How ownership affect bank performance

(1)

Faculty of Economic and Business Faculty of Social Science

— Evidence from BRIC countries

Master Thesis – MSc. International Financial Management

Yue Ma s1996533

Supervisor: Dr. B. Qin Co-assessor: Dr. H. Gonenc

(2)

1

Abstract

The main idea of this study is to identify whether any difference can be observed regarding the performance of commercial banks in BRIC countries. To this, a two-tie comparison of bank performance has been investigated between state-owned versus private-owned banks and foreign-owned versus domestic-owned banks. In particular, I first apply Data envelopment analyses method to compute bank efficiency score which used as the main indicator of bank performance. Besides, accounting based measure, Earnings before interest and taxes also introduced as additional dependent variable. Empirical analyses has been conducted by using data of 207 commercial banks in BRICs during the period of 2000 - 2006, I find that ownership matters. More specifically, government-owned banks appear to be less efficient and less profitable than their private-owned counterparts, while foreign-owned banks are superior to domestic-owned banks in terms of bank efficiency.

(3)

2

1. Introduction

The previous studies of how ownership structure affects corporate performance can mainly specified into two categories: one is the concentration of ownership, firm can be recognized according to the degree of concentration, whether a firm’s ownership is concentrated or diffused, the fraction of shares owned by the inside or the outside shareholder, these factors will influence the decision making of those shareholders and hereby, the performance of the firms (Jensen & Meckling, 1976; Bearle & Means, 1932; Iannotta et al., 2007). The second one, that’s what I try to investigate in this paper, is the nature of the shareholder (Shleifer & Vishny, 1995; Shleifer, 1998; Cornett et al., 2010). To be more specific, the firm may be owned by government, domestic private investors, or otherwise, foreign investors. The above categories of owners operate firm with different approaches and incentives; thereby it is really interesting to identify which type of enterprise is with the best performance.

I choose banking sector in BRIC countries as sample to explore how the nature of owners influences bank performance. There is no doubt that banking industry serves a central role in contemporary economy and financial market, both the external environment and internal factors affect their way of doing business and outcomes as well. Levine and Sara (1998) provide evidence that highlight the crucial importance of bank efficiency for the sustainable development of an economy. Moreover, as banking sector has undergone its great transformations during last decades, the analysis of bank performance is of great significance for bank management, government as well as the financial market.

(5)

4

heat up the internal competition; external factors like technology innovation and financial crisis enforce banking industry to consider whether their old way of doing business is good enough to stand up to those new challenges. On the other hand, domestic industry especially in emerging market benefit a lot from financial liberalizations (e.g. easy access to international financial market, the import of innovative techniques and financial instruments). Those mix impacts also impel banking industry in BRICs become an eye catching market.

Concerning the performance of government owned banks, development view implies that state-owned banks suite best in dealing with market failure, with the aim of maximizing broader social objectives, government-owned banks are of great importance for financial and economic development (Gerschenkron, 1962 ), while few empirical researches support this argument. In contrast, we saw more scholars argue that state-owned banks are used as a mechanism to purse private interests (La Porta et al., 2002; Micco et al., 2007; Cornett et al., 2010). Perhaps the agency view and political view can better explain this circumstance; both of the two theories link government-owned banks with misallocation of resources, but with different intentions.

On the other hand, as there was an increasing trend of privatization in banking sector, studies arise to analyze the reason for the fundamental reassessment of government ownership and whether privately owned banks have better performance than that state-owned peers (Megginson, 2005; Clarke et al., 2005; Ianotta et al., 2007; Micco et al., 2007). Moreover, the financial liberalization that opens up for foreign investments also accelerates the achievement of privatization. The growing foreign ownership stimulate the host country economy and speed up cross-border competition among domestic banking market, it will be an attractive topic for scholars to explore how they perform and get a slice of the promising market.

Hereby, the research question of this paper is:

Whether different types of bank ownership can explain different levels of bank performance in BRIC countries.

(6)

5

bank, then I introduce them as well as earnings before interest and taxes as bank performance measures to conduct empirical test. I use data of 207 banks in BRIC countries during 2000 to 2006. After controlling bank characters, country specific and time specific effects, OLS pooled regression has been applied to test performance difference between government-owned banks and private-owned banks, foreign-owned banks and domestic-owned banks. The empirical results reveal that the performance of government-owned bank is inferior to private-owned banks both from efficiency and profit perspectives, while foreign-owned banks is superior to their domestic peers in terms of bank efficiency, no significant difference is detected regarding profitability between to the two subgroups.

My study is in line with the agency and political views suggested that government-owned banks were not performed as well as they planned, mainly because they serve more for other self-purposes instead of the financial market. while the bank privatization improve the bank performance with their superior monitoring and incentive mechanism in dealing with agency problems as well as their clear profit maximization target, which in turns, prevent banks from engaging in political affairs. Another finding of my study that foreign-owned banks associate with higher efficiency than their domestic peers provide additional support for the global advantage hypothesis, which claims that foreign-owned banks can mitigate the distance rendered problems by operating their better management and technological skills.

(7)

6

instead of comparing merely state-owned and private-owned banks (Shleifer, 1998; Cornett et al., 2010), a two-tie comparison has been investigated which take foreign ownership into consideration, the aim is to exam the influence of growing foreign investors since the financial liberalization.

The rest of this paper is organized as follow: Section 2 provides a general introduction of the banking industry in BRIC countries; Section 3 reviews the main theories related with my study followed with empirical evidence conducted by previous scholars as well as hypotheses development; Section 4 is about data source and description; Section 5 explains the methodology I applied for analyze. Section 6 specifies how I conduct empirical test followed with results interpretation as well. And finally are the conclusion and limitation of this paper.

2. Background —banking industry in BRIC countries

The recent ten years have seen remarkable changes of banking industry among BRIC countries. This section provides an overview of BRICs banking industry merely regarding the ownership structure and the how they transform during recent decades. Those changes consist with the deregulation of state-owned banks, the privatization of centralized ownership and the reform of financial liberalization, result in increasing competition and industry transformation.

2.1 Bank deregulations

(8)

7

the government-appointed Narasimhan committee, which include the deregulation of interest rates on the deposits and lending and easy entry restrictions into the banking industry (Sanyal & Shankar, 2011). To a large extent, those changes severed the competition on banks sector among emerging countries and result in dramatically transformations in the ownership structure of banking sector.

2.2 Privatization

A trend of transformation from state-owned to private-owned banks began with 1990s among emerging countries. One primary reason is that compelling evidence arise to reveal the poor performance of government-owned banks, e.g. intermediation costs and huge portfolio of non-performing loans (Megginson, 2005); another reason is a gradually perception that the government lead banking system is associate with slower pace of financial development.

(9)

8

Note: GOBs refer to banks that more than 50% of equity is held by the government. POBs refer to banks that more than 50% of equity is held by private sectors.

Source: Vernikro, 2007; Staub, et al., 2010; Gupta, et al., 20111.

