Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 Efficiency of Kuwaiti Banks: A Comprehensive X-Efficiency Framework Jamal Al-Khasawneh, Khiyar Abdullah Khiyar and Mohammed Z. Shariff This study tracks the nonparametric X-efficiency based measures and compares the efficiency dynamics of conventional and Islamic Kuwaiti banks over the period 1996-2010. The findings show that, in terms of all efficiency measures, Kuwaiti banks kept losing efficiency: a trend that started in 2004 and worsened after the 2008 economic crisis. However, two efficiency stages can be noted: The first trend is indicated for the period 1996-2003, during which conventional banks had clear efficiency superiority over Islamic banks. The second trend, from 2004 until 2010, shows Islamic banks having superiority over conventional banks. This superiority of Islamic banks was mainly due to the loss of efficiency in conventional banks and not to any gained efficiency in Islamic banks. 1. Introduction The Kuwaiti banking system witnessed structural economic shocks during the last two decades. The Iraqi invasion of Kuwait in 1990 and the consequent exponential increase in the defense bill in the 1990s, combined with relatively low oil revenues, made the Kuwaiti economy go on a slow-down phase. However, the government‟s unlimited support in the aftermath of the first Gulf War was able to re-establish confidence in the financial system, and banks started showing healthy operating signs and positive profitability four years after the liberation of Kuwait. By 2003, luck had coincided with luck; clouds totally cleared when national security concerns became a thing of the past after the collapse of the hostile regime in Iraq and the record increase in oil prices. The period of 20032008 represents a booming modern era in the economic history of Kuwait and the Gulf Cooperation Council (GCC) countries. According to the 2004 Financial Sector Assessment Report, issued by the World Bank, Kuwait had made an impressive recovery from the economic damage caused by the 1990-91 war; oil production and exports had been restored, and fiscal and external current account balances since the mid-1990s had reached surpluses to the order of 20% of GDP. The World Bank‟s report also indicated that by the end of 2002, the total assets of financial institutions amounted to more than 200% of nominal GDP, while stock market capitalization reached about 100%. ______________________________________ Jamal Al-Khasawneh, Department of Economics and Finance, Gulf University for Science and Technology (GUST), Email: Alkhasawneh.j@gust.edu.kw, Khiyar Abdullah Khiyar, Department of Economics and Finance, Gulf University for Science and Technology (GUST), Email: Abdalla.K@gust.edu.kw, Mohammed Z. Shariff, Department of Economics and Finance, Gulf University for Science and Technology (GUST), Email: shariff.m@gust.edu.kw, Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 The institutional structure of the banking and financial system in the state of Kuwait consists of units subject to the Central Bank of Kuwait‟s supervision and oversight. These units consist of local banks (commercial, specialized and Islamic), investment companies (conventional and Islamic), exchange companies, and investment funds. The term “local banks”, as used by the Central Bank of Kuwait (CBK), encompasses the group of fourteen commercial banks registered with it. Of these, eleven are Kuwaiti banks and the rest are branches of foreign banks. By the end of 2010, of the eleven Kuwaiti banks, five were conventional commercial banks, five were Islamic commercial banks, and one was specialized, the Industrial Bank of Kuwait. Table (1) shows the number of financial institutions and their classifications. However, Al-Hassan el al. indicate that the banking industry is highly concentrated, with the top three banks having 63% of the market share, and the top five banks having 81%. They also indicate that the banking sector is largely domestically owned, due to entry barriers and licensing restrictions for foreign banks, including GCC banks. Table 1: Number of Kuwaiti Financial Institutions and their Classifications Type Total No. Conventional Islamic Commercial Banks Investment Companies Investment Funds Total 10 100 112 222 5 46 57 108 5 54 55 114 The latest shock came from outside Kuwait‟s borders; this time the 2008 worldwide economic crisis exposed the Kuwaiti economy as well as the banking system to unexpected potential losses. Given the unique economic history of the state