Ownership Structure, Performance and Risk in the Indonesian Banking Industry Maria Ulpah Student Number: 1740032 Supervisor:

(1)

Ownership Structure, Performance and Risk in the Indonesian Banking Industry

Maria Ulpah

Student Number: 1740032

Supervisor: Dr. Lammertjan Dam Prof. Dr. C.L.M. (Niels) Hermes

University of Groningen Faculty of Economics and Business MSc BA Finance-Corporate Finance

(2)

Abstract

This study investigates the effect of ownership structure on bank performance and risk in Indonesia. In particular, using Generalized Least Squares (GLS) I first check whether Government Owned Banks (GOBs) outperform Publicly-Owned Banks (POBs), and second, whether listed banks outperform non-listed banks. I use data of 109 banks in Indonesia for the period 2002-2006. I use various performance measures: Efficiency score based on the Data Envelopment Analysis (DEA) approach and accounting ratios of performance: Income, Profit and Expenses. For bank risk taking behaviour, I use Loanloss ratios. I find that Government Owned Banks (GOBs) are more profitable and more efficient than the Privately-Owned Banks (POBs) and listed banks are less profitable, less efficient, and take less risk loans than non-listed banks.

(3)

I. Introduction

Most of the theoretical research defines ownership structure in two main dimensions. The first is the category of ownership or nature of the ownership: a firm can be owned by the public, the state or a foreign company. The second is the ownership concentration: a firm can have dispersed ownership or concentrated ownership. This study is both about the effect of the nature of ownership and ownership concentration on bank efficiency, performance and risk in Indonesia banking industry. I specifically focus on Government Owned Banks (GOBs) Vs Privately-Owned Banks (POBs) and Listed Vs Non-Listed Banks.

Regarding government ownership, Alchian (1965) notes that the nature of ownership can be explained by the property right hypothesis. He defines a property right as an exclusive authority to determine how a resource is used, whether that resource is owned by government or by individuals (private ownership). According to the property right hypothesis, private ownership is superior to public ownership. Sapienza (2004) mentions three views that might explain the role of the government ownership: The Social, the Political and the Agency view. The Social view assumes that government owned banks are a solution to address market failures and might foster economic development. Therefore, there might be a positive relationship between government ownership and financial and economic development. The

Political view states that politicians will behave based on personal interest and a

government owned bank is merely a tool to achieve their personal objective. The

Agency view assumes that there is a separation between the management and the

owners of banks. Managers - the agents- might misbehave and shirk when they find that a project will reduce their private benefits and/or compensation. The Social view assumes that government ownership might improve efficiency and performance of government owned bank, whereas the Political and the Agency view imply that there is misallocation of resources that lead to the inefficiency.

Some of the existing empirical studies find that Government Owned Banks are less profitable and less efficient than Privately-Owned Banks (Berger et al., 2005, Sapienza 2004, Ianotta et al. 2007). According to Shleifer and Vishny (1997), the

bureaucrats have concentrated control rights of the state firms, and not the public, and

(4)

Political and Agency view. In contrast, some of the studies support the social view of

the role of government ownership, namely that Government Owned Banks outperforms Privately-Owned Banks (Bhattacharya et al. 1999, Altunbas et al.2001, Isik and Hassan 2003).

Regarding ownership concentration, Bearle and Means (1932) were the pioneers in studying this subject by examining the ownership structure in the United States. They emphasize a negative relationship between ownership concentration and performance. Moreover, they also mention that separation between management and shareholder would eliminate the control of the shareholder over the management, potentially leading to misbehaviour of the manager. Jensen and Meckling (1976) give the foundation of this agency problem approach in examining the effect of ownership concentration on performance. The agency problem might induce managers to misbehave and will lead to inefficiency and less profitable firms.

Firms might be assumed to have a dispersed ownership when listed in a stock exchange. Several studies have investigated the effect of going public to the risk and performance of the firm and result in different conclusion. Sarkar et al. (1993) study the effect of ownership on performance on The Indian banking industry. They find there is no difference in performance between listed and non-listed banks, whereas Isik and Hassan (2003) note that being a publicly traded company appears to be positively associated with an increase efficiency and performance. In contrast, Kwan (2004) finds that listed banks tend to be less profitable than non-listed banks.

There are a limited number of studies that examine the effect of ownership structure on performance and risk in developing countries, such as Indonesia. This lack of empirical studies and the need to link the theory to literature in banking is important to do, especially for the developing countries where healthy banks are needed to support the economic development. In this context, this study tries to investigate the effect of ownership structure on bank efficiency, risk and performance in the Indonesian banking sector. The results of this study might be useful in the context of regulation and policy regarding banking supervision, competition policy and prudential regulation.

(5)

based on Barr et al (1999) who use the Intermediation approach (see the methodology section for more detail). Next, I investigate the effect of ownership on bank efficiency, traditional performance measurements and risk. A Generalized Least Squares (GLS) regression analysis is used to find the effect of ownership structure on risk and performance from the 109 banks during period 2002-2006.

This paper is organized as follows. I provide an overview of existing literature on performance, risk and ownership in section 2. Section 3 gives the data description and the methodology is presented in section 4. Empirical results and robustness check will be provided in section 5 and section 6 respectively. Section 7 concludes this paper.

II. Performance, risk and ownership structure

Generally, the nature of banks ownership can be categorized into three categories: Government Owned Banks, Privately-Owned Banks and Foreign owned banks. Government ownership is one example of the government participation on the financial market. La Porta et al. (2002) find that on average government ownership is still large and such ownership is commonly found in countries with low levels of income per capita, underdeveloped financial systems, heavy government intervention, and less secure property rights protection. Therefore, government ownership in still has a key role in many countries.

Recently, Sapienza (2004) mentions three views regarding the role of government ownerships on banking sector: The Social, the Political and the Agency

views. The Social view, or Development view, was firstly introduced by Alexander

Gerschenkror (1962). He suggests that governments should jump in through financial and economic development to address market failure. This view assumes that government banks can collect funds and directly allocate them to the long term profitable and strategic projects, thus generating demand and benefits and encouraging growth. According to this view, Government owned banks have a positive role in development and might establish that Government Owned Banks are profitable. Therefore they can contribute to economic development.

(6)

government owned banks are only political tools to achieve individual interests and eventually will generate inefficient banks.