2.3 Foreign entry

Financial liberalization on the international level results in a growing presence of foreign ownership in emerging domestic banking market. This is driven by self-requirements of this industry and a series of global forces, e.g. the need of financial development and economic growth, the large growing cases of international trade and foreign investment activities and the aim of expand international market of bank strategies. As a result, we saw an increasing number of countries remove the entry restriction and open the door for foreign institutions, especially in developing countries.

Figure 2 illustrates the shares of bank total assets (by percentage) that controlled by foreign and domestic owners respectively. Brazil and experienced a most distinctive rising from 6% in 1990

1

The data of China is for author’s calculation.

Figure 1: Share of banks in total assets (as %) in 2006 by Ownership Groups

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Brazil Russia India China

(10)

9

to 27.5% at the end of 2006; Russia increasing from 6% to about 10% during this period. It should be highlight that China operated with no foreign-owned banks in 1990 and experienced dramatic present of foreign investors since the entrance of World Trade Organization in 2001(reaching 10% of occupation in 2006); India illustrates a slowing pace of growing (from 5% to 8%).

Note: Foreign-owned banks refer to banks that more than 50% of equity is held by foreign investors.

Source: Hawkins & Mihaljek, 2001; Vernikro, 2007; Staub, et al., 2011; Gupta, et al., 20112.

3. Literature review and Hypotheses development

3.1 The changing of ownership structure and bank performance

The relevance of arguments regarding firm ownership has been extensively investigated in theoretical literature (Shleifer & Vishny, 1997; Shleifer, 1998). This section focus on theories regarding different categories of bank ownership followed with relevant empirical evidence to explore how it works in practice.

2

The data of China in 2006 is for author’s calculation.

Figure 2: Share of total assets controlled by foreign-owned banks (as %)

0 10 20 30 40 50 60 70 80 90 100

Brazil Russia India China

(11)

10

Theories in favor of state-owned banks

Government ownership enjoyed special favor during half century ago due to the existence of market failures, as pointed out by Shleifer (1998), ―where monopoly power, externalities, or distributional issues raise concerns, private profit-maximizing firms may fail to address these concerns, whereas the aim of state-owned firms is to allocate resource to general welfare to achieve social objectives maximization‖. This is true in banking industry as well. Alexander Gerschenkron (1962) initially introduces development view to describe this trend, the author point out that privately owned banks have been justified in industry countries as their advantage of channeling savings into industry, but for countries where financial institutions is not sufficiently developed, state ownership is more favorable on the ground that it better satisfied the general welfares by directing the collected savings to foster long term projects, while at the same time promote financial development and economic growth. Gerschenkron then take Russia as an example, at the time privately owned banks have inadequate funding and risk tolerance to afford long-term operation, especially in an economy where ―fraudulent bankruptcy had been almost elevated to the rank of a general business practice‖ (Gerschenkron, P.20). It is the government that fulfills the role of supporting bank system with sufficient capital for long term standing and growth.

Theories against state-owned banks

On the other hand, Theories opposite with development view claim that state-owned banks doesn’t work as expected, mainly due to their behavior of misallocating resource for pursuing self-interests.

(12)

11

working in the best interest of their principals, managers are more inclined to pursue their personal benefits on the expense of banks’ higher expected social welfares. Specifically, this can be driven by following reasons: Firstly, public managers are less likely to get performance based compensation, in the absence of an close tie for connecting managers’ effort and productivity to observable firm performance, managers have less incentive and less diligent in serving the owners’ interests (Meulbroek, 2000; Megginson, 2005); Secondly, unlike shareholders who seek for higher risk with maximum value, managers are generally risk aversion. This is especially true in state-owned banks as the rewarding mechanisms encourage safe behavior. When concerning their performance and career development, public managers are more likely to forgive projects with higher risk but also higher expected returns. Moreover, as managers generally engaged in short-term employment contract, they consider more on the firm’s short term financial target instead of long-term achievement, it is in this context that the decision making of managers will not serve the best interests of the enterprise.

On the other hand, the monitoring mechanism in state-owned banks generates highly agency cost but often with little effect. This is because principals in state-owned banks have no clear responsibility for monitoring (Clarke et al., 2005); the real implementation cannot be guaranteed. Another reason is that state-owned banks are more distributed than their privately peers because all citizens of a countries jointly owned the state-owned banks, as individual citizen have no right to sell their shares in a state-owned banks, they earn less from monitoring performance, and therefore with less motivations (Alchian, 1965).

(13)

12

La Porta et al., 2002). The state-owned banks China, as an example, generally exhibit a higher level of non-performing loan ratio, predominantly because central and local governments manipulated them to engage in negative net present value policy loans for political interests, This in turn result in the depression of the assets qualities (Dobson & Kashyap, 2006). However, after two major legislative reforms occurred in 19953, they transformed more toward commercial business based on market principles instead of policy-lending activities (Berger et al., 2009).

State-owned banks in practice

It is clear that state-owned banks used to be widespread, especially in developing countries. However, they have been experienced a decrease over the last two decades. At this time, massive empirical evidence arises with the arguments that the behavior of state-owned banks is inconsistence with development theory rendered efficiency justification.

Cornett et al. (2010) in their study find that government-owned banks maintain obviously higher levels of government securities to total assets than that of private-owned counterparts’ in countries associate with heavily government involvement. Their finding indicates that instead of financing private projects to enhance social welfare, state-owned banks play a more active role in financing the government. Using data of bank ownership around 92 countries, La Porta, Lopez-de-Silanes and Shlerer (2002) find government ownership of banks is still common around the world till 1995 and win special favor in countries with underdeveloped financial system and heavy government intervention in the economy. In contrast with the development view motivation, their finding of higher government ownership of banks associate with lower productivity support the political theory on the effects of government interference in markets. Sapienza (2004) focus on the lending activities of the two types of banks in Italy market. His study shows that, contrary to the social welfare maximizing view, the lending behavior of state-owned banks favor mostly large companies. Moreover, the author’s evidence of higher involvement of the political party associate with cheaper the interest rates charged support the argument that government ownership of banks has distorting effects on financial allocation of

3_{The 1995 Central Bank Law of China confirmed PBOC as the central bank and substantially reduced the influence}

(14)

13

resources. This is consistence with the finding that resource allocation from a political perspective reduces bank efficiency and hinders economic development (La Porta et al., 2002).

Trends towards privatization

Supported by compelling evidence, the era that highlight strategic role of national commercial banking in funding the nation’s economic development is of the past. As a result, a western country-leads privatization aim at divesting government ownership has been launched since then.

Compared with state-owned banks, the private counterpart enjoys a competitive position for the following reasons. Firstly, as the primary goal of private-owned banks are costs minimization and profits maximization, the incentives of managers closely meet banks’ financial achievement and thus weak the agency problem to a large extent. Secondly, the conflict between shareholders and managers can be minimized by enhancing the monitoring level of managers’ actions, due to the easier way of getting managers’ private information by shareholders (Barry et al., 2010). Finally, to some extent the most important, privatization is the optimal way to prevent politician from pursuing their political goal or self-interests (Megginson, 2005; Clark et al., 2005).