of Kuwait, this paper examines cost, revenue, and profit in the state of Kuwait during the period 1996-2010. Data Envelopment Analysis (DEA) is used to compute absolute efficiencies, which can be used to characterize the efficieny changes in Kuwaiti banks. We believe that the study of the dynamic changes in efficiencies, after critical political and economic shocks in this country, is missing in the literature. Furthermore, there is a unique phenomenon in Kuwait where some conventional banks conventional banks where some conventional banks did convert and others are in the process of conversion into Islamic banks 2. Literrature Review In this section, we present some of the relevant literature related to the GCC region and Middle East and North Africa countries MENA . Darrat et al. (2003) found that Kuwaiti banks fail to optimally utilize a significant proportion of their resources. The sources of bank inefficiency appear to be both allocative (regulatory) and technical (managerial) in nature. The results also indicate that larger banks in Kuwait are less efficient than smaller ones, although all banks have improved their efficiency levels and experienced some gains in productivity, indicating that the loss of efficiency is due mainly to pricing decisions. Hassan et Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 al. (2004) investigated the efficiency of the Bahraini banking sector. Their findings indicate that the cost inefficiency is due mainly to technical inefficiency rather than allocative inefficiency, while the major source of the total technical inefficiency for Bahraini banks is pure technical inefficiency (input waste) and not scale inefficiency. Maghyereh (2004) investigated the effect of financial liberalization on Jordanian commercial banks‟ efficiency. The findings indicate high technical efficiency across Jordanian banks and that larger banks have lower scale efficiency and higher pure technical efficiency than small and medium banks. Jordanian banks were investigated further by Bdour et al. (2008), who indicated that, with minor exceptions, the X-efficiency scores of most of the banks showed a consistent increase after the implementation of the liberalization policy. Ariss (2008) investigated the evolution of bank efficiency in the Lebanon subsequent to a period of deregulation liberalization. The findings indicated positive cost-efficiency improvement with higher efficiency improvement in consolidating banks. Gattoufi et al. (2009) analyzed the performance of Oman's banking sector, using panel data for the period of 2000-2005. The findings show that Omani banks have high pure technical efficiency and that local commercial banks outperform their foreign counterparts. Srairi (2009) indicated that banks in the Gulf region are relatively more efficient at generating profits than at controlling costs, and that conventional banks are more efficient than Islamic banks. Furthermore, the findings demonstrate a consistent increase in profit efficiency since 2000, while showing decreasing cost efficiency since 2004. Al-Delaim et al. (2006) employed DEA to measure the relative cost efficiency of 24 Islamic banks from 14 countries. Overall findings indicate that most Islamic banking institutions are efficient and the rest are on the way to improving their efficiency. Kabir (2006) investigated the efficiency of Islamic banks around the world. The findings indicated that, on average, Islamic banks are less efficient than conventional banks in terms of all efficiency measures. Using DEA, Fadzlan et al. (2007) analyzed the comparative performance of foreign and domestic Islamic banks in Malaysia. The authors found that foreign banks exhibit higher technical efficiency compared to domestic Islamic banks. Furthermore, the results show that technically more efficient banks are larger in size, have greater loans intensity, and on average have less non-performing loans. In a study carried out by Kamaruddin et al. (2008), the cost and profit efficiency of fullyfledged Islamic banks and the Islamic window operations of domestic and foreign banks in Malaysia were analyzed using DEA. The findings indicated that Islamic windows outperform Islamic banks in the term of cost efficiency and profitability. Ariss (2010) indicated that Islamic banks are more efficient in terms of financing activities but still less competitive (they have a small market share) than conventional banks. In another study carried out by Abdul-Majid et al. (2010), it was found that Islamic banks in different countries had no advantage over other conventional banks, as Islamic banks suffered poor allocative efficiency, which is a determinant or subcomponent of cost efficiency. Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 3. Methodology 3.1. The Non-parametric Data Envelopment Analysis Charnes, Cooper, and Rhodes (1978) coined the term „data envelopment analysis‟ (DEA). There has since been a multitude of works that have applied and extended the DEA methodology. DEA constructs a frontier based on the sample data rather than using an assumed production function. This nonparametric approach shows how a particular decision making unit (DMU) operates relative to other DMUs by providing a benchmark for the best practice technology based on the DMUs in the sample. Because DEA makes no assumptions about inefficiency distributions, it is subject to data problems and inaccuracies created by accounting rules (Isik, 2000). However, DEA works better than the parametric approach when the sample size is small. Following Rangan et al. (1988), Berger et al. (1992), Elyasiani and Mehdian (1992), Fare et al. (1994), Grabowaski et al. (1993), Leightner and Lovell (1998), Wheelcock and Wilson (1995), Isik and Kabir (2003), and others, DEA is used in this paper to measure U.S. banks‟ efficiency scores. This choice is motivated by the small sample size during some years of our data set. Some other reasons for this choice are: 1) most studies that have used both Stochastic Frontier Approach (SFA) and DEA have found that both approaches preserve the efficiency ranking of the DMUs (see Isik and Kabir, 2002, 2003; and Al-Sharkas et al. (2008). Since the purpose of this paper is to use the efficiency scores to rank merging banks according to their efficiency characteristics, we use DEA rather than SFA; 2) the non-parametric DEA is the better choice when the industry has experienced a series of reforms and/or shocks because we can assume variable returns to scale (which is not an option in SFA); and finally, and most importantly, 3) under DEA, profit efficiency scores can be broken down into more basic components (cost efficiency, revenue efficiency, etc.). Farrell (1957) decomposes the overall cost efficiency (CE) of each DMU into two components: a) technical efficiency (TE), which shows the ability of a DMU to achieve the maximum output for a given production set; and b) allocative efficiency (AE), which shows management‟s ability to construct an optimal product mix, given their respective prices which is related mainly to the regulatory environment or macroeconomic conditions (Lovell, 1993). Assuming constant returns to scale (CRS), CE is decomposed as follows: CE = TE AE (1) Banker et al. (1984) proposed the variable returns to scale frontier (VRS), in which the frontier changes over time due to technological progress, financial crises, higher industry concentration due to mergers and acquisitions, and financial deregulation (Isik and Kabir, 2003). However, Banker et al. further decompose TE into two components: a) pure technical efficiency (PTE), which indicates the proportional reduction in input usage if inputs are not wasted; and b) scale efficiency (SE), which represents the proportional output reduction if the bank achieves CRS. So, equation 1a can be re-written as follows: Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 CE = PTE SE AE (2) As shown previously, cost efficiency can be estimated by summing input prices rather than output quantities. Consider n DMUs, where each DMU uses m inputs to produce s outputs. The general form of the cost minimization problem is then: m min p i x i* i 1 s.t. n j 1 j x ij x i* (3) n j 1 i 1,2,........, m; j y rj y r r 1,2,........, s; j , x i* 0 j 1 Assuming (VRS). where pi is a vector of input prices for the j-th DMU and xi* is the cost minimization vector of input quantities for the j-th DMU, given the input prices and the output levels. The first constraint places a restriction on the input side, requiring the use of inputs in a linear combination at the efficient frontier to be less than or equal to the use of the inputs by the i-th bank. The second constraint shows that the observed outputs of DMUj must be less than or equal to a linear combination of outputs, xi* , of the DMUs forming the efficient frontier. The third constraint assures the feasibility of the solution. The fourth constraint imposes the VRS assumption. The only way to derive a more cost efficient DMU is by getting it closer to the efficient frontier. This can be achieved by