examine determinants of banking efficiency and the results of this analysis are presented at Table 4. This table explains that market share with respect to deposits of banks positively contribute to technical efficiency. Dacanay III (2007) also found same results in his analysis. Bad cost management make adverse impact on technical efficiency as the NIE proxy of the cost management, contain negative sign and also significant at 1 percent level. Issuance of loans, which is one the key source of income for banks, enhances technical and scale efficiency of banks. This finding is consistent 1184

with the results of Sufian (2009). Non-interest income variable is used to assess the impact of income diversifications on the performance of banking sector. It reveals that income diversifications significantly enhance technical, pure technical and scale efficiency. Significant positive impact of this variable on all three efficiency variables illuminates the importance of income diversification in banking sector and dictate that instead of sole reliance on interest income, diversification in income would be better strategy for banks because this diversification not only enhances efficiency of this sector but also beneficial for financial intermediary role of banks. Prevailing economic condition is also crucial for efficiency of banking sector of any country. In recent years, macroeconomic condition of this country was not remarkable. Therefore, to envisage the impact of current economic situation on the performance of banking sector, we include GDP growth rate as an explanatory variable in the model and the results illumine that recent deteriorated economic growth exerted hampering impact on technical as well scale efficiencies. The coefficient of this variable contains negative signs for both these two efficiency measures and also significant at 5 percent and 1 percent significant levels respectively. It means current unhealthy economic progress postulate negative impact on banking efficiency, therefore, to improve the efficiency, economic stability should also be taken into consideration. A qualitative variable for global financial crisis 2008 is included in the model. The results explain that Pakistan banking sector safely escaped from its negative impact. Foreign banks in host countries may be one of the important causes for transmitting negative contagious affect of developed countries financial crisis to developing countries through capital in or out flow. Foreign banks have tiny share in Pakistan banking sector because branches of these banks have merely 1 percent share in total bank branches network vii. Moreover, major banking business of foreign banks is to provide financial facilitation for foreign trade. Therefore, due to these reasons efficiency of Pakistan banking sector could safely escaped from the negative signals of recent global financial crisis. This very fact can also be confirmed from efficiency analysis where banking efficiency improved from 2007 to 2008 and this positive pace remain continued in 2009. Large banks are negatively associated with scale efficiency because its coefficient is significant at 5 percent level and contains negative sign. This negative impact indirectly transmits hampering impact on technical efficiency. This finding is also consistent with the finding of scale efficiency analysis of large banks group where this group is observed as most scale inefficient as compare to medium and small banks groups. 5. Conclusions This study analyzes efficiency of Pakistan banking sector in post-banking reforms era by taking data of individual banks from 2004 to 2009. It has been found that technical efficiency decreases in middle, however, increasing trend prevails in last period of data. Small banks are the most technical and sale efficient. Efficiency, return to scale and efficiency determinants analyses illuminates that large banks are least efficient with respect to scale operations. Return to scale analyses elucidate that small banks enjoy economies of scale as compare to medium and large banks. Efficiency determinants analysis explains that market share with respect to deposits, income diversifications, and issuances of loans positively enhance efficiency. On the other hand current unhealthy economic conditions of Pakistan and bad costs management of banking sector exert negative impact on efficiency of this sector. 1185