(15)

14

theories to explain states’ choices, that is, states rely much on budget transfer from the federal government, and government manipulate state banks as development agencies and those whose banks were already under federal control are more likely to giver control to federal government.

On the other hand, it will be too extremely to say that there is no researches exist that go opposite with above arguments. Empirical researches conducted by previous scholars regarding bank ownership and performance in each of the BRIC countries are list in Appendix A. One study found that go against with above arguments was carried out by Bhattacharya, Lovell and Sahay (1997), the authors analyze bank efficiency in Indian market during the 1986 to 1991, and they found during the sample period state-owned banks perform with most efficient, followed by foreign-owned banks and private-owned banks at last. As state by Megginson (2005), this is surprising as it show dramatic opposite result with that of non-financial enterprises in India. Berger et al. (2004) argue that this phenomenon may partially explained by the accounting practices of government auditors.

In sum, although opposite evidence exist, those analysis are focus on more far back period than my study; More importantly, considering the academic evidence that favor the performance of private-owned banks account for clearly large proportion, I assume the hypotheses as follow:

H1: The performance of state-owned banks is inferior to that of privately-owned counterparts.

The effect of foreign banks entry

(16)

15

When concerning the performance of domestic banks versus foreign banks, Berger et al. (2000) employ comparative advantage hypotheses to explain the exist differences between the two sub-groups. Under the home field advantage hypothesis, domestic banks should be superior to their foreign owned peers on the ground that organizational diseconomies to operating and monitoring in a long distance result in operating problems. Apart from distance, other barriers like differences in language, culture and regulation also put foreign banks in a lower position.

In contrast, the global advantage hypothesis holds the opinion that foreign banks have the ability to overcome these cross-border disadvantages and exhibit to be more efficient compared with their domestic counterpart. It is not difficult to observe that foreign investors are generally from developed countries which are strong enough to expand their cross-border market. As a result, they always subject to superior managerial skills which allow them providing more favorable and high quality services for customers; their better investment and risk management skills enable them to get higher returns for investment; the broader ways of getting resources also result in lower costs, Those advantages in turns make it possible for foreign-owned banks exhibit a better performance.

The study conducted by Demirguc-Kunt & Huizinga (1999) is the most representative one. After controlling bank specific characteristics and macro environment leverage, the authors found that foreign banks have lower margins and profitability compared to domestic banks in developed countries, while the opposite holds in developing countries. They further explained this finding may reflect that in developing countries, a foreign bank’s technological edge is relatively strong and strong enough to overcome any informational disadvantage. This situation can also be explained by lower tax and provisioning compared with domestic banks.

(17)

16

than other foreign- owned banks. Thereby, they drew the conclusion that instead of facilitating the transfer of technology and modern bank management, international institutional investors are more interested in cherry-picking banks with high financial achievement.

It should mention that previous researches go mix results, there is also empirical evidence that favor home field advantage hypothesis (Staub, Souza and Tabak, (2010) list in Appendix A), that’s why this line of research needs further exploration. Considering the empirical evidence that support foreign-owned banks stand obviously a larger proportion, my study follow the argument of Demirguc-Kunt & Huizinga (1999) and hereby assume the second hypothesis as follows:

H2: The performance of foreign-owned banks is superior to that of domestic-owned peers.

4. Methodology

In order to explore the correlation between ownership structure and bank performance, two steps have been considered in this process. First, to apply data envelopment analyses to calculate bank efficiency score, which used as the main indicator of bank performance in this study. Then an Ordinary Least Squares (OLS) model is employed to examine how bank performance explained by different types of ownership structure.

4.1 DEA approach to estimate bank efficiency

(18)

17

In the DEA model, each bank represented by one Decision Making Units (DMU), the aim of DEA is to construct a non-parametric envelopment frontier based on the provided data points (that is the DMU) such that all observed points lie on or below the production frontier (Coelli, 1996). In other words, the purpose is to compute a maximum efficient score for each bank by measuring one DMU relative to all others.

According to Charnes et al. (1978)’ definition, each DMU is measured by the maximum ratio of weighted outputs to weighted inputs4, subject to the condition that the similar ratios for all other DMUs ≤1 (Sathye, 2002). Assume that J outputs and K inputs are applied for each of the N banks, the efficiency score of bank i can be expressed as:

Max EFFi = (∑

u

ij

y

ij ) / (∑

v

ik

x

ik) (1) Subjest to: (∑

u

i j

y

nj) / (∑

v

ik

x

nk) ≤1 n= 1,…..N （2）

u

i j,

v

ik ≥0 ; j=1,….J; k=1,……K.

y

n

j is a proxy of output matrix for bank n;

x

nk is a proxy of input matrix for bank n, n=1,...N;

u

i

j

and

v

ik are the variable weights that determined by the liner programming (1). Each bank be

evaluated with an efficiency score E (0 ≤ E ≤1), E close to 0 implied that the bank is relatively less efficient; E close to 1 describe bank with high efficient; E equal to 1 means the banks is the best performed among the entire sample.

The calculating process was conducted by DEAP Version 2.1 program, variable returns to scale (VRS) input - oriented model has been selected. This paper select input and output variables based on intermediation approach which identify banks as financial intermediaries and employ volume of deposits, loans and other variable as inputs and outputs (Sathye, 2002), the exact set variable also consider the data availability. Herby, the input variables include total deposits, interest expense, and personal expense, while the outputs variables include total loan and interest

4_{The outputs are the primary business of the organization, that is, the main services that banks provided; while the}

(19)

18

income. It should be mentioned that the efficiency scores were computed by each of the four countries in order to avoid the country specific influence.

4.2 Regression analysis

A set of panel data is introduced to conduct empirical research, as panel data provide information across section while at the same over times; it allows researchers identify bank performance from cross-broader while take time effect into consideration as well. The Pooled Ordinary Least Squares approach (POLS) model is employed to test the hypotheses, which is widely used by previous researchers (Rivas et al., 2006; Patti & Hardy, 2005). As far as other models for conducting empirical research of bank performance: GLS (Boubakri et al., 2005), GMM (Ataullah & Le, 2006), or Tobit (Casu & Molineux, 2003), the last one is widely applied on the ground that the efficiency score calculate by DEA method censored around zero to one, any other non-censored estimations will lead to biased results. While Ataullan and Le (2006) as well as other researchers point out the biased result can be avoid by taking the natural logarithm of [efficiency score/(1- efficiency score)] (Fethi & Pasiouras, 2009). As a result, in my study OLS model has been applied after rewrite EFFICIENCY in natural logarithm form.