The last view, the agency view, assumes that government owned banks are there to address market failures, but on the other hand banks also have multiple non-measurable objectives that might alter the manager’s incentives and benefits. Therefore, under this view, Government Owned Banks would allocate funds to profitable and strategic projects, but managers of these banks may provide less effort because it would reduce their personal benefits, such as private benefits and perks (Sapienza 2004). According to the Social and Political view, there is a misallocation of resources but for different reasons. Personal objectives of the politician is the reason for the misallocation in the Political view, whereas in The Agency view misallocation happens because managers misbehave, either they shirk or use the firm assets for personal use (Sapienza 2004)..

Cornett et al. (2000) study the performance differences between privately-owned banks and state-privately-owned banks in South Korea, Indonesia, Malaysia, the Philippines and Thailand for the period 1994-1997. The main finding of this study is that state ownership is associated with significantly inferior performance. The decrease in the performance of state-owned banks is larger than the decrease in performance of privately-owned banks during the crisis period. Barth et al. (2002) study the relationship between bank ownership and performance by using data from 60 countries. This study finds that there is no relationship between ownership structure and performance and there is a negative relationship between government ownership and the financial development. Another interesting finding is the inverse relationship of the degree of innovation and government ownership.

(7)

Altunbas et al. (2001) study the relationship of performance and ownership structure in the German banking industry. They find that Government Owned Banks outperform Privately-Owned Banks. This is consistent with the study by Hart, Shleifer, and Vishny (1997) that give an argument for efficiency of government ownership. Sapienza (2004) conducted a study on the effect of government ownership on bank lending. He focuses on the lending relationship instead of using the overall activities. Using data from Italian banks and firms, he find that State owned banks charge much lower interest rate than privately-owned banks. This main finding fully supports the political view of Government Owned Banks, since he find that the lending behaviour of the Government Owned Banks are influenced by the electoral results of a party affiliated with the banks.

Moving to the empirical studies that examine efficiency of the financial institution, Berger and Humprey (1997) examine efficiency of 130 financial institutions in 21 countries. They find that the various methods to calculate efficiency of financial institution will yield different results. As such, the survey gives the implication of efficiency calculation on Government policy. Barr et al. (1999) examined the productive efficiency of the commercial banks in United States over period 1984-1998. In recent years, a large number of studies have been conducted to assess the banking sector efficiency. There are two approaches that are mostly used to assess the efficiency: the Parametric and Non-Parametric technique. The

Parametric approach can be divided into three approaches: Stochastic Frontier

Approach (SFA) (Greene, 1993, Lovell, 1993), the Thick Frontier Approach (TFA) (Berger and Humprey, 1992) and the Distribution Free Approach (DFA) (Berger, 1993). Data Envelopment Analysis (DEA) (Sathye 2003, Isik and Hassan, 2003) is labelled as The Non parametric approach to calculate the efficiency score.

(8)

private banks and foreign banks. Isik and Hassan (2003) who study the effect of ownership on efficiency of the Turkish banks also found that public banks are better than the private banks in term of technical efficiency.In contrast; Berger et al. (2005) found that Government owned banks in China underperformed than the privately-owned banks and foreign banks. The study also found the financial reform in China results the decreasing number of the Government banks and effect the efficiency level.

Moving to the relationship between Ownership concentration -Listed Vs Non-Listed banks- and performance, Isik and Hassan (2003) state that the market discipline hypothesizes that publicly traded banks are more efficient than private banks. By using the Turkish banking industry, they find that publicly traded banks are more efficient. Loderer et al. (2006) and Ianotta et al. (2007) note that publicly traded banks appear to be positively associated with bank performance. Loderer et al. (2006) state that listed firms might give a better care of the minority investor therefore listed firms might be performed better. In contrast, Kwan (2004) finds that listed banks tend to be less profitable and less efficient than non-listed banks. He claims that this result is related to the agency theory view that assumes that the separation between management and shareholder is negatively associated with firm performance. It is consistent with Jensen (1978) who argues that diffused ownership is associated with the problem of agency costs that lead to inefficiency.

III. Data and Descriptive Statistics

A sample of banks in Indonesia is drawn from Bank Indonesia database and BankScope database. Variables are extracted from the Balance sheet, Income statement and Ownership information from Bank Indonesia database and BankScope database. I start from the complete sample of banks in Indonesia: 133 banks. I exclude banks that do not have data for the whole research period to create a balanced panel; therefore I end up with a smaller sample. The final data set consist of 109 banks with 545 bank year observations.

(9)

performance besides the traditional measurement. The Efficiency score1 is calculated as the total productivity factor (TFP) using the software package Deap (Ver 2.1) by Tim Coelli. Income, Profit and Expenses as the traditional measurement indicator are mostly used in the study of the relationship between ownership and performance. These traditional measurements of performance are used to see how ownership structure affects the bank performance.

The second model tries to investigate determinants of banking sector risk in Indonesia. Loan loss reserve as a fraction to total earning assets (Loanloss) is used to measure of a bank’s risk based on accounting information. A higher level of bank provision might indicate a lower level of risk. The advantage of the loan loss variable is that it uses accounting data; therefore it can be applied to non-listed banks.

Again, to define bank ownership, I employ the following ownership characteristics: Government Owned Banks (GOBs) versus Privately-Owned Banks (POBs) and listed banks versus non listed banks. To identify government versus private ownership is usually straight forward, but due to some specific Indonesian banking characteristic, I had to make a few judgments. First, I classify Regional Development banks as Government owned banks because the regional development banks are owned by local government. Second, I classify foreign ownership as private ownership. There are 5 states owned banks and 26 regional development banks, but due to data availability, I end up with 21 banks in the sample that are government owned banks. Privately owned banks in Indonesia consist of foreign exchange commercial banks, non-foreign exchange commercial banks; joint venture banks, Islamic banks and foreign owned banks. It might be found that ownership structure may affect the performance and risk taking behaviour of banks. As we can observe from figure 1, the assets of Regional Development Banks have grown rapidly during the period 2002-2006 relative to other banks. It indicates that the Regional Development Banks provide significant contribution to the development of banking industry in Indonesia

1

(10)

Figure 1

Asset growth of banking industry in Indonesia during the period 2002-2006

Asset Growth 2002-2006 -20.00% -10.00% 0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% 2003 2004 2005 2006 St at e Owned bank

Foreign exchange commercial banks Non-foreign Exchange commercial banks Regional Development banks Joint vent ure banks Foreign owned banks

Source: Bank Indonesia 2006

Besides Government Owned Banks versus Privately Owned Banks, I also examine the effect of listed firm and non-listed firms on banking risk and performance. All the listed banks are listed in Jakarta Stock exchange.

I use several control variables that give specific characteristics of banks that are believed to be highly important in explaining the banks performance and risk. These variables are Size, Loans, Deposits, Capital and Loanloss. These control variables are suggested by Berger et al. (2005) and Ianotta et al. (2007). Loan loss is only used as control variable in model 1. More detail on the variables is in table 1 of the Appendix.