using input equal to X rather than X 1 , holding the output fixed (the bold horizontal arrow shows this choice). Finally, the cost efficiency of the each DMU can be obtained as follows: m p x i 1 m i * i p x i 1 i M inimum virtual cost 1, Observed cost (4) i where the cost efficiency value will be equal to one for the DMUs that lie on the efficient frontier. The cost efficiency scores take values in the range (0,1). We can also obtain the allocative efficiency using equation (1a) as follows: CE CE TE AE AE TE Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 3.2. Estimation of Revenue Efficiency Using the same considerations as in the previous section, we can obtain the revenue efficiency (RE) scores for each DMU. The revenue maximization problem maximizes the vector of output quantities, y * , in the first step. Then, the revenue-maximizing problem is calculated as follows: s max q r y r* r 1 s.t. n j 1 j x ij x i j y rj y (5) n j 1 i 1,2,........, m; r 1,2,........, s; * r j , y i* 0 n j 1 j 1 Assuming (VRS). where q r is a vector of output prices for the j-th DMU, and y r* is the maximization vector of output quantities of the DMUs forming the efficient frontier. The first constraint indicates that the use of the inputs in a linear combination of efficient DMUs must be less than or equal to the use of inputs of the j-th DMU. The second constraint shows that the observed outputs of the j-th DMU must be less than or equal to the linear combination of the DMUs forming the efficient frontier. The last two constraints are well defined in the previous section. After solving the above problem, we can obtain RE as follows: s RE q r 1 s q r 1 r yr (6) r y * s s where q r 1 r y r is the observed/actual revenue of the DMU, and q r 1 r y * is the virtual efficiency profit that could be achieved if the DMU were situated on the efficient frontier. The value of the profit efficiency scores will always fall in the range (0,1). 3.3. Estimation of Profit Efficiency Summing the cost and revenue efficiencies generates the profit efficiency (PE) concept, which seeks to minimize costs and maximize revenue simultaneously. Unlike cost and revenue efficiencies, PE is obtained by allowing inputs and outputs to vary. The profit maximization problem can be described as follows: Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 s m r 1 i 1 max q r y r* pi xi* s.t. n j 1 j xij xi* i 1,2,........, m; j y rj y r* r 1,2,........, s; n j 1 (7) xi* xi , y r* y r j 0 n j 1 j 1 Assuming (VRS). where the first constraint indicates that the use of the inputs in a linear combination of efficient DMUs must be less than or equal to the use of inputs of the j-th DMU. The second constraint shows that the observed outputs of the j-th DMU must be less than or equal to the linear combination of the DMUs forming the efficient frontier. However, the two constraints in this problem are solved simultaneously. The third constraint is imposed to assure that the revenue maximization and cost minimization are both achieved. This constraint requires that the inputs of the j-th DMU must be greater than or equal to the output of the DMUs on the efficient frontier, and it indicates that the output of the j-th DMU must be less than or equal to the outputs of the DMUs on the efficient frontier. This constraint is important because it is possible to maximize profit efficiency by minimizing costs only. In this case, profit maximization will be equivalent to cost minimization. The same argument is valid for the revenue efficiency. Finally, the profit efficiency can be obtained using the following equation: s q r 1 s q r 1 m s where q r 1 r m r y r p i xi r y pi x i 1 m * r i 1 , (8) * i y r p i x i represents the observed profitability of DMUi. This value i 1 could be negative for DMUs with losses. s m r 1 i 1 q r y r* pi xi* , on the other hand, represents the virtual profitability that could be achieved if the DMU is located on the efficient frontier. Accordingly, the profit efficiency values must lie in the range ( ,1) . Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 4. Data, Sample Characteristics, and Variable Defination DEA needs a set of inputs and outputs in order to measure efficiency, and therefore, relative productivity. Accordingly, we model commercial banks as multi-product firms, producing two outputs and employing two inputs. All variables are measured in millions of U.S. dollars except prices, which are measured as ratios. The outputs include (1) net loans; and (2) other earning assets, which consist of loans to special sectors, interbank loans, and investment securities (treasuries and other securities). All output prices are estimated as proxies. These are calculated as follows: 1) the price of loans is defined as total interest income to net loans; and 2) the price of other operating income is defined as other operating income to other earning assets weighted by the proportion of other earning assets over the total of other earning assets plus off-balance-sheet items. Inputs include (1) personnel expenses; and (2) loanable funds, which are defined as the sum of demand and time deposits and non-deposit funds as of the end of the respective year. Also, input prices are estimated as proxies. The price of labor is calculated as personnel expenses over total assets. The price of capital is calculated as non-interest expense over total assets. Finally, the price of funds is calculated as total interest expense over loanable funds. To examine the X-efficiency scores of Kuwaiti banks, we manually gathered data for the years 1996-2010; all accounting data were obtained from the annual reports of each bank as posted on each single bank‟s website. Furthermore, we compared these data to the data posted by the Kuwaiti Central Bank, where we found a full match between the two sources. Tables (2) and (3) show the summary statistics for the total input and output variables of sample banks. Table (2) shows the absolute dollar value of input variables. To make the comparison easier, we calculated the percentage increase of each variable as of 1996 and 2010. The results are shown in panel B of Table (2). Table (2) shows that personnel expenses in 2010 increased by 0.766% from those in 1996, while its price increased very marginally by 0.26%. More importantly, deposits increased 963% with a decrease in interest paid of 0.79%. Table (3), which represents outputs and their prices, shows that loans increased by 622%, which is basically less than the percentage increase of deposits, where it accounts only for two thirds of deposits. These idle deposits could be a possible source of inefficiency for Kuwaiti banks if no other investment avenues are used. However, interest earned decreased by almost 1% from its 2010 figure. Comparing the general downward interest trend, it can be said that the interest margin decreased by 0.2%. Accordingly, Kuwaiti banks increased their other earning assets by 450%; those earnings show an encouraging increasing return of around 0.75%. Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 5. Revenue, Cost and Profit Efficiencies Results 5.1. Revenue Efficiency Results Table (4) represents the average revenue scores and their descriptions: conventional, Islamic, and the whole sample. Table (4) shows that revenue efficiency was in a state of consistent increase for the period 1996-2004, with a clear advantage for conventional banks, whose efficiency scores were 5% to 10% higher than those of Islamic banks. Starting from 2005, efficiency scores converted backwards with all Kuwaiti banks‟ inefficiencies increasing significantly with a marginal superiority given to Islamic banks. Comparing the efficiency scores of 2002 with those of 2009, we can indicate -9%, -20%, and 11% efficiency changes for all, conventional, and Islamic banks respectively. Year 2010 looks totally different, and we can indicate 11%, 45%, and -32% efficiency changes from 2009 for all, conventional, and Islamic banks respectively. Such a significant efficiency change can be explained by three possible reasons: Firstly, Islamic banks usually invest in longer-term investments than conventional banks, especially in real estate. Accordingly, it is expected that Islamic banks‟ losses may take a longer time to be actualized/re-evaluated. The second reason is the number of Islamic banks which have only recently converted to Islamic banking and, therefore, have limited experience in Islamic banks‟ investment opportunity sets and, consequently, a limited ability to operate successful Islamic funds. Thirdly, the Kuwaiti government supported the banking system after the economic crisis by issuing government bonds and treasury bills, a choice Islamic banks could not accept. 5.2. Cost Efficiency Results Table (5) represents the cost-efficiency scores of Kuwaiti banks. We can see that there was a general trend of efficiency loss over time for all Kuwaiti banks, with two distinctive stages to be noted: 1997-2003 and 2004-2010. In the first stage, conventional banks achieved a higher efficiency score than Islamic banks. In the second stage, however, we can see that, with the exception of 2007, Islamic banks outperformed conventional banks. This superiority of Islamic banks was mainly due to the loss of efficiency in conventional banks and not to any gained efficiency in Islamic banks. 