The model I applied follows the research of Iannotta et al., 2007 and Micco et al., 2007:

Pi,t= α+ß0*GOBi+ß1*FOBi+ ß2Year t + ß3 Countryi+ ß4*CVi,t+ εit

Where Pit is the measure of bank performance for bank i in year t; α is the constant variable;

Explanatory variables include a Dummy variable GOB for state-owned banks and a foreign ownership dummy variables FOB to capture foreign-owned banks in equation. Yeart and

Countryi are introduced to control time and country fixed effect respectively. CVi,t is a measure

(20)

19

The independent measurements consist with two dummy variables to reflect the ownership groups:

A Dummy variable GOB for Government-owned banks and a foreign ownership dummy variables FOB to capture foreign-owned banks. As this paper focus on the impact of different types of ownership on bank performance, only banks with constant ownership during the sample period have been considered. It should be mentioned that the definition of government-owned banks is follow the research of Micco et al., (2007) to classify a bank as government-owned if the bank ultimate owner5 is either a national or local government; otherwise, the bank should be classified as private bank. The same goes with the definition of foreign-owned banks.

The dependent variable is bank performance, Efficiency score6 and EBIT (earnings before interest and taxes) were applied respectively aim at reflect bank performance from two dimensions. The first measure is introduced as the main indicator of bank performance from the efficient aspect, which is calculated by using a mathematical programming techniques — Data Envelopment Analysis method (describe in later process), while the later one, EBIT, is employed as financial return measure that reflect bank profitability without influenced by tax effect. It is widely used in academic researches (Demirguc-Kunt & Huizinga, 1999; Micco et al., 2007; Bonin et al., 2004).

The country specific and bank specific control variables are list as follow:

GDP: the national GDP growth rate was choose to control country effect. Based on the argument that banks located in countries with faster growth are associate with better performance and higher competitive (Bonin et al., 2004), I suppose better performance can be detected for banks located in higher GDP growth rate.

SIZE: the logarithm of total assets is employed to capture scale bias (Berger et al., 2005; Micco et al., 2007; Bonin et al., 2004). Based on the arguments that larger banks associate with larger

5_{The ultimate owner refers to the shareholder own more than 50% of the bank’s equity capital with no other single}

shareholder owning a larger share (Micco et al., 2007).

(21)

20

market power (Maudos & Magore, 2005) and enjoy a higher level of cost advantage (Iannotta et al., 2007). I expect a positive correlation between size and bank performance.

CAPITAL: The ratio of book value of equity to total assets account for bank capitalization (Demirgus-Kunt & Huizinga, 1999; Iannotta et al., 2007). Bank with well-capitalized need to borrow less to afford a certain level of assets (Demirguc-Kunt & Huizinga, 1999), I expect capital ratio is positive correlated with bank performance.

LOANS: loans to assets ratio is introduced for capturing loan portfolio orientation (Cornett et al, 2010; Barry et al., 2010; Zhang, 2009). As loan is relatively more profitable compared with other types of assets, a positive relationship between loans and bank performance is expected.

DEPOSITS: deposits to total assets (Barry et al., 2010; Bonin et al., 2004; Iannotta et al., 2007). Due to the reason that the more deposits, the more expenses will arise, I expect banks with larger deposits enjoy less efficiency and profitability.

NONINT: the ratio of non-interest income to total assets (Demirgus-Kunt & Huizinga, 1999; Micco et al., 2007). NONINT is included to proxy for bank income produced from non- lending activities since we have seen an increasing trend of commercial banks engaged in fee-based services. No certain sign of coefficient has been assumed for this variable.

Followed the research of Iannotta et al.(2007) and Micco et al. (2007), year dummy and country dummy was include in the regression in order to control the time specific and country specific effects.

Year dummy: D2000, D2001, D2002, D2003, D2004, D2005, D2006, each of them take value one during the correspondent year and zero otherwise7.

Country dummy: D_bra, D_rus, D_ind, D_chi8, each of them takes value one for correspondent country and zero otherwise.

(22)

21

5. Data

5.1 Data source

In this paper, bank-level annual data and ownership information are mainly obtained from the Fitch-IBCA Bank Scope (BSC) database, which provides bank-level annual financial information for 179 countries around the world. The sample contains commercial banks in Brazil, Russia, India and China that cover the years 2000-2006. The government and foreign ownership information for each bank is hand collected by checking the ultimate shareholder from the section of ―shareholder information‖ or through other sources as well.

Before conduct statistical analysis the dataset require further editing and improving. The criteria of sample selection are: Firstly, due to this paper focus on commercial banks, the original sample includes 1592 banks after excluding banks belongs to other types. The second requirement is to ensure banks provide at least one year financial information, banks missing too much crucial information and observations are also dropped, The remaining sample consists of 773 banks; the last requirement, which is the most important, is that banks with clear ownership information and stable ownership structure during the sample period I analyses. I identified bank ownership via all kinds of sources, however, still some banks are hard to be coded, and thereby be removed from the dataset. Moreover, considering other effects that may influences the outcome, banks that changed their ownership type during the sample period e.g. banks that went bankruptcy or be taken over are excluded as well. Hereby, the final dataset end up with 207 commercial banks from the four countries for a total of 1449 bank-year observations. Table 1 presents the process of data selection.

(23)

22

Table 1 Summary of sample selection

5.2 Data description

Table 2 gives a brief view of the entire sample adopted in this paper. The mean and Stand deviation of main variables are provided classified by regions. The value of continuous variables on the 1, 25, 75 and 95 percentiles are also reported. For those banks in my sample, private-owned banks account for majority part. In case of the foreign and domestic differentiation, the number of banks nearly the same for the two sub-groups.

The dependable variable EFFICIENCY range from about 0.33 to over 0.52, while the average EBIT of the whole sample is 1.86, which close to the research of Micco et al.(2007) with mean ratio of 1.36 during the period of 1995-2002. The ratio of China and India remains relative lower than that of all samples, with the mean value of 0.79 and 0.95 respectively. Regarding other controlling variables, the average loans to total assets ratio is 0.57, the average deposit to total assets ratio is 0.80, and the average ratio of non-interest income to total assets is 0.02. Obvious difference exist in capital ratio among the four countries, with a stand deviation of 15.11, which caused by the substantial low ratio of India banks, merely one third of that of other countries. The controlling variables are comparable with previous researches in the field (Clarke et al., 2005; Iannatto et al., 2007).

9_{The reason of the final dataset end up with a substantial fewer observations than the original raw sample is that the}

information of commercial banks in developing countries are not sufficient enough on Bankscope database.

Commercial banks in BRIC countries 1592

Less: banks unable to provide sufficient financial information 690

Less: banks missing too much crucial data ₁₂₉

Less: banks have no clear classification of bank ownership type

566 Less: banks change their ownership type during the sample period

Total: banks included 207

(24)

23

Table 3 illustrates the correlation matrix of the independent variables adopted in the regression. As can be seen from the table, higher correlation can be only observed between capital ratio and total assets as well as the two dummy variables GOB and FOB10. The former can be explain by larger banks process larger amount of deposits and thereby less capitalized than smaller banks. The two dummy variables GOB and FOB will separately and jointly introduced into the regression in order to avoid collinearity bias.