(11)

Table 1

Descriptive statistics of banks characteristics

Full Sample 2002 2003 2004 2005 2006 Mean 12370 10072 10762 11621 13630 15765 Std.Dev 33368 30205 31162 32329 34696 38135 Mean 5090 3239 3796 4979 6235 7203 Std.Dev 12221 7685 9208 12029 14156 15886 Mean 1276 1214 1141 1063 1281 1669 Std.Dev 3457 3824 3312 2853 3192 3995 Mean 10021 8159 8789 9456 11033 12666 Std.Dev 27427 24614 25607 26164 28793 31628 Mean 696 826 659 527 597 872 Std.Dev 2039 2822 2080 1450 1503 2048 Mean 186 118 148 182 226 255 Std.Dev 534 345 448 509 626 672 Mean 3.35 3.24 3.28 3.33 3.41 3.49 Std.Dev 0.82 0.80 0.80 0.81 0.83 0.83 Mean (%) 1.91 1.87 2.07 8.60 8.26 7.30 Std.Dev (%) 3.67 7.13 1.81 67.05 66.32 58.86 Mean (%) 2.42 1.88 2.55 2.74 2.67 2.27 Std.Dev (%) 3.86 6.10 2.73 3.74 2.86 2.80 Mean 0.56 0.39 0.50 0.68 0.65 0.59 Std.Dev 0.21 0.20 0.18 0.18 0.17 0.17 Mean (%) 5.77 5.11 4.75 5.81 5.29 7.90 Std.Dev (%) 14.35 4.87 3.16 6.71 4.40 30.53 Mean (%) 1.75 2.83 1.64 1.59 1.28 1.43 Std.Dev (%) 3.47 6.56 2.56 1.98 1.93 1.33 Mean (%) 51.38 52.82 53.63 71.59 67.99 64.71 Std.Dev (%) 21.07 24.47 23.40 107.01 93.34 74.25 Mean (%) 78.70 77.44 78.14 83.13 78.00 76.78 Std.Dev (%) 29.15 15.97 15.23 57.18 15.29 16.29 Mean (%) 13.91 13.22 15.23 14.18 13.45 13.46 Std.Dev (%) 12.30 9.90 17.34 12.01 10.87 9.94 545 109 109 109 109 109 109 109 109 109 109 109 Number of banks Loans/Total earning assets Deposits/Total earning assets Capital/Total assets Number of observations Income/Total earning assets Efficiency

Expenses/Total earning assets LoanLoss/Total earning assets Personal Expenses(Rp in millions) Variabels used in GLS regression Size (Log Asset)

Profit/Total earning assets Interest income (Rp in millions) Input Variables (DEA Estimator) Deposits (Rp in millions) Interest expenses(Rp in millions)

Variables Assets (Rp in millions )

Output Variables (DEA Estimator) Loans(Rp in millions)

Size as log of total assets shows an increasing trend from 2002 to 2006. The

log of total assets reaches a number of 3.49, whereas it is only 3.24 in 2002. The increasing number shows a significant development in the Indonesian banking sector. The increasing number of total assets also associated with the increasing number of total loans per total asset. A higher level of the ratio total loans per total assets (Loans) is reached at 2004 with 71.59%. The same pattern also happens for the variable Deposits, Profit, Expenses and Capital. The average of Efficiency has increased from 59% to 65% during period 2002-2005 and it reach number 59% at 20062. The level of Efficiency is consistent with the profitability measurement. As can be seen from the table, Profit and Income have increased during 2002 until 2006. The growth of a bank automatically will increase the bank’s expenses. Loan loss ratio shows an interesting finding, since the value will decrease when banks is getting larger (in terms of total asset). Borio (2003) states that banks turn bank provision

2

(12)

down when economic condition increase and the higher lever of the reserve might be interpreted as lower risks, but Altunbas et al. (2007) argue that banks who have higher level of Loanloss expect that the risk in the future is higher.

Table 2 provides the correlation matrix for the dependent and independent variables that will be used in the GLS regression. As observed from the table, Size has positive correlation with the Profit, income and Efficiency. Negative and significant correlation can be found between Size and Loans, Size and Loan loss and Size and

Capital. It means that larger banks are less capitalized than the small banks and it also

give fewer loans. Deposits and Size has a positive correlations, it might be because of larger banks have a good reputation; therefore people put too much trust to save their money at these large banks. The higher correlation can be seen between Income and

Profit.

Table 2

Correlation matrix among variables

Size Profit Expenses Loanloss Loans Deposits Capital GOBs Listed Efficiency Income

Size 1 Profit 0.020 1 Expenses -0.042 0.040 1 Loanloss -0.116** 0.059 0.129** 1 Loans -0.083* 0.049 0.203** 0.219** 1 Deposits 0.072* -0.080* -0.009 -0.124** 0.006 1 Capital -0.390** 0.164** 0.019 0.055 -0.069 -0.181** 1 GOBs 0.280** 0.131** 0.067 -0.112 -0.103 0.025 -0.101** 1 Listed 0.396** -0.059 -0.027 -0.077* -0.069 0.034 -0.096* -0.100* 1 Efficiency 0.008 0.300** 0.022 0.115* 0.173** -0.252** 0.362** 0.055 -0.087** 1 Income 0.084** 0.825** -0.020 -0.067* 0.085* -0.018 0.083* 0.152** -0.148** 0.292** 1 **. Correlation is significant at the 0.01 level (2-tailed).

*. Correlation is significant at the 0.05 level (2-tailed).

Correlations

Table 3 shows the summary of t-test equality of Government Owned Banks (GOBs) versus Privately-Owned Banks (POBs) and Listed Vs non-listed bank variables means. Significant differences between GOBs and POBs refer to Size,

Profit, Income, Expenses, Loanloss and Capital. I find that GOBs are more profitable

(13)

account the ownership variables that might impact bank efficiency. However, it is unclear whether, for example, a size effect might drive these differences and therefore I turn to regression analysis to control for such characteristics.

Significant differences between listed and non-listed banks refer to Size,

Income, Efficiency, Deposits and Capital. The listed banks have higher assets value

but have less profit and more expenses than the non-listed banks. The listed banks are less capitalized and have a less percentage of loans than the non-listed banks. Non-listed banks have better performance and more efficient than the Non-listed banks.

