The choice between private banks and state-owned banks

(1)

University of Amsterdam -‐ Amsterdam Business

Bachelor in Economics and Business, Economics and Finance track

The Choice Between Private Banks and State-Owned Banks

Bachelor Thesis

July 2015

Max Kranendijk (101251200)

Thesis Supervisor: C. van der Kwaak

(2)

Abstract

The nationalization processes in recent years as a consequence of the financial crisis have reignited the debate on which ownership structure should be preferred, private or state-‐owned banks. In this literature review, the differences between state-‐owned banks and private banks are examined based on empirical and theoretical studies. In general, most studies find that private banks should be preferred over state-‐owned banks. Firstly, countries with a relatively higher degree of private banks than state-‐owned banks are associated with more economic growth. Secondly, in most countries private banks perform better than state-‐owned banks in terms of profitability, riskiness and efficiency. Thirdly, private banks are not influenced as much as state-‐owned banks by political incentives. However, an advantage of state-‐owned banks over private banks is their stabilizing effect during a crisis. During a crisis state-‐owned banks often have a positive impact on the economy, because state-‐owned banks then tend to keep lending out money to firms, so that projects are continued. In addition, an example of the German banking system shows that in a country with “good governance” state-‐owned banks could possibly perform better in terms of profitability than private banks.

(3)

country would be so severe, that governments had no choice, but to nationalize these banks. This happened especially in developed countries (examples are Northern Rock in the UK, Kaupting and Landsbanki in Iceland, Anglo Irish bank in Ireland, and ABN AMRO in the Netherlands), where in general the amount of private banks relative to state-‐ owned banks is higher. Criticizers of the banking system in developed countries, with its high degree of private banks, claim that in this situation profits of banks are privatized, while the losses are nationalised. Proponents of private banks point out that countries with high economic growth, are in general countries with a high degree of private banks (La Porta 2002, Barth, Caprio and Levine 1999, Cole 2009). The nationalization

processes in recent years have reignited the debate on which ownership structure should be preferred, private or state-‐owned banks (Bertay, 2014., Cull 2013). This resulted in the following research question; “should state-owned banks be preferred over private banks?”

In this paper I try to give a well-‐considered answer to this general question, by executing a diverse literature review. To give some background information, I will give a short overview in chapter 1 on how the financial system works. This will be done by giving a simplified description on how banks work, how banks reduce information asymmetry and what causes banks to be in danger. In chapter 2, theoretical arguments in favour of private banks are discussed, followed by theoretical arguments in favour of state-‐owned banks. In chapter 3, the two ownership forms of banking are judged in three different ways. First, bank performance will be considered based on economic growth, profits and efficiency. Second, the difference in lending behaviour between state-‐owned and private banks will be examined. Third, the effects of bank ownership on the economy during a crisis will be evaluated.

When examining the effects of bank ownership on economic growth we can assume that the prevalence of relatively many private banks in a country seems to go along with more economic growth and a more efficient financial system in the long run

(5)

(La Porta, 2002., Barth, Caprio and Levine 1999., Cole 2009). When looking at the individual performances of banks based on profit, riskiness and efficiency, the results are more mixed. Most studies that evaluate the performance differences between state-‐ owned banks and private banks across countries (both empirical and theoretical) find that state-‐owned banks are less profitable and take more risk in particular in times of economic downturn (Ianotta, 2007., Cornett 2010, Boyd 2005). The results of Figueira (2009) are an exception, which show that in Latin America private banks did not outperform state-‐owned banks. However, when we look at the differences in

performance of state-‐owned banks and private banks on a per country basis, we find mixed results. In India (Bhattacharyya, 1997) state-‐owned banks were more efficient but less profitable, while in Germany (Altunbas, 2001) state-‐owned banks had slight cost advantages and were more profitable.

Looking at the differences in lending behaviour between private and state-‐owned banks, the findings of Dinc (2005) and Sapienza (2001) give a clear preference for private banks. They show that in most countries state-‐owned banks are politically

biased, by charging lower interest rates and increase lending behaviour during elections. However, in most developed countries an increase in lending behaviour cannot be

detected during elections.

However, in times of a crisis state-‐owned banks seem to have a positive impact on the economy of a country. Brei (2014), Betray (2014) and Coleman (2014) et al showed that contrary to private banks, state-‐owned banks keep lending out money to firms during a crisis, so that firms can continue their projects. These articles proved that in times of a crisis, countries or local areas with a relatively high degree of state-‐owned banks generated higher economic growth than countries or local areas with a relatively low degree of state-‐owned banks. The drawback of this outperformance during crisis periods is that in times when there are no recessions countries with relatively many private banks seem to generate more economic growth (Brei, 2014., Betray, 2014., Coleman, 2014.))

II. Chapter 1

How does the Financial System work?