10_{As suggested by Cohen (1988), a coefficient value in excess of 0.5 is an indication of higher correlation.}

Table 2 Descriptive statistics of variables applied in this study

Brazil Russia India China All 1th 25th 75th 99th

Variable 1 2 3 4 5 6 7 8 9

Independent variable

Dummy GOB No.GOB 8 10 20 6 44

No.POB 68 66 10 19 163

Dummy FOB No.FOB 38 47 11 16 102

(25)

24

Table 3 illustrates the correlation matrix of the independent variables adopted in the regression. As can be seen from the table, higher correlation can be only observed between capital ratio and total assets as well as the two dummy variables GOB and FOB11. The former can be explain by larger banks process larger amount of deposits and thereby less capitalized than smaller banks. The two dummy variables GOB and FOB will separately and jointly introduced into the regression in order to avoid collinearity bias.

Figure 3 shows the mean value of dependable variables EBIT and EFFICIENCY with different types of ownership over years from 2000 to 2006.

It can be seen from Figure 3 that the ratio of EBIT for GOBs is lower than that of POBs for the whole sample period. Even though the fluctuation of EBIT for POBs is more dramatic, its lowest point in 2003 is relatively higher than its’ POBs peers. The same trend can be seen also from the movement of EFFICIENCY, which the ratio of GOBs wave below that of POBs all the time.

11_{As suggested by Cohen (1988), a coefficient value in excess of 0.5 is an indication of higher correlation.}

Table 3 Correlation matrix among variables

FOB GOB CAPITAL DEPOSITS GDP LOANS SIZE NONINT

FOB 1 GOB -0.511*** 1 CAPITAL 0.191*** -0.216*** 1 DEPOSITS -0.081*** 0.122*** -0.357*** 1 GDP -0.021 0.201*** -0.111*** 0.181*** 1 LOANS 0.090*** -0.028 0.040 0.100*** 0.240*** 1 SIZE -0.204*** 0.421*** -0.567*** 0.147*** 0.202*** -0.057** 1 NONINT -0.017*** 0.320*** -0.016*** 0.094 -0.012*** -0.034*** -0.095*** 1

(26)

25

Figure 4 compares the performance of FOBs and DOBs, from the fluctuation of the first graph we can see that, the EBIT of FOBs is higher than that of DOBs a with a large variation in the beginning, while the gap between the two subgroups narrow down and almost disappear during the following three years, finally, the EBIT of DOBs experience a increasing trend during the last year and surpass its peer group. The wave of EFFICNCY is relatively stable for both of the two subgroups, around 0.5(FOBs) and 0.4 (DOBs) respectively.

Figure 3: The fluctuation of performance measures by years (GOBs vs POBs)

(27)

26

6. Results

6.1 Tests for difference in mean and median

By comparing the performance between government-owned banks Vs. private-owned banks and foreign-owned banks Vs. domestic-owned banks, I first perform a t-test and a Mann-Whitney U

Figure 4: The fluctuation of performance measures by years (FOBs Vs DOBs)

(28)

27

test respectively to identify the joint equality of each variables across those two pairs of ownership groups.t-test is a parametric test of the joint equality of means for each variable while Mann-Whitney U test is a non-parametric test which require less stringent assumptions, it used to examine the equality of median of each variables. Before apply t-test, a Leven’s Test for homogeneity of variances was conducted, only EFFICIENCY show significant sign of unequal variances, thereby, t-statistic which does not assume equal variance is reported for EFFICIENCY, while the results assumes equal variance are reported for all other variables. Table 4 illustrates the results for conducting t-test and Mann-Whitney U test.

Table 4 t- test and Mann-Whitney U test between Government-owned & private-owned banks; Foreign -owned & Domestic-owned banks.

Mean and Median values of bank performance measures are reported, as well as P-value for each of the two test.

t-test Mann-Whitney test

Variables GOBs POBs FOBs DOBs GOBs POBs FOBs DOBs

Efficiency 0.33 0.47 0.51 0.36 0.17 0.44 0.48 0.25 (0.000)*** (0.000)*** (0.000)*** (0.000)*** EBIT (%) 1.30 2.06 2.01 1.73 1.10 1.64 1.67 1.17 (0.000)*** (0.000)*** (0.000)*** (0.000)*** CAPITAL (%) 10.51 17.97 18.55 13.55 5.88 13.83 14.01 8.55 (0.000)*** (0.000) *** (0.000)*** (0.000)*** DEPOSITS 0.89 0.77 0.77 0.83 0.95 0.78 0.77 0.90 (0.001)*** (0.009)*** (0.001)*** (0.000) *** LOANS 0.56 0.56 0.59 0.55 0.54 0.61 0.63 0.56 (0.000) *** (0.001)*** (0.010)** (0.001)*** SIZE 15.39 13.34 13.32 14.41 15.50 13.27 14.65 13.29 (0.025)** (0.000)*** (0.000)*** (0.000)*** NONINT 0.23 0.25 0.27 0.22 0.016 0.014 0.014 0.016 (0.005)*** (0.208) (0.059)* (0.716) NO. of Obs 258 797 557 509 258 797 557 509

GOBs: government-owned banks; POBs: private-owned banks; FOBs: foreign-owned banks; DOBs: domestic-owned banks. Bank performance indicators are EFFICIENCY and EBIT. Other variables contain: SIZE: the log of total assets; CAPITAL: the ratio of book value of equity to total assets; LOANS: the ratio of loans to total assets; DEPOSITS: the ratio of deposits to total assets; NONINT: the ratio of non-interest income to total assets;

(29)

28

First looking at the comparison between GOBs vs. POBs, significant difference can be found in t-test for all of the variables on different significant levels. The Mann-Whitney test presents the similar results. The two performance indicators of government-owned banks exhibit substantially lower average and median value than that of privately owned banks. For all the other variables, state-owned banks exhibit relatively larger size and deposits but tend to be less capitalized and exhibit smaller amount of loans and non-interest incomes than their privately owned peers.

On the other hand, higher efficiency and EBIT can be found for foreign-owned group than their domestic counterparts in terms of both mean and median values. Apart from this, FOBs are smaller but enjoy larger amount of loans and well capitalized than DOBs, deposits are remarkably lower for FOBs than for DOBs. Finally, there is no significant difference regarding non-interest incomes between the two ownership groups.

To sum up, based on the results of t-test as well as Mann-Whitney U test, there is enough evidence to conclude that significant difference can be detected among nearly all the variables between two pairs of ownership groups. Moreover, the mean and median values of the performance indicators suggest that government-owned perform way below their private-owned peers on both efficient and earnings from assets measures, whereas foreign-owned banks are unambiguously better performing than domestic ones by either measure.