Table 3

Test for difference in mean of Government owned banks (GOBs) Vs Private owned banks (POBs) and Listed Banks Vs Non-Listed Banks

GOBs POBs Listed NonListed

3.82 3.24 3.94 3.17 2.90 1.67 1.50 2.03 3.63 2.13 1.37 2.73 0.59 0.56 0.53 0.57 7.75 5.30 5.06 5.99 0.96 1.94 1.26 1.90 46.55 65.87 52.81 64.93 80.22 78.34 80.53 78.15 11.35 14.52 11.74 14.55 No.of obs 105 440 125 420 No.of Banks 21 88 25 84

**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).

Variables

0.006** 0.012*

0.177 0.000**

Size (Log Asset) Profit (%) 0.173 0.048* Loans (%) Deposits (%) Capital (%) Income (%) Efficiency Expenses (%) LoanLoss (%) _0.000** _0.250 0.154 0.142 0.02* 0.040* 0.476 t-statistics t-statistics 0.009** 0.000** 0.017* 0.50 0.021** IV. Methodology

It is common to measure bank performance by various financial ratios. Sherman and Gold (1985) note that the financial ratios do not capture long

term performance. Therefore in their study, Sherman and Gold (1985) measure bank performance by using one of the frontier analysis methods to measure the efficiency level. They were the first to apply the Data Envelopment Analysis (DEA) methods in the banking industry (Sathye 2003). Therefore, in this study I will focus on Efficiency as the main indicator of bank performance and Income, profit and expenses variables, as traditional measurement of performance, to see whether it supports how the ownership structure affects the banking sector performance.

(14)

indicators: Efficiency. The discussion how to measure Efficiency by DEA technique will be provided below. The second stage is assessment of the effect of structure ownership on efficiency, performance and risk of Indonesian banking industry by using the regression analysis, which will be discussed in the subsequent section.

Estimating Bank Efficiency

There are two empirical ways to measure the efficiency: namely parametric and non-parametric approach. Data Envelopment Analysis (DEA) is a non-parametric approach and deterministic methodology to determine the relative efficient frontier based on the empirical data of chosen input and output variables, called Decision Making Units (DMU). DEA determines which banks in the sample produce a particular output combination at the given input prices at least cost (Mester 1996).

DEA has been widely used to assess the banking sector efficiency. I use the most frequently used model of Data Envelopment Analysis (DEA) approach by Charnes, Cooper and Rhodes model (see Charnes, Cooper, and Rhodes, 1978). The CCR model defines efficiency as a function of the weight of input and output combination (Charnes et al. 1978). The efficiency of the entities can be solved by the linear programming. Let’s assume there are m banks in the sample and try to estimate efficiency score for bank k. The maximization problem reads:

k Max EFF =(

∑

u y_r _rk) /(

∑

v x_j _jk) Subject to: (

∑

u y_r _rm) /(

∑

v x_j _jm) 1;≤ m=1,...n 0, 1, 2... r u ≥ r=

∑

s m 0, 1, 2... j v ≥ j=

∑

The above model evaluate the relative efficiency of bank k based on performance of the others banks. Where y is the observed amount of output _rk

variables of bank k and y is the observed amount of the output variables of bank _rm

m=1, 2…..n, whereas x is the observed amount of input variables of bank k and jk x rm

is the observed amount of the input variables of bank m=1, 2…..n. The variables and are the weight that will be determined by the above linear programming. Each

r u j

(15)

bank’s maximum efficiency score will be less than or equal to 1, if a bank has a score of 1, it means that the bank is “the best practice’ bank.

In order to estimate value of Efficiency of 109 banks, I use the intermediation

approach to determine input-output variables. Following the study by Barr et al. (2002), I select the input-output variables based on the intermediation approach3. The outputs variables are: Loans and interest income, whereas the input variables are deposits, personal expenses and total operating expenses. Descriptive statistics of the input and output variable can be found in table 1.

The Effect of Ownership Structure on Efficiency, Performance and Risk

I specify 2 models to evaluate: i) the effect of ownership structure on: bank efficiency, performance and ii) the effect of ownership structure on bank risk. These models are based on Lannotta et al. (2007). The first model examines the effect of ownership structure on performance:

t

P = +α βOS +χCV_t +ε_t (1)

Where is the observed performance variable of bank at time t. The performance measurements are Efficiency, Income, Profit and Expenses. is the ownership variables: Government Owned Banks (GOBs) versus Privately-Owned Banks (POBs) and Listed versus Non-Listed Banks. is the control variables. The control variables are Size, Loans,Loan loss, Deposit and Capital as control variables. I

am specifically interested in the parameter

t P

OS

t CV

β,which reflect the effect of ownership. Next, I also estimate the effect of ownership on risk taking behaviour based on the following equation:

t

R = +α βOS +χCV_t+ε_t (2)

Where R is the observed variable as a proxy of banks’ risk at time. I

employed Loanloss as a proxy of bank’s risk. OSis the ownership variables. OSis

3

(16)

the ownership variables : Government Owned Banks (GOBs) versus Privately-Owned Banks (POBs) and Listed versus Non-Listed Banks. is the control variables. The control variables are Size, Loans, Loan loss, Deposit and Capital as control variables.

Based on the existing literature that study the effect of ownership, I conclude the expected relationship between ownership structure and risk and performance of Indonesian banking industry on table 4 below.

t CV

Table 4

Expected sign of dummy and control variables in regression model

Expected Sign on Dependent variables (+,- or unknown) Independent variables

Profit Income Expenses Efficiency Loan loss

GOB - - + - + Listed +/- +/- +/- +/- _ Size + + - + +/- Loans +/- +/- + +/- +/- Loan loss +/- +/- + +/- NA Deposit +/- +/- + +/- +/- Capital +/- +/- +/- +/- +/- Notes:

1. + indicate that the independent variable might have a positive coefficient 2. - indicate that the independent variable might have a positive coefficient

3. +/- indicate that the independent variable might have a positive or negative coefficient

V. Empirical Results

To assess the relationship between ownership structure, performance and risk, I apply Generalized Least Square (GLS) multiple regressions for pooled data. GLS regression computes consistent estimates of the covariance matrix allowing for heterosdatisticity. As mentioned in the literature, when Efficiency is used as

dependent variable whereas its value attain between zero and one, logistic form regression might be applicable to use (Isik and Hassan 2003 ). In some cases, the

efficiency score might be censored around zero or one, as a result Tobit regression

(17)

poses not a real problem to apply GLS regression. Hereafter the result discussion based on the GLS regression in table 5, 6 and 7.