References Ariff, M. & Can, L. (2008). Cost and profit efficiency of Chinese banks: A non-parametric analysis. China Economic Review, 19, 260-273. Ataullah, A. & Le, H. (2006). Economic reforms and bank efficiency in developing countries: the case of Indian banking industry. Applied Financial Economics, 16, 653-663. Banker, R. D., Charnes, A. & Cooper, W. W. (1984). Some models for estimating technical and sale inefficiencies in data envelopment analysis. Management Science, 30 (9), 1078-1092. Berger, A. N. & Mester, L. J. (1997). Inside the black box: What explains differences in the efficiencies of financial institutions? Journal of Banking and Finance, 21, 895-947. Charnes, A., Cooper, W.W. & Rhodes, E. (1978). Measuring the efficiency of decision making units. European Journal of Operational Research, 2, 429-444. Chen, X., Skully, M. and Brown, Kym (2005). Banking efficiency in China: Application of DEA to pre- and postderegulation eras: 1993-2000. China Economic Review 1, 229-245. Coelli, T. J. (1996). A guide to DEAP Version 2.1: A data envelopment analysis (computer) program. CEPA working paper No.8/96, Australia: Department of Econometrics, University of New England. Coelli, T. J., Rao, D. S. P, O Donnel, C. J. & Battese, G. E. (2005). An introduction to efficiency and productivity analysis (2 nd Edition). USA: Springer Science + Business Media, Inc. Damar, H. E. (2006). The effects of shared ATM networks on the efficiency of Turkish banks. Applied Economics, 38, 683-397. Dacanay II, S. J. O. (2007). Malmquist index and technical efficiency of Philippine commercial banks in the post- Asian financial crisis period. Philippine Management Review, 14, 93-114. Das, A. & Ghosh, S. (2006). Financial deregulation and efficiency: An empirical analysis of Indian banks during the post reforms period. Review of Financial Economics, 15, 193-221. Farrell, M. J. (1957). The measurement of productive efficiency. Journal of the Royal Statistical Society. Series A (General), 120 (3), 253-290. Government of Pakistan (2009). Economic survey of Pakistan 2008-2009. Islamabad: Ministry of Finance. Hardy, D. C. & Patti, E. B. (2001). Banking reforms and banking efficiency in Pakistan. IMF Working Paper: WP/01/138. Hsiao, H., Chang, H., Cianci. A. M. & Huang, L. (2010). First financial restructuring and operating efficiency: Evidence from Taiwanese commercial banks. Journal of Banking & Finance, 34, 1461-1471 Hao, J., Hunter, W. C. & Yang, W. K. (2001). Deregulation and efficiency: the case of private Korean banks. Journal of Economics and Business, 53, 237-254. Hussain, I. (1999). Pakistan: The Economy of an Elitist State. Karachi: Oxford University Press. 1186

Isik, I. & Hassan, M. K. (2002).Technical scale and allocative efficiencies of Turkish banking industry. Journal of Banking & Finance, 26, 719-766. Isik, I. & Hassan, M. K. (2003). Financial deregulation and total factor productivity change: An empirical study of Turkish commercial banks. Journal of Banking and Finance, 27, 1455-1485. Khan, M. Z. (1996) Pakistan: prospects for private capital flows and financial sector development. The Pakistan Development Review 35 (4), 853-883. Khan, M. A. and S. Khan (2007) Financial sector restructuring in Pakistan. Lahore Journal of Economics (Special Edition), 97-125. Limi, A. (2004) Banking sector reforms in Pakistan: economies of scale and scope, and cost complementarities. Journal of Asian Economics 15, 507-528. Llanto, G. M. & Badiola, A. R. (2009). The impact of global financial crisis on rural and microfinance in Asia. Discussion Paper Series No.2009-24. Philippine Institute for Development Studies. Meenai, S. A. (2010) Money and Banking in Pakistan. (6 th Edition). Karachi: Oxford University Press, Pakistan (Revised and expanded by Javed A. Ansari). Miller, M. S. & Noulas, A. G. (1996). The technical efficiency of large bank production. Journal of Banking & Finance, 20, 495-509. Mostafa, M. M. (2009). Modeling the efficiency of top Arab banks: A DEA-neural network approach. Expert System with Application, 36, 309-320. State Bank of Pakistan (2003). Financial sector assessment 1990-2000. Karachi: SBP Staub, R. B., Souza, G. S. S. & Tabak, B. M. (2010). Evolution of bank efficiency in Brazil: A DEA approach. European Journal of Operational Research, 202, 204-213. Sufian, F. (2009), Determinants of bank efficiency during unstable macroeconomic environment: Empirical evidence from Malaysia. Research In International Business and Finance, 23, 54-77. Zaidi, S. A.(2005) Issues in Pakistan s Economy. (2 nd Edition). Karachi: Oxford University Press. Endnote i For details see Khan & Khan (2007). ii Both Parametric and non-parametric techniques have been used in finance literature in order to assess efficiency of banking sector. iii Both are public sector banks. iv In efficiency analysis literature these are called decision making units (DMUs). In our case, firms represent banks. v Non increasing return to scale vi GDP Deflator with base year 1999-00 is used to make nominal variable into real form. vii In 2008, out of more than 7700 branches, numbers of foreign bank branches was only 64. Almost same situation prevailed throughout the history of Pakistan banking industry. 1187