6.2 Regression analysis

In order to get further exploration, OLS pooled regression analyses have been conducted between pairs of ownership groups: GOBs vs. POBs and FOBs vs. DOMs, The two dummy variables were separate and joint incorporated into the regression regarding the two dependent variables respectively in order to check the sensitive of the results. P-value is reported in the parentheses with Table 5. It should be highlighted that the year dummy and country dummy are all taken into the regression to control the time fixed and country fixed effects while not reported in tables.

(30)

29

lower efficiency as compared with their privately counterparts, the result is consistent with the researches of (Bonin et al., 2004; Nakane & Weintraub, 2005; Beck et al., 2003). Likewise, the results go against with researches of (Staub et al., 2010; Bhattacharya et al., 1997). Also as I expected, test reported in Column 2 and 3 shows that the sign of FOB dummy variable is positive and statistical significance with EFFICIENCY at 1% level, indicating foreign-owned banks exhibit higher efficiency than that of their domestic peers (Bonin et al., 2004; Styrin, 2005; Karas et al., 2010b).

The result is of economically importance as well (Micco et al., 2007). Focusing on the joint regression (column 3), on average, the efficiency of government-owned banks (measured by EFFICIENCY) is 43.9% points lower than that of their private peers, whereas the efficiency of foreign-owned institutions is 42.6% higher than that of the domestic ones.

(31)

30

more investors in this sector of those areas, hence server the competitiveness of the banking industry, the result is in contrast with my hypotheses, while in line with the research of Bonin et al. (2004). At last, LOANS show positive correlation with EBIT but not reach at a significant level, indicating LOANS have no effect in explaining EFFICIENCY.

Table 5

Regression for bank performance and ownership structure variable: Government owned versus Privately owned banks; Foreign owned versus Domestic owned banks;

POLS model is employed and regression coefficient and p-value are reported.

EFFICIENCY EBIT Variables 1 2 3 4 5 6 Constant -3.166*** -2.968*** -3.187*** 4.997*** 5.148*** 4.996*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) GOB -0.696*** -0.439*** -0.251*** -0.303*** (0.000) (0.000) (0.000) (0.000) FOB 0.565*** 0.426*** -0.007 -0.098 (0.000) (0.000) (0.605) (0.124) CAPITAL 0.014*** 0.011*** 0.013*** 0.037*** 0.036*** 0.037*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) DEPOSITS -0.019* -0.024** -0.005 -0.223*** -0.214*** -0.224*** (0.079) (0.024) (0.612) (0.000) (0.000) (0.000) LOANS 0.025 0.016 0.001 0.571*** 0.587*** 0.576*** (0.154) (0.360) (0.932) (0.000) (0.000) (0.000) SIZE 0.011*** 0.007*** 0.014*** 0.093*** 0.075*** 0.092*** (0.000) (0.002) (0.000) (0.000) (0.000) (0.000) NONINT 0.572*** 0.466*** 0.511*** 8.363*** 8.321*** 8.391*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) GDP -0.019*** -0.016*** -0.014*** 0.003 0.002 0.002 (0.000) (0.000) (0.000) (0.593) (0.678) (0.957) R-squared 0.12 0.13 0.13 0.11 0.11 0.11 N 895 895 895 1042 1042 1042

Result for Efficiency and EBIT. Independent variable are: a dummy variable GOB equal to 1 if the bank is government-owned, otherwise zero and a dummy variable FOB equal to 1 if the bank is foreign-government-owned, otherwise zero.

Control variables include SIZE: the log of total assets; CAPITAL: the ratio of book value of equity to total assets; LOANS: the ratio of loans to total assets; DEPOSITS: the ratio of deposits to total assets; NONINT: the ratio of non-interest income to total assets; and GDP: the national GDP growth rate.

(32)

31

In terms of the regression that taking EBIT as performance measure (Column 4, 5 and 6), the dummy variable GOBs exhibits significant and inverse correlation with EBIT in both of the two cases (Column 4, 6) ( La porta et al., 2002; Micco et al., 2007; Cornett et al., 2010), this is also in line with EFFICIENCY regression. Again, the finding is both statistically and quantitatively significant (Micco et al., 2007). The government-owned banks are 0.30% (Column 6) point lower in earnings before interest and taxes than that of their privately owned counterparts. What come into notice is that the results suggest no significant difference was detected between foreign-owned banks and domestic-owned banks in terms of EBIT. Moreover, although the correlation values in both the two regression are extremely small, the negative sign is contradict with the EFFICIENCY regression.

Most of the controlling variables show same results with the EFFICIENCY regression, except for LOANS and GDP. The former variable illustrate strong significant and positive sign of correlation with EBIT, since loan, as the main source of bank assets, is more valuable compared with others (Iannotta et al., 2007). Finally, GDP show no correlation with EBIT in all three cases.

To sum up, the use of different performance indicators does not lead to mix results in terms of the dummy variable GOBs. The result strongly support my first hypothesis that government-owned banks that associate with their inefficiency and lower earnings go far away from the development view rendered superior performance. In contrast, agency and political view can better explain why they did not work as planned. My finding is parallel with the research of Iannotta et al., 2007, who find the poorer performance of state-owned banks stem both from their lower earnings from assets but also inferior efficiency. Cornett et al. (2010) point out the performance gap between government-owned banks and private-owned banks can be explained by the perverse incentives of managers/political bureaucrats of government-owned banks.

(33)

32

cost in providing same quality of services than domestic peers or for sending home country managers abroad (Berger, 2000). The research that conducted by Bonin et al. (2004) strengthen my argument. The authors use empirical test to prove foreign-owned banks show higher efficiency but lower cost efficiency than their domestic counterparts.

6.3 Robustness test

In this section I conduct some sensitive checks in order to test whether the results of my study are robust to alternative specification.

Firstly, for further exploiting the panel data, the Generalized Least Squares (GLS) model was employed to replace the original OLS model. One study conducted by Isir & Hassan (2003) employ GLS model instead of OLS ones for regression analyses, the authors further explain that the dependent variable, that is the efficiency score, estimated by nonparametric approach may cause heteroscedasticity (This is similar with my study, but efficiency score is calculated by different technique), whereas GLS regression can deal with this problem by computing a consistent estimate of the covariance matrix. In addition, as I mentioned earlier in the paper that the efficiency score measured by DEA approach may lead to biased result, the dependent variable EFFICIENCY will be introduced into the regression after rewritten in logistic functional form.

Secondly, as suggest by previous scholars (Dietrich & Wanzenried, 2011; Athanasoglou, et al., 2005), bank profits exhibit a tendency of persistence over time. Due to this reason, I introduce a one year lagged performance indicator as an additional explanatory variable in each of the two regressions. The results of GLS regression are list in Table 6.

(34)

33

up, the sensitive tests illustrate similar results with the initial ones, implying the results of my study are robust.

Table 6

Regression for bank performance and ownership structure variable: Government owned versus privately owned banks; Foreign owned versus Domestic owned banks; PGLS model is employed and regression coefficient and p-value are reported.