Table 5 reports the estimation results of equation (1) and use Efficiency as the

Dependent variable. As can be observed in column 3 table 4, the estimated coefficient for variable GOBs is positive and significant. This indicates that the Government owned banks in Indonesia are more efficient that the privately owned banks4. This empirical result is in contrast with the study by Altumbas et al. (1994) and Ianotta et al (2007). Likewise, Bhatttacharya et al. (1997) found the same result where the public banks in India were found to be more efficient than the private banks. This result is surprisingly different from study by Hadad et al. (2003) that found there was no relationship between ownership structure and performance in Indonesian banking industry5.

Table 5

Bank Efficiency and ownership structure variables: Government owned banks (GOBs) versus Privately-owned banks (POBs) and listed banks versus Non-Listed banks

Efficiency as indicator of bank’s performance. P-values in parenthesis. GlS is used to estimate the coefficient

1 2 3 Constant 0.280*** 0.297*** 0.289*** (0.000) (0.000) (0.000) GOBs 0.052*** 0.040*** (0.000) (0.007) Listed -0.043*** (0.001) Size 0.060*** 0.051*** 0.057*** (0.000) (0.000) (0.000) Loans 0.086*** 0.067*** 0.068*** (0.000) (0.000) (0.000) Deposits -0.119*** -0.114*** -0.116*** (0.000) (0.000) (0.000) Capital 0.825*** 0.794*** 0.796*** (0.000) (0.000) (0.000) Loanloss 0.086 0.314 0.115 (0.411) (0.178) (0.664) R-square 0.41 0.43 0.44

GOBs A dummy variable that equals 1 if the banks is Goverment owned banks and zero otherwise Listed A dummy variable that equals 1 if the banks is listed and zero otherwise

Income Ratio of operating income per earning asset Profit Ratio of net income per total earning asset Efficiency Based on DEA calculation

Expenses Ratio of operating expenses per earning asset Size Log of total asset

Loans Ratio of loans per total earning asset Deposits Ratio of total deposits per total earning asset Capital Ratio of equity per total asset

***,**,* indicate statistical significance at 1%, 5% and 10% level, respectively

Variables Efficiency

4

It is supported by the mean of the efficiency score of GOBs that is higher than the mean of the POBs 5

(18)

The positive effect of Government ownership on performance in Indonesia banking also supported by the development of the regional development banks with respect to the application of the local government autonomy. Some of Government Owned banks in Indonesia are large banks, therefore there simply size effect on bank performance. The estimated coefficient for Listed is negative and significant

(respectively at 1%) suggesting that banks who are listed in stock market might be related with the agency theory where the separation between control and ownership will evolve the incentive problem (Iannotta et al. 2007)..

As expected, Size, Loans, and Capital have a positive and significant effect on Efficiency, whereas the estimated coefficient for Loanloss is positive and not

significant. Size acts as a proxy for bank’s ability to diversify. The bigger amount of

bank’s assets, the well diversified and will be less likely to fall than the small banks (Mester 1993). The higher level of loans will increase the efficiency level by reason of the banks’s loans are more valuable than the securities. Moreover, it might be indicating the market power of banks in loan market6 ((Isik and Hassan 2003).

Capital as a proxy for the level of bank’s capitalization. Efficient banks typically have

higher profit which might lead to higher level of capital (Isik and Hassan 2003} and Ianotta et al. (2006) assume that a higher level of capital might represent banks with riskier assets. Indonesian economy is still in the process of recovery after the monetary crisis in 1999; Regulatory pressure might have induced banks to hold more risk. Additionally, banks with riskier operations which might be lead to higher gain and/or profit tend to have more equity to increase the safety and for investor protection reason. Deposit has a negative and significant sign indicating that the

higher amount of deposit lead to higher interest cost, thus decreasing the profitability and may be decrease the efficiency level.

To sum up, the above explanations therefore indicate that Government Owned Banks (GOBs) are outperforming Privately-Owned Banks (POBs) in terms of

efficiency. Listed banks in Indonesia are less efficient than non-listed banks.

Table 6 present the result for regression of Income, Profit and Expenses,

respectively. In line with Efficiency, variable GOBs has a positive and significant

6

(19)

coefficient. Thus, Government owned banks (GOBs) are more profitable than the Privately-Owned Banks (POBs). The sign of Listed dummy variable is negative and

Statistically significance. Thus, banks with dispersed ownership7 might have lower income and profit and still agency problem might be the main reason of it. This result differs from study by (Iannotta et al. 2007) that find that listed banks in European banking industry are more profitable than non-listed banks.

Table 6

Traditional indicator of Bank Performance and ownership structure variables: Government Owned Banks (GOBs) versus Privately-Owned Banks (POBs) and Listed banks versus non-listed banks

Income, Profit and Expenses as indicator of bank’s performance. P-values in parenthesis. GlS Regression is used to estimate the coefficient

1 2 3 4 5 6 7 8 9 Constant -0.004 0.002 -0.002 -0.001 0.002 -0.006 0.031*** 0.027*** 0.038*** (0.310) (0.637) (0.522) (0.712) (0.709) -0.262 (0.000) (0.007) (0.000) GOBs 0.014*** 0.014*** -0.009** 0.013*** 0.033022*** 0.028*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Listed -0.003*** -0.013 0.009*** (0.000) (0.000) (0.000) Size 0.005*** 0.003*** 0.004*** 0.006*** 0.004*** 0.007*** -0.001*** -0.005*** -0.008*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) 0.0088 (0.000) (0.000) Loans 0.003*** 0.003*** 0.003*** 0.004*** 0.005*** 0.005*** 0.035*** 0.037*** 0.037*** (0.004) (0.005) (0.007) (0.006) (0.000) (0.001) (0.000) (0.000) (0.000) Deposits -0.007** -0.009*** -0.007** -0.010** -0.009** -0.010** -0.001 0.004 0.003 (0.03) (0.006) (0.026) (0.017) (0.033) (0.018) (0.988) (0.594) (0.628) Capital 0.070*** 0.065*** 0.070*** 0.077*** 0.075*** 0.077*** 0.008 0.023** 0.014* (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.346) (0.035) (0.092) Loanloss -0.020 -0.021 -0.020 -0.060* -0.062** -0.055* 0.356*** 0.390*** 0.384*** (0.370) (0.332) (0.363) (0.066) (0.046) (0.088) (0.000) (0.000) (0.000) R-square 0.4 0.56 0.59 0.34 0.44 0.69 0.66 0.72 0.76 GOBs A dummy variable that equals 1 if the banks is Goverment owned banks and zero otherwise

Listed A dummy variable that equals 1 if the banks is listed and zero otherwise Income Ratio of operating income per earning asset

Profit Ratio of net income per total earning asset Efficiency Based on DEA calculation

Income Profit Expenses

Most of the controlling variables for bank characteristic are statistically significant, except for the Loan loss variables. This variable gives different result

from the Efficiency regression. Loanloss has a negative and significant coefficient in Income regression, but this variable is not significant in Profit regression. this

7

(20)

negative sign of Loanloss indicate that the higher level of income and/or profit will be

followed by the lower level ratio loan loss per total assets. This result is evidence of possibility of poorer asset quality, thus will reduce level of income8. Another explanation for this result might be related to the moral hazard theory , especially in relation with the compensation problem for manager (Anandarajan et al. 2005), therefore it will reduce the income and profit

As we can see from table 6 columns 9, the regression result when Expenses is

used as the dependent variable. These results are consistent with the prior result of

performance regression indicators. As we can observe, variable GOBs and Listed have positive and significant coefficient. This finding indicates that Government Owned Banks (GOBs) and listed banks have higher cost. The controlling variable of bank characteristics: Size, Loans, Loan loss and Capital have a significant coefficient.