5.3. Profit Efficiency Results Table (6) represents the profit efficiency scores of our sample banks. Profit efficiency sums up both revenue and cost efficiencies. We can still see the previously indicated trends in the last two sections. The period of 1996-2003 shows conventional banks having a clear superiority over Islamic banks; they scored about 85%, while Islamic banks achieved 60%. For 2004-2010, the whole scene looks different; Islamic banks achieved higher average efficiency scores of Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 73%, with conventional banks scoring 57%. Overall, the profit efficiency trend of Islamic banks looks less volatile than that of conventional banks. 6. Conclusions Given the uniqueness of the Kuwaiti banking system, and the economic history of the state of Kuwait, this paper examines cost, revenue, and profit in the state of Kuwait during the period 1996-2010. Data Envelopment Analysis (DEA) is used to compute absolute efficiencies. The results of all efficiency measures show that there was a general trend of efficiency loss over time for all Kuwaiti banks, with two distinctive stages to be noted: 1997-2003 and 2004-2010. In 1997-2003, conventional banks maintained a higher efficiency score than Islamic banks. The second stage witnessed serious inefficiencies increase of conventional banks and a slight efficiencies increase of Islamic banks, putting Islamic banks on the lead for the latest stage. Overall, Islamic banks‟ cost, revenue, and profit efficiency performance look more stable than that of conventional banks. Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 Table 2: Banks’ Inputs and Inputs’ Prices 1996 Inputs Average(Whole Sample) Max Min Stdev Skewness Kurtosis Input Prices 2003 2010 Growth 2010-1996 Per Exp Tdep Per Exp Tdep Per Exp Tdep Per Exp Tdep 10.32 47.2 4.2 14.05 2.83 8.24 928.11 3275.9 141.6 947.16 2.24 5.82 17.6 31.5 9.3 9 0.7 -1.04 1833.65 3222.6 935.2 876.38 0.63 -0.26 89.38 285.33 12.95 38.57 1.55 1.18 9867.33 26157 177.18 3920.39 1.14 0 7.66 5.05 2.08 9.63 6.98 0.25 0.26 -0.79 1996 Per Exp Tdep Average(Whole Sample) 0.006 0.083 Max 0.011 0.147 Min 0.006 0.072 Stdev 0.002 0.034 Skewness 2.71 1.02 Kurtosis 7.707 0.424 Per Exp: Personnel expenses, and Tdep: Total deposits. 2003 Per Exp 0.004 0.006 0.003 0.001 -0.055 -1.726 Tdep 0.035 0.079 0.02 0.022 2.188 4.954 2010 Per Exp 0.007 0.009 0.005 0.001 0.137 -0.94 Tdep 0.018 0.038 0.004 0.009 1.022 2.377 Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 Table 3: Banks’ Outputs and Outputs’ Prices 1996 Outputs Average(Whole Sample) Max Min Stdev Skewness Kurtosis NetL 710.14 3649.2 183.2 1130.11 2.74 7.76 NetL 1318.08 2478.4 445.8 741.07 0.5 -0.18 1996 Output Prices Average(Whole Sample) Max Min Stdev Skewness Kurtosis OEA 1262.3 6611 182 2026.95 2.88 8.47 2003 NetL 0.128 1.974 0.145 0.596 2.972 8.876 OEA 0.018 0.055 0.013 0.015 2.296 5.836 NetL: Net Loans, OEA: Other Earning Assets. OEA 5120.875 14051 2014.25 1902.18 1.78 2.78 2010 NetL 6937.225 19633.33 826.575 2471.3 1.29 1.12 2003 NetL 0.152 0.54 0.062 0.19 2.437 5.952 OEA 0.019 0.024 0.012 0.005 -0.325 -1.862 OEA 2643.7 8944.15 195 1230.34 1.52 1.23 2010 NetL 0.065 0.093 0.051 0.014 1.572 1.727 OEA 0.071 0.354 0.003 0.115 2.273 5.213 Growth 2010-1996 NetL 6.21 OEA 4.50 Growth 2010-1996 NetL -0.977 OEA 0.747 Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 Table 4: The Nonparametric Revenue Efficiency of Kuwaiti Banks (1996-2010). Average(Whole Sample) Max Min Stdev Skewness Kurtosis Average(Non Islamic Banks) Max Min Stdev Skewness Kurtosis Average(Islamic Banks) Max Min Stdev 1996 0.70 1.00 0.53 0.20 1.04 -0.14 1997 0.76 1.00 0.63 0.13 1.34 1.93 1998 0.79 1.00 0.67 0.14 1.05 -1.00 1999 0.82 1.00 0.70 0.13 0.85 -1.40 2000 0.82 1.00 0.72 0.11 0.99 -0.32 2001 0.84 1.00 0.72 0.10 0.34 -1.31 2002 0.87 1.00 0.68 0.14 -0.60 -1.68 2003 0.92 1.00 0.76 0.12 -0.98 -1.81 2004 0.71 1.00 0.41 0.23 -0.16 -2.06 2005 0.62 0.84 0.49 0.13 0.72 -1.00 2006 0.66 1.00 0.46 0.24 0.23 -1.12 2007 0.72 1.00 0.27 0.25 -0.42 -0.48 2008 0.63 1.00 0.26 0.26 0.03 -1.12 2009 0.61 1.00 0.33 0.26 0.26 -1.68 2010 0.82 1.00 0.34 0.24 -1.30 0.66 0.70 1.00 0.53 0.20 1.04 -0.14 0.77 1.00 0.63 0.13 1.11 1.55 0.67 0.80 1.00 0.67 0.16 0.90 -1.89 0.76 0.83 1.00 0.70 0.14 0.57 -2.12 0.71 0.83 1.00 0.72 0.11 0.74 -0.85 0.73 0.87 1.00 0.75 0.09 0.23 -1.45 0.72 0.90 1.00 0.68 0.13 -1.30 0.91 0.71 0.96 1.00 0.78 0.10 -2.24 5.00 0.76 0.69 1.00 0.41 0.25 0.15 -2.47 0.82 0.65 0.84 0.50 0.14 0.29 -1.95 0.53 0.58 0.46 0.13 0.67 1.00 0.52 0.26 0.22 -1.22 0.64 0.82 0.46 0.25 0.70 1.00 0.27 0.28 -0.40 -0.21 0.77 1.00 0.51 0.24 0.66 1.00 0.26 0.25 -0.43 1.26 0.57 0.95 0.32 0.34 0.50 0.83 0.33 0.22 1.05 -0.61 0.88 1.00 0.76 0.17 0.95 1.00 0.77 0.09 -2.30 5.35 0.56 0.81 0.34 0.23 The revenue efficiency scores presented in this table are obtained by the following equation: s RE q r 1 s q r 1 r yr s where r y* q r is the price outputs and y r is the quantity of output r. q r y r is the observed/actual revenue of the DMU and r 1 s q r 1 r y * is the virtual efficiency profit that could be achieved if the DMU lies on the efficient frontier. The profit efficiency scores take values in the range (0, 1). Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 Table 5: The Nonparametric Cost Efficiency of Kuwaiti Banks (1996-2010). Average(Whole Sample) Max Min Stdev Skewness Kurtosis Number Of banks Average(Non Islamic Banks) Max Min Stdev Skewness Kurtosis Number of Banks Average(Islamic Banks) Max Min Stdev Number of Banks 1996 0.84 0.93 0.64 0.12 -1.58 2.30 5 1997 0.81 0.98 0.67 0.13 0.09 -1.98 7 1998 0.85 1.00 0.71 0.11 0.20 -1.66 7 1999 0.88 1.00 0.66 0.12 -1.09 1.24 7 2000 0.83 0.99 0.66 0.12 -0.36 -0.69 7 2001 0.80 1.00 0.59 0.15 -0.22 -1.37 7 2002 0.80 1.00 0.55 0.16 0.07 -0.87 7 2003 0.86 1.00 0.66 0.16 -0.45 -2.28 7 2004 0.63 1.00 0.30 0.27 0.07 -1.77 7 2005 0.57 0.76 0.37 0.15 0.04 -1.65 8 2006 0.59 1.00 0.44 0.23 0.75 -0.48 8 2007 0.56 1.00 0.28 0.23 1.11 0.61 9 2008 0.53 0.82 0.28 0.20 0.31 -1.01 9 2009 0.61 1.00 0.36 0.25 0.91 -0.84 9 2010 0.60 1.00 0.42 0.17 1.90 4.16 9 0.84 0.93 0.64 0.12 -1.58 2.30 5 0.83 0.98 0.67 0.12 -0.30 -1.61 6 0.68 0.87 1.00 0.75 0.10 0.08 -1.97 6 0.71 0.92 1.00 0.83 0.07 -0.09 -2.26 6 0.66 0.86 0.99 0.69 0.10 -0.58 1.44 6 0.66 0.84 1.00 0.69 0.12 -0.27 -1.43 6 0.59 0.84 1.00 0.70 0.14 0.47 -2.22 6 0.55 0.90 1.00 0.70 0.13 -1.11 -0.43 6 0.66 0.60 1.00 0.30 0.28 0.48 -1.60 6 0.81 0 1 1 1 1 1 1 1 1 0.52 0.76 0.37 0.14 0.80 -0.26 6 0.71 0.77 0.45 0.02 2 0.58 1.00 0.44 0.24 0.97 -0.03 6 0.61 0.77 0.45 0.23 2 0.59 1.00 0.28 0.28 0.67 -1.00 6 0.49 0.60 0.42 0.09 3 0.58 0.82 0.28 0.22 -0.05 -1.22 6 0.45 0.62 0.30 0.16 3 0.55 0.92 0.36 0.22 1.63 2.87 6 0.78 1.00 0.56 0.31 3 0.61 1.00 0.42 0.20 1.93 4.32 6 0.58 0.71 0.49 0.12 3 The cost efficiency scores presented in this table are obtained by the following equation: m CE px i 1 m * i i px i 1 M inimum virtual cost 1 where p i is the price of input i, x * is the optimal input quantity and x i is the actual inputs quantity. Observed cost i i The cost efficiency score will be one for DMUs on the efficient frontier. The cost efficiency scores take values in the range (0,1). Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 Table 6: The Nonparametric Profit Efficiency of Kuwaiti Banks (1996-2010). Average(Whole Sample) Max Min Stdev Skewness Kurtosis Average(Non Islamic Banks) Max Min Stdev Skewness Kurtosis Average(Islamic Banks) Max Min Stdev 1996 0.65 1.00 0.20 0.35 -0.20 -1.90 1997 0.64 1.00 0.33 0.27 0.56 -1.38 1998 0.71 1.00 0.36 0.29 -0.03 -2.28 1999 0.80 1.00 0.51 0.22 -0.38 -2.29 2000 0.71 1.00 0.49 0.19 0.59 -0.94 2001 0.78 1.00 0.55 0.20 -0.09 -2.30 2002 0.79 1.00 0.46 0.22 -0.35 -1.50 2003 0.94 1.00 0.66 0.14 -2.45 6.00 2004 0.60 1.00 0.22 0.29 -0.13 -1.72 2005 0.49 0.77 0.32 0.14 1.08 1.56 2006 0.60 1.00 0.42 0.27 0.48 -0.57 2007 0.67 1.00 0.15 0.30 -0.34 -0.75 2008 0.54 1.00 0.09 0.33 0.14 -0.99 2009 0.53 1.00 0.11 0.40 0.12 -2.12 2010 0.83 1.00 0.32 0.26 -1.27 0.18 0.65 1.00 0.20 0.35 -0.20 -1.90 0.66 1.00 0.33 0.29 0.31 -2.05 0.53 0.72 1.00 0.36 0.32 -0.17 -2.92 0.65 0.84 1.00 0.51 0.21 -0.98 -1.04 0.58 0.72 1.00 0.49 0.20 0.27 -1.40 0.60 0.81 1.00 0.55 0.19 -0.58 -1.71 0.58 0.82 1.00 0.46 0.22 -0.92 -0.12 0.59 1.00 1.00 1.00 0.00 -2.24 5.00 0.66 0.58 1.00 0.22 0.31 0.18 -2.06 0.74 0.49 0.77 0.32 0.16 1.23 1.99 0.50 0.58 0.58 0.24 0.59 1.00 0.42 0.31 0.55 -1.57 0.62 0.66 0.58 0.06 0.62 1.00 0.15 0.33 0.04 -0.90 0.79 1.00 0.54 0.23 0.57 1.00 0.09 0.29 -0.36 1.98 0.47 1.00 0.15 0.46 0.40 1.00 0.11 0.38 1.26 0.59 0.87 1.00 0.73 0.19 0.94 1.00 0.68 0.13 -2.27 5.20 0.61 1.00 0.32 0.35 The profit efficiency scores presented in this table are obtained by the following equation: s q r 1 s q r 1 m r r y r p i xi i 1 m y r* pi xi* s where q r 1 m r y r p i x i is the observed profitability of the banki and i 1 s q r 1 m r y r* p i x i* is the virtual profitability that i 1 i 1 could be achieved if the DMU is located on the efficient frontier. The profit efficiency score takes values in the range ( ,1) . Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 References Abd-Majid, M. Saal.D. and Battisti. G., (2010), “Efficiency in Islamic and Conventional Banking: An International Comparison”, Journal of Productivity Analysis, Vol. 34, pp. 25-43. Available at SSRN: http://ssrn.com/abstract=1668966 Al-Delaim, K .S. and Al-Ani, A., (2006), “Using Data Envelopment Analysis to Measure Cost Efficiency with an Application on Islamic Banks”, Scientific Journal of Administrative Development, 4: 134-156. Al-Sharkas, A. A., M. K. Hassan and S. Lawernce, (2008). “The Impact of Mergers and Acquisitions on the Efficiency of the U.S. Banking Industry: Further Evidence”, Journal of Business Finance & Accounting 35, pp. 5070 Ariss, R., Financial liberalization and bank efficiency: evidence from post-war Lebanon, Applied Financial Economics, 18 (2008), pp. 931-946. Ariss, R., (2010), “Competitive behavior in Middle East and North Africa banking systems”, The Quarterly Review of Economics and Finance, Volume 49, Issue 2, May 2009, Pages 693-710. Bdour, J & Al-khoury, A (2008) „Predicting Change in Bank Efficiency in Jordan: A Data Envelopment Analysis‟, Journal of Accounting and Organizational Change, 4, 162–181 Berger, Allen N., and David B. Humphrey(1992). „„Megamergers in Banking and the Use of Cost Efficiency as an Antitrust Defense,‟‟ Antitrust Bulletin, vol. 37, pp. 541–600 Charnes, A. Coopers,W. and Rhodes, E. (1978), “Measuring the Efficiency of Decision Making Units”, European Journal Operational Research, 3, 429 – 444 Darrat A.F., C. Topuz, and T. Yousef, 2003, “Assessing Bank Efficiency of Banks in an Emerging Market: The Kuwaiti Experience in the 1990s”, Studies in Economics and Finance, 21, 1-21. Elyasiani, E., Mehdian, S (1992). Productive efficiency performance of minority and nonminority-owned banks: A nonparametric approach. Journal of Banking and Finance 16, 933–948. Fadzlan, S. and Muhd-Zulkhibri, A. M. (2007), “Bank Ownership, Characteristics and Performance: A Comparative Analysis of Domestic and Foreign Islamic Banks in Malaysia”, Munich Personal RePEc Archive, 12131 Fare, R., Grosskopf, S., Norris, M., Zhang, Z., 1994. Productivity growth, technical progress, and efficiency change in industrialized countries. The American Economic Review 84, 66–83. Farrell, M.J. (1957), “The Measurement of Productive Efficiency”, Journal of the Royal Statistical Society, 120(3): 253-281. Gattoufi, S., Al-Hatmi, S., (2009). A Productivity Analysis of the Omani Banking Industry: A Data Envelopment Analysis (DEA) Approach to Decompose and Analyze Its Technical Efficiency. Forthcoming, International Journal of Accounting and Finance Proceedings of 23rd International Business Research Conference 18 - 20 November, 2013, Marriott Hotel, Melbourne, Australia, ISBN: 978-1-922069-36-8 Grabowski, R., Rangan, N., Rezvanian, R., 1993. Organizational forms in banking: An empirical investigation of cost efficiency. Journal of Banking and Finance 17, 531-538. Isik, I., 2000. Efficiency and performance in international financial institutions. A dissertation, Department of Economics and Finance Isik, I., Hassan, M.K., 2002. Technical, scale and allocative efficiencies of Turkish banking industry. Journal of Banking and 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. Kabir, M.H. (2006), “The X-Efficiency of Islamic Banks”, Islamic Economic Studies, Vol. 13, No. 2, February 2006. Kamaruddin, B.,Safa, M. and Rohani, M. (2008), “Assessing Production Efficiency of Islamic Banks and Conventional Bank Islamic Windows in Malaysia”, International Journal of Business and Management Research, 10670: 31-48 Leightner, E J and C A K Lovell (1998): „The Impact of Financial Liberalisation on the Performance of Thai Banks‟, Journal of Economics and Business, 2, 115-132. Maghyereh, A 2004, „The Effect of Financial Liberalization on the Efficiency of Financial Institutions: The Case of Jordanian Commercial Banks‟, Working Paper Series. SSRN: http://ssrn.com/abstract=436420 Srairi S. (2009). 'A Comparison of the profitability of Islamic and conventional banks: The case of GCC countries', Bankers, Markets and Investors, nº 98, pp.16-27. Wheelcock, D. and P. Wilson (1995), “Explaining bank failures: Deposit insurance, regulation, and efficiency” , Review of Economics and Statistics, 77, pp. 689-700.