EFFICIENCY EBIT Variables 1 2 Constant -1.493*** 1.919*** (0.000) (0.000) GOB -0.214*** -0.162*** (0.000) (0.000) FOB 0.306*** 0.176*** (0.000) (0.000) CAPITAL 0.003*** 0.022*** (0.000) (0.000) DEPOSITS -0.014*** -0.236*** (0.000) (0.000) LOANS 0.240*** 0.239*** (0.000) (0.000) SIZE 0.020*** 0.125*** (0.000) (0.000) NONINT 0.017*** 5.076*** (0.006) (0.000) GDP EFFICIENCY (-1) EBIT (-1) -0.032*** (0.000) 0.540*** (0.000) 0.032*** (0.000) 0.536*** (0.000) R-squared 0.99 0.99 N 895 1042

Result for Efficiency and EBIT. Independent variable are: a dummy variable GOB equal to 1 if the bank is government-owned, otherwise zero and a dummy variable FOB equal to 1 if the bank is foreign-government-owned, otherwise zero.

Control variables include SIZE: the log of total assets; CAPITAL: the ratio of book value of equity to total assets; LOANS: the ratio of loans to total assets; DEPOSITS: the ratio of deposits to total assets; NONINT: the ratio of non-interest income to total assets; and GDP: the national GDP growth rate.

(35)

34

7. Conclusion

The banking system in BRICs experienced their dramatic transformations during the late 1990s. An increasing presence of mergers and acquisitions took place which in turns cause a trend of bank privatization, at the same time attracting not only domestic investors but also foreign players joint in this industry. Hereby, the main idea of this paper is to investigate how bank performances vary in response to those changes of bank ownership structure among BRIC countries.

The results shed light two issues: First, consistent with my hypothesis, government-owned banks exhibit lower efficient and lower profitability than that of private-owned banks. This finding is go opposite with development view justified the better achievements of state-owned banks in bank operating and balancing social and economic objectives. Instead, the analyses more favor the agency and political theories. Specifically, the weaker motivational mechanism for managers in pursuing bank financial outcomes and imperfect monitoring system of government-owned banks are more likely to cause agency problems, hence, the real implementation of the banks cannot be guaranteed. The less intense supervision mechanism as well as the less development regulatory system in turns made government-owned banks a remarkably effective tool for politicians to achieve their self-interests (Megginson, 2005).

Second, even though no significant difference was detected regarding earnings before interest and taxes between foreign-owned banks and domestic-owned banks, the former exert higher efficiency on the extremely significant level. The result to some extent favor more of global advantage hypothesis than the home field advantage hypothesis, that is, the superior managerial skill and leading techniques of foreign-owned banks offset their cross-border barriers and exhibit higher efficiency than their domestic competitors.

(36)

35

lack of clear ownership information. Another limitation is that not all of the commercial banks in BRICs are included in the sample, the reason is that explore the nature ownership for every bank was difficult based on the exits data set, hereby, I delete some of banks missing clear ownership information. This may to some extent affect my empirical results.

(37)

36

Reference

Alchian, A., 1965. Some economics of property tights. I1 Politico, 30, 816-829

Ataullah, A., Le, H., 2006. Economic reform and bank efficiency in developing countries: the case of the Indian banking industry, Applied Financial Economics, Vol. 16, 653-663.

Athanasoglon, P., Brissimis, S.N., Delis, M.D., 2005. Bank-specific, industry-specific and macroeconomic determinants of bank profitability, Bank of Greece, Working Paper, No.25. Barry, A.T., Lepetit, L., Tarazi, A., 2010. Ownership structure and risk in publicly held and privately owned banks, Journal of Banking & Finance, Vol. 35 (2010), 1327-1340.

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

Beck, T., Crivelli, M.J., Summerhill, W., 2003. State Bank Transformation in Brazil - Choices and Consequences, World Bank Policy Research Working Paper 3619,2005.

Berger, A.N., 1995. The relationship between capital and earning in banking, Journal of Money, Credit and Banking, Vol. 27, 432-456.

Berger, A., De Young, R., Genay, H., Udell, G., 2000. Globalization of financial institutions: evidence from cross border banking performance, Brookings Papers on Economic Activity Vol. 2, 23-158.

Berger, A.N., Hasan, I., Klapper, L.F., 2004. Further evidence on the link between finance and growth: An international analysis of community banking and economic performance, Journal of Financial Services Research Vol. 25, 169-202.

Berger, A., Clarke, G., Cull, R., Klapper, L., Udell, G., 2005. Corporate governance and bank performance: A joint analysis of the static. selection and dynamic effects of domestic, foreign, and state ownership, Policy Research Working Paper 3632. The World Bank.

Berger, A.N., Hasan, I., Zhou, M., 2009. Bank ownership and efficiency in China: what will happen in the world’s largest nation? Journal of Banking & Finance Vol. 33 (2009) 113-130. Bhattacharyya, A., Lovell, C.A.K., Sahay, P., 1997. The impact of liberalization on the productive efficiency of Indian commercial banks, European Journal of Operational Research, Vol. 98 (1997) 332-345.

(38)

37

Boubakri, N., Cosset, J.C., Fischer, K., Guednami, O., 2005. Privatization and bank performance in developing countries, Journal of Banking & Finance Vol. 29, 2015-2041.

Casu, B., Molyneux, P., 2003. A comparative study if efficiency in European banking, Applied Economics, Taylor and Francis Journal, Vol.35 (17), pages 1865-1876.

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

Clarke, Cull, M., Shirley, 2005. Bank privatization in developing countries: A summary of lessons and findings, Journal of Banking & Finance, Vol. 29 (2005) 1905-1930.

Coelli, T.J., 1996. A Guide to DEAP Version 2.1: A Data Envelopment Analysis (Computer) Program, Centre for Efficiency and Productivity Analysis (CEAP) Working Papers. NO. 8/96. Cohen, J., 1988. Statistical power analysis for the behavioral science (2nd ed.), Hillsdale, NJ: Lawrence Erlbaum Associates.

Cornett, M.M., Guo, L., Khaksari, S., Tehranian, H., 2010. The impact of state ownership on performance differences in privately-owned versus state-owned banks: An international comparison, J. Finan. Intermediation, Vol.19(2010)74-94.

Demirguc-Kunt, A., Huizinga, H., 1999. Determinants of Commercial Bank Interest Margins and Profitability: Some International Evidence, The World Bank Economic Review, Vol. 13, 379-408. Dietrich, A., Wanzentied, G., 2011. Determinants of bank profitability before and during the crisis: Evidence from Switzerland, Journal of International Financial Market, Institutions & Money, Vol. 21, 307-327.

Dobson, W., Kashyap, A.K., 2006. The contradiction in China’s gradualist banking reforms, Brookings Panel on Economic Activity, Vol. (2), 1-60.