Small bank tends to have higher expenses than large banks. The existence of the economies of scale would bring in the inverse relationship between Size and Expenses

(Iannotta et al. 2007)9.

The use of different proxies as indicator of banking sector performance does not lead to different and ambiguous results. The main findings are Government owned banks (GOBs) are more efficient and more profitable than privately owned banks (POBs). But in the other side, Government owned banks also have higher expenses that suggesting cost inefficiency. However, the higher expenses is compensated with the higher level of profit and income, therefore it yield high level of efficiency. And it is simply a size effect. Kumar et al. (2001) state that firm’s size relate to the economies of scale and scope. Mahjundar (1997) also find that larger firms have better performance than the small firms These empirical results strengthen the social view of the role of government ownership. The result is similar to the study by (Bhattacharyya et al. 1997) that find that Private Banks is India do not outperformed public banks in term of profit and income. Altunbas et al. (2001) also find that public banks in Germany are more profitable than private banks. Conversely, the listed banks in Indonesia are less efficient and less profitable but have higher expenses than non-listed banks. Listed banks might be negatively affected by the market discipline mechanism and could be suffer from the agency problem.

8

There could be a motive that managers tend to use loan loss provision to smooth reported income 9

(21)

Moving to the analysis of Risk versus ownership structure, I estimate the equation (2) and provide the results of Loan loss regression table 7. Column 1 table 7

presents regression results without take into account any difference in ownership structure, while in column 2 table 7 provides regression result with GOBs as an indicator of ownership structure. The last column presents regression result that includes GOBs and Listed variables. GOBs and Listed both have strongly negative and significant coefficient.

Table 7

Bank Risk and ownership structure variables: Government Owned Banks (GOBs) versus Privately-Owned Banks (POBs) and listed banks versus non-listed banks

LoanLoss as indicator of bank’s risk. P-values in parenthesis. GlS is used to estimate the coefficient

1 2 3 Constant 0.023*** 0.016*** 0.015*** (0.000) (0.000) (0.000) GOBs -0.002*** -0.003*** (0.000) (0.000) Listed -0.002*** -0.002 Size -0.003*** -0.002*** -0.001*** (0.000) (0.000) (0.000) Loans 0.008*** 0.008*** 0.008*** (0.000) (0.000) (0.000) Deposits -0.007* -0.001 -0.001 (0.052) (0.549) (0.549) Capital 0.011** 0.017*** 0.017*** (0.034) (0.000) (0.000) Loanloss R-square 0.32 0.39 0.40 GOBs A dummy variable that equals 1 if the banks is Goverment owned banks and zero otherwise Listed A dummy variable that equals 1 if the banks is listed and zero otherwise Income Ratio of operating income per earning asset

Profit Ratio of net income per total earning asset Efficiency Based on DEA calculation

Expenses Ratio of operating expenses per earning asset

Size Log of total asset

LoanLoss

This result is consistent with the test of mean differences for Loan loss, where

(22)

theory, it could be that because of the dispersed ownership firms will take less risk and will have optimized the shareholders wealth10. It also might be caused by the market discipline11

Size, Loans and Capital has a strong significance effect on risk taking

behaviour. The small banks tend to have more loss provision. On the other hand, large banks through economics of scale mechanism tend to have less loan loss provision. Small bank take less risk activities. Banks with higher capital have a propensity to have higher loan loss provision. It confirms that there might be an association between size and potential diversification benefit (Hughes et al, 2001). The results show that Loan has positive relationship with the Loan loss suggesting that the higher

level of loans will lead to the higher level of loan loss provision. Deposit has a

negative influence on Loan loss, but this result was not found to be statistically

significant. In contrast, (Anandarajan et al. 2005) find that there is inverse relationship

between loans and loan loss provision, even the result is not significant.

In conclusion, I find that Government Owned Banks are more profitable than Privately-Owned Banks. This confirms the social view of the government ownership that suggests that Government Owned Banks (states or regional government banks) have to contribute to the economic development and social welfare (Stiglitz, 1993). Non-listed banks are taking more risk than listed banks. Market discipline could be a reason of the listed bank behaviour.

The findings of this study concerning the performance and risk lead to conclusion that Government Owned Banks (GOBs) outperform Privately-Owned Banks (POBs) in term of profitability and efficiency level. I also find that Government Owned Banks is less risky. The rapid development of regional development banks contribute to this finding. As noted by Bank Indonesia in Indonesia bank statistic per December 2007, the regional development banks charge lower interest rate than the private banks and market share of the regional development banks grow faster than the states owned banks, therefore people prefer the regional and state owned banks than the private one (see figure 1 in appendix). This result also strengthens the social view of Government owned banks role.

10

Managers of the listed banks tend to take less risk to secure their compensation and benefits. 11

(23)

VI. Robustness checks

The purpose of this section is to test whether our results are robust to alternative specifications. First, to exploit the panel structure of Banks-fixed effect is not possible because variable GOBs dummy’s do not change overtime, thus to exploit the panel structure, we can do a time-fixed effect and see whether the results are robust. Second, as suggested by Ianotta et al.(2007) that bank profits and income do persevere over time I include a one year lagged performance as explanatory variable in regression model 1, while in regression model 2 I include the one year Loan and Loan loss variables to figure out the influences of last year Loans to the bank risk

taking behaviour in the following year.

Table 8

Ownership structure and performance and risk: Time Fixed effect regression

Efficiency, Income, profit and expenses as indicator of bank’s performance, whereas Loanloss as indicator of bank’s risk. Panel time fixed effect is used. P-values in paranthesis.