(6)

According to Megginson (2001) banks have three main functions. Banks are responsible for the transfer of money, act as financial intermediaries and banks are able to evaluate and monitor credit allocation. Banks generate money by accepting money as deposits, or by borrowing from money markets. Deposits come from individuals, businesses,

financial institutions and governments with surpluses. Banks use their deposits and borrowed money (liabilities) to make loans or buy securities (assets). Loans are made to individuals, businesses, financial institutions and governments with deficits.

(www.frbsf.org/education/publications/doctor-‐econ/2001/july/bank-‐economic-‐

function., 05-‐06-‐2015)

Banks are originally created so that people and businesses can put their money at a safe place, in which it also gets interest for. At the same time people and businesses, which are in need of money for their investments can borrow at banks. Generally banks make money by lending money at higher interest rates, than the interest they pay on deposits. So banks make money by collecting interest on loans and interest payments from the securities they own, while interest rates on deposits cost banks money. The difference between the interest banks receive on loans and the interest banks pay on deposits is known as the “spread”, which is the net interest income for banks.

(www.investopedia.com/university/banking-‐system/banking-‐system3.asp., 05-‐06-‐ 2015)

1.2- How do banks reduce information asymmetry?

Banks act as intermediaries in the financial system and reduce the asymmetric problem in contracts between borrowers and lenders. Information asymmetry implies a

situation, in which one party in a transaction has more or superior information than the other party. The asymmetric information problem arises in the situation of money lending/borrowing, because borrowers know more about the risk of their investments than lenders do. Also lenders cannot observe if borrowers will act in the best interest of the lenders. This moral hazard is known as the principal-‐agent problem (principal is the lender and agent is the borrower). (Berk and De Marzo, 2011). When borrowers are monitored, lenders can more easily know which borrowers will have a higher chance of default. Banks reduce the information problem, as they provide both credit and liquidity

(7)

and so are able to collect far more information than other lenders (Berk and De Marzo, 2011).

1.3- What causes banks to be in danger?

According to Cippolini and Fiordelisi (2012) credit risk, liquidity risk and bank market power are the main factors of banks getting into danger. Credit risk refers to the risk of default by the issuer of any bond or loan. It is a measure to indicate to which degree the bond’s or loan’s cash flows are not known with certainty. Liquidity risk refers to the risk of not being able to buy or sell an investment quickly enough to prevent a loss. In case of illiquidity, the party is forced to liquidate an investment at a loss, because cash is

required to pay for another obligation. Bank market power refers to the ability of a bank to manipulate prices (in this case interest rates) by influencing supply and or demand of money (Berk and DeMarzo, 2011).

Banks are in danger when they are in financial distress. Cippolini and Fiordelisi (2012) measure this by looking at the effect of credit risk, liquidity risk and market power on Shareholder Value Ratio. The shareholder value ratio (SHVR) is the ratio between economic value added and the shareholder capital invested in the previous period. Economic value added, or EVA, is calculated by computing the difference between the bank net operating profits minus the cost of capital (expressed as a percentage) times the capital invested in the previous period.

Cippolini and Fiordelisi (2012) measure credit risk by focusing on the ratio of loan loss provisions to total loans. They find in all their models an increase in the probability of a distressed shareholder value ratio, when the ratio of loan-‐loss

provisions to total loans increase. Liquidity risk is measured by focusing on the ratio of liquid assets to total assets. When accounting for country fixed effects there is a negative relationship between liquid assets and observing a distressed shareholder value ratio.

Bank market power is estimated by using the Lerner index, which indicates the degree to which market power allows to fix a price above marginal cost. Chippolini and Fiordelisi (2012) find that a distressed shareholder value ratio has a negative significant link with the Lerner index, suggesting that an increase in market power lowers chances of financial distress. So an increase in credit risk and liquidity risk lead to a higher

(8)

chance of banks getting in financial distress. Berger et al (2009) state that there is a negative link between bank market power and financial distress.

III. Chapter 2:

Bank Ownership and the Reasons for Nationalizing or Privatizing.

2.1- What are the reasons for privatizing banks?

Privately owned banks are commercial and generally aim at generating maximum profits. Government-‐owned banks are in the hands of the public and would ideally aim at generating a maximum social welfare. Private ownership should in general be preferred over public ownership, when the incentives to innovate and to contain costs are strong (Shleiffer, 1998).

As biggest evidence Shleifer (1998) points out the difference in success of capitalism over socialism. During the first decades after the Second World War

economists almost all favoured government ownership of firms as soon as any market imperfections were suspected. This was especially the case for “strategic” sectors, as mines, hospitals, schools and banks. However, the last 30 years the desire to privatise grew enormously in market economies and communism collapsed almost everywhere in the world. Governments proved to be imperfect, by preferring political goals, such as patronage or simply maximizing the income of politicians (Shleiffer, 1998). More interestingly Shleifer and