Fethi, M.D., Pasiouras, F., 2009. Assessing bank performance with operational research and artificial intelligence techniques: A survey, University of Bath School of Management Working Paper Series. http://www.bath.ac.uk/management/research/papers.htm

Gerschenkron, Alexander, 1962, Economic Backwardness in Historical Perspective, Harvard University Press, Cambridge, MA.

(39)

38

Gupta, P., Kochhar, K., Panth, S., 2011. Bank Ownership and the Effects of Financial Liberalization: Evidence from India, IMF Working Paper, International Monetary Fund WP/11/50.

Hassan, M.H., Bashir, A.H., 2003. Determinants of Islamic Banking Profitability, ERF Paper. Hawkins, J., Mihaljek, D., 2001. The banging industry in the emerging market economies: competition, consolidation and systemic stability-an overview, BIS Papers NO.4.

Iannotta, G.,Giacomo, N., Sironi, A.,2007. Ownership Structure, Risk and Performance in the European Banking Industry, Journal of Banking and Finance, Vol. 31, 2127-2149.

Isik, I., Hassan, M.K., 2003. Efficiency, ownership and market structure, corporate control and governance in the Turkish Banking Industry, Journal of Business Finance & Accounting, Vol. 30 (9) & (10), Nov. Dec. 2003, 0306-686X.

Jensen, M, C., Meckling, W., 1976. Theory of the Firm: Managerial Behavior, Agency Costs, and Ownership Structure, Journal of Financial Economics, Vol. 3, 305-360.

Jin O’Neill, 2001. Building Better Global Economic BRICs, Global Economics Paper, No.66. Karas, A., Schoors, K., Weill, L., 2010b. Are private banks more efficient that public banks? Evidence from Russia‖, Economics of Transition, Vol. 18, 209-244.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A.A., 2002. Government ownership of banks. J. Finance, Vol. 57, 265-302.

Levine, R., Sara Z., 1998. Stock Markets, Banks, and Economic Growth, American Economic Review, Vol. 88, pp.537-558.

Lin, X., Zhang, Y., 2009. Bank ownership reform and bank performance in China, Journal of Banking and Financ, Vol. 33, 20-29.

Maudos, J., Nagore, A., 2005. Explaining market power differences in banking: a cross country study, IVIE Working Paper 10.

Megginson, W. L., 2005. The economics of bank privatization, Journal of Banking & Finance, Vol. 29 (2005) 1931-1980.

Meulbroek, L.K., 2000. Executive Compensation Using Relative- Performance–Based Options: Evaluating the structure and costs of indexed options, Harvard Business School Working Paper No. 02-021.

(40)

39

Nakane, I.M., Weintraub, B.D., 2005. Bank privatization and productivity: Evidence for Brazil, Journal of Banging & Finance, Vol. 29 (2005) 2259-2289.

Patti, E.B., Hardy, D.C., 2005. Financial sector liberalization, bank privatization, and efficiency: Evidence from Pakistan, Journal of Banking & Finance, Vol. 29 (2005) 2381-2406.

Rivas, A., Ozuna, T., Policastro, F., 2006. Does the use of derivatives increase bank efficiency, International Business & Economic Research Journal, Vol. 5, 47-56.

Sanyal, P., & Shankar, R., 2011. Ownership, competition, and bank productivity: An analysis of Indian banking in the post-reform period, International Review of Economics and Finance, Vol. 20 (2011) 225-247.

Sathye, M., 2002. Efficiency of banks in a developing economy: The case of India, European Journal of Operational Research, Vol. 148 (2003) 662-671.

Sathye, M., 2005. Privatization, performance and efficiency: A Study of Indian Banks, VIKALPA Vol. 30 NO.1.2005.

Sapienza, P., 2004. The effect of government ownership on bank lending. J. Finan. & Econ., Vol. 72, 357-384.

Sherman, H.D., Gold, F., 1984. Bank Branch Operating efficiency: Evaluation with Data Envelopment Analysis, Journal of Banking & Finance, Vol. 9, 297-315.

Shlerfer, A., Vishny, R.W., 1997. A survey of corporate government, J. Finance, Vol. 52, 737-783.

Shleifer, A., 1998. State versus private ownership, J. Econ, Vol. 12, 133-150.

Stark, D., 1997. Recombinant Property in East European Capitalism in: Grabher, G. and D. Stark (eds.), Restructuring Networks in Post-Socialism, Oxford: Oxford University Press, 35-69. Staub, B.R., Souza, S.G., Tabak, M.B., 2010. Evolution of bank efficiency in Brazil: A DEA approach, European Journal of Operational Research, Vol.202 (2010) 204-213.

Styrin, K., 2005. What explain difference in efficiency across Russian Banks? Economics Education and Research Consortium, Russia and CIS, Final report, Moscow.

(41)

40

Appendix

Appendix A: Summary of empirical studies regarding the influence of ownership groups on

bank performance in each of the four countries in BRICs respectively.

Study Country and sample description Summary of empirical findings and

conclusion Nakane & Weintraub

(2005)

Brazil (1990-2001)

Using intermediate input as a proxy of productivity

Stated-owned banks are less productive than their privately peers; no significant performance difference can be detected between banks acquired by foreigners and those acquired by domestic investors.

Beck, Crivelli, Summerhill (2003)

Brazil (1995-2003)

Use ROA, ROE and Overhead cost/assets to measure performance

Private banks improve banks performance and efficiency; the authors apply political economy theories to explain the transformation of Brazilian banks.

Staub, Souze and Tabak (2010)

Brazil (2000-2007)

Employ data envelope analysis to compute efficient scores

Banks with foreign participation are the least cost efficient compared to that of domestic peers, this finding in favor of the home field advantage hypothesis; the agency theory hypothesis is reject as state-owned banks are more efficient than private banks

Styrin (2005) Russia (1998-2002)

Adopt stochastic frontier approach to compute X-inefficiency

Foreign banks tend to be more efficient than their domestic counterparts, supporting the global field advantage hypothesis.

Karas, Schoors and Weill (2010 b)

Russia (2002 and 2006)

Adopt stochastic frontier approach to compute cost efficiency

(42)

41

Bhattacharyya, Lovell and Sahay (1997)

India (1986-1991)

Employ data envelope analysis to compute efficient scores

State-owned banks perform with most efficient, followed by foreign-owned banks and private-owned banks at last.

Sathye (2002) India (1997-1998)

Using the non-parametric technique -- DEA

Foreign banks appear to be more efficient than banks with other two type of ownership, and the private Indian banks are the least efficient.

Berger, Hasan and Zhou (2009)

China (1994-2003)

Investigate the profit and cost efficiency of banks operation in China

Foreign banks are the most profit efficient, followed by private domestic banks and the state-owned institutions are least efficient.

Lin & Zhang (2009) China (1997-2004)

Joint analysis of the static, selection and dynamic effects of domestic private, foreign and state ownership.

How ownership affect bank performance — Evidence from BRIC countries