1 2 3 4 5

Efficiency Income Profit Expenses LoanLoss

Constant 0.3839*** -0.0037 -0.0140 0.0459 0.0250*** (0.0000) (0.7034) (0.1736) (0.2374) (0.0062) GOBs 0.0356** 0.0123*** 0.0093** 0.0439*** -0.0074* (0.0419) (0.0034) (0.0323) (0.0078) (0.0559) Listed -0.0449*** -0.0048 -0.0185*** 0.0109 -0.0046 (0.0089) (0.2348) (0.0000) (0.4971) (0.2167) Size 0.0480*** 0.0045* 0.0097*** -0.0108 -0.0006 (0.0000) (0.0583) (0.0001) (0.2448) (0.7609) Loans 0.0401*** 0.0033 0.0066*** 0.0370*** 0.0104*** (0.0000) (0.1318) (0.0036) (0.0000) (0.0000) Deposits -0.1423*** -0.0064 -0.0017 0.0041 -0.0133*** (0.0000) (0.2362) (0.7492) (0.8440) (0.0071) Capital 0.6736*** 0.0603*** 0.0510*** 0.0248 0.0092 (0.0000) (0.0000) (0.0004) (0.6476) (0.4711) Loanloss 0.7657*** 0.0592 -0.0926* 0.4237** (0.0000) (0.2058) (0.0572) (0.0211)

GOBs A dummy variable that equals 1 if the banks is Goverment owned banks and zero otherwise Listed A dummy variable that equals 1 if the banks is listed and zero otherwise

Income Ratio of operating income per earning asset Profit Ratio of net income per total earning asset Efficiency score Based on DEA calculation

(24)

As we can observe from table 9, the regression results that include the lagged performance and risk measurement are similar to those obtained from the initial regression in table 5, 6 and 7. Indeed, the lagged performance and risk measurement have a positive and significant coefficient in Efficiency, Income, Profit, Expenses and Loan loss. It indicate that the level of income, profit, expenses and efficiency score

follow the previous year trend. Lagged Loans have a negative and significant

coefficient on Loanloss regression indicating that the higher level of loans on the

previous loans will decrease the value of the loan loss provision in the following year. I might assume that the banks risk taking behaviour is highly affected by the change of loan level. Overall, these two specifications certainly show that the result of this study is robust.

Table 9

Ownership structure and performance and risk: Lagged Dependent variables

Efficiency, Income, profit and expenses as indicator of bank’s performance, whereas Loanloss as indicator of bank’s risk. Panel time fixed effect is used. P-values in paranthesis.

1 2 3 4 5

Efficiency Income Profit Expenses LoanLoss Constant 0.2534*** -0.0070* -0.0115*** 0.0097* 0.0092*** (0.0000) (0.0683) (0.0098) (0.0607) (0.0000) GOBs 0.0186* 0.0089*** 0.0077*** 0.0090*** -0.0004 (0.0975) (0.0000) (0.0000) (0.0070) (0.3997) Listed -0.0219 -0.0042*** -0.0084*** 0.0003 -0.0023*** (0.0997) (0.0000) (0.0000) (0.7315) (0.0027) Size 0.0109* 0.0048*** 0.0063*** -0.0017*** -0.0016*** (0.0600) (0.0000) (0.0000) (0.0060) (0.0000) Loans 0.0341*** 0.0029** 0.0028* 0.0209*** 0.0112*** (0.0000) (0.0262) (0.0890) (0.0000) (0.0000) Deposits -0.0663*** -0.0062** -0.0038 0.0014 0.0002*** (0.0020) (0.0268) (0.2460) (0.7562) (0.8630) Capital 0.4139*** 0.0589*** 0.0552*** 0.0055 0.0001 (0.0000) (0.0000) (0.0000) (0.2217) (0.9735) Loanloss 0.4199 0.0443* 0.1683*** 0.1654*** (0.1078) (0.0915) (0.0000) (0.0000) efficiency (-1) 0.4898*** (0.0000) Income (-1) 0.1818*** (0.0000) Profit (-1) 0.2964*** (0.0000) Expenses (-1) 0.5458*** (0.0000) Loanloss(-1) 0.3379*** (0.0000) Loan(-1) -0.0042*** (0.0000) GOBs A dummy variable that equals 1 if the banks is Goverment owned banks and zero otherwise

Listed A dummy variable that equals 1 if the banks is listed and zero otherwise Income Ratio of operating income per earning asset

Profit Ratio of net income per total earning asset Efficiency score Based on DEA calculation

(25)

VII. Conclusion

The results of this study indicate that Government Owned Banks are more profitable and more efficient than Privately-Owned Banks. In addition, Government Owned Banks take less risk and only pick the best loans. The growth of Regional Development banks provides the major contribution to the performance of Government Owned Banks in Indonesia. Most of the Government Owned Banks in Indonesia are categorized as the largest bank in Indonesia; therefore a size effect really matters in influencing efficiency and performance of Government Owned Banks in Indonesia. This result strongly supports the social view of the role of government ownership and supports the previous research by Bhattarcharya et al. (2003) and Isik and Hassan (2003). Likewise, this study is in contrast with the study by Altunbas et al. (1994) and Ianotta et al. (2007)

Regarding the effect of ownership concentration on risk and performance, this study shows that listed banks are less profitable and less efficient than non-listed banks. Moreover, listed banks will take less risky loans. Separation between agent and owner might influence the risk taking behaviour of listed banks. This study gives a different result compared to the study by Isik and Hassan (2003) and Ianotta et al. (2007) who find that listed banks are more profitable and more efficient that non-listed banks.

Naturally, there are several limitations to this study. First, due to data availability, this study does not include all banks in Indonesia. Second, I do not take into account macro economic variables that might influence the banking risk and performance. It can be a suggestion for further research to take into account macroeconomic variables in the model estimation. Third, I only use the Loanloss variable as the main proxy for bank risk, while this variable is mainly calculated from accounting information. It might be better to use a risk indicator based on market data, such as insolvency risk by using market prices of a bank’s stock. Fourth, Efficiency calculation uses the non-parametric approach. For theoretical reasons, it might be better to use the parametric approach as well. As such, we can see whether different approaches will yield different results or not.

(26)

than the mean of the world efficiency score. Related to issues of privatization, Governments tend to privatize firm based on an efficiency argument, but the decision to implement the policy of privatization needs a very serious and rigorous assessment. Finally, on a more speculative note, I expect that the increasing growth of the Indonesian banking industry will foster the economic development in Indonesia.

References:

Alchian, A.A., 1965, Some economics of property rights, Il Politico, 816-829

Altunbas, Y., Molyneux, P., 1996, Economies of scale and scope in European banking, Applied Financial Economics 6, 367–375.

Altunbas, L. Evans and P. Molyneux., 2001, Bank ownership and efficiency, Journal of Money, Credit and Banking 33, pp. 926–954.

Altunbas, Santiago C, Edward PMG and P Molyneux., 2007, Examining the relationships between Capital, Risk and Efficiency in European Banking, European Financial Management 13(1), 49-70

Anandarajan A, Hasan I, Lozano-Vivas A., 2005, Loan loss provision decisions: An empirical analysis of the Spanish depository institutions, Journal of International Accounting, Auditing and Taxation 14, 55-77.

Barr, Richard, K. Killgo, F. Siems and S. Zimmel., 2002, Evaluating the Productive Efficiency and Performance of U.S. Commercial Banks, Managerial Finance 28, 3-25 Barth, J.R., Nolle, D.E., Phumiwasana, T., Yago, G., 2003. A cross-country analysis of bank

supervisoryframework and bank performance. Financial Markets, Institutions and Instruments 12, 67–120.

Bearle Jr., A.A., Means, G.C., 1932. The Modern Corporation and Private Property. Macmillan, New York.

Berger, A.N., Hanweck, G.A., Humphrey, D.B., 1987. Competitive viability in banking: scale, scope, and product mix economies. Journal of Monetary Economics 20, 501– 520

Berger, A.N., & Timothy H. Hannan, 1993. Using efficiency measures to distinguish among alternative explanations of the structure-performance relationship in banking, Finance and Economic Discussion Series 93-18, Board of Governors of the Federal Reserve System (U.S.).

Berger. A.N., Humphrey, D.B., 1997. Efficiency of financial institutions: International survey and direction for future research. European Journal of Operational Research 98(2), 175-212

(27)

Bhattacharyya A, Lovell CAK, Sahay P., 1997, The impact of liberalization on the productive efficiency of Indian commercial banks, European Journal of Operational Research 98, 332-345.

Claudio E. V. Borio, 2003, Towards a macroprudential framework for financial supervision and regulation?, BIS Working Papers 128, Bank for International Settlements.

Casu, Barbara & Philip Molyneux, 2003, Acomparative study of efficieny in European banking, Applied Economics, Taylor and Francis Journals, vol. 35(17), pages 1865-1876,

Cestona, MG, Surroca, J., 2008, Multiple goals and ownership structure: Effect on the performance of Spanish savings banks, European Journal of Operational Research 187,582-599

Charnes A, Cooper WW, Rhodes E. , 1978, Measuring the efficiency of decision making units, European Journal of Operational Research 2, 429-444.

Cornett, M.M., Guo, L., Khaksari, S., Tehranian, H., 2000. Performance differences in privately-owned versus state-owned banks: An international comparison. Working paper, World Bank, Southern Illinois University at Carbondale, Suffolk University and Boston College.

Dermirguc-Kunt, A., Huizunga, H., 1999.Determinant of commercial bank interest margin and profitability: some international evidence. The World Bank Economic Review 13,379-408

Greene W.H., 1993, The Econometric Approach to Efficiency Analysis.” In H. Fried, C.A.K. Lovell

Hadad, M., Sugiarto, A., 2003, Kajian mengenai kepemilikan bank di Indonesia, Bank Indonesia Working Paper.

Hart, Oliver & Shleifer, Andrei & Vishny, Robert W, 1997, The Proper Scope of Government: Theory and an Application to Prisons," The Quarterly Journal of Economics, MIT Press, vol. 112(4), pages 1127-61,

Hughes, J.P., Mester, L.J., 1998, Bank capitalization and cost: evidence of scale economies in risk management and signaling,The Review of Economics and Statistics 80, 314–325. Iannotta G, Nocera G, Sironi A, 2007, Ownership structure, risk and performance in the

European banking industry, Journal of Banking & Finance 31, 2127-2149.

Isik I, Hassan MK., 2003, Efficiency, Ownership and Market Structure, Corporate Control and Governance in the Turkish Banking Industry, Journal of Business Finance & Accounting 30, 1363-1421.

Jensen, M., Meckling, W., 1976, Theory of the firm: Managerial Behavior, agency cost and ownership structure, Journal of Financial Economics 3, 305-360

(28)

La Porta, R., F. Lopez-de-Silanes and A. Shleifer, 1999, Corporate Ownership around the World, Journal of Finance 54, 471-517.

Lewis, Ben W, 1935, Berle and Means on the modern corporation, Journal of Political Economy 43, 548-554

Loderer, Claudio, and Urs Waelchi., 2006, Protecting minority shareholder :Listed versus unlisted firms, Finance working paper ECGI no 133.

Lovell, C.A.K, 1993, Production Frontiers and Productive Efficiency. In H. Fried, C.A.K. Lovell, and S.S. Schmidt (eds.), l&e Measurement of Productive efficiency Techniques and Applicarims. New York: Oxford University Press.

Majumdar, Sumit K., 1997, The Impact of Size and Age on Firm-Level Performance: Some Evidence from India, Review of Industrial Organization 12 (2), 231-241

Macey, J.R., and Ohara, M., 2001, the corporate Governance of Banks, Federal Reserve Bank of New York Economic Policy Review

Mester LJ., 1993, Efficiency in the savings and loan industry., Journal of Banking & Finance 17, 267-286.

Mester LJ., 1996, A study of bank efficiency taking into account risk-preferences, Journal of Banking & Finance 20, 1025-1045.

Rajan, Raghuram G., Zingales, Luigi and Kumar, Krishna B., 2001, What Determines Firm Size, CRSP Working Paper No. 496 , http://ssrn.com/abstract=170349

Rivas, Andres, Teofilo Ozuna, Jr. and Felice Policastro, 2006, Does the Use of Derivatives Increase Bank Efficiency: Evidence from Latin American Banks, International Business and Economics Research Journal 5, 47-56.

Sarkar,Jayati., Sarkar, Subrata., and Bhaumik, Simon., 1998, Does ownership always matter-Evidence from Indian banking insustry, Journal of comparative economics 26, 262-281

Sapienza P., 2004, The effects of government ownership on bank lending, Journal of Financial Economics 72, 357-384.

Sathye M., 2003, Efficiency of banks in a developing economy: The case of India, European Journal of Operational Research 148, 662-671.

Sherman HD, Gold F, 1985, Bank branch operating efficiency : Evaluation with Data Envelopment Analysis, Journal of Banking & Finance 9, 297-315.

Shleifer A, Vishny RW., 1997, A Survey of Corporate Governance, The Journal of Finance 52, 737-783.

Welch, Emma. 2003., The relationship between ownership structure and performance in listed Australian companies, Australian journal of Management Vol 28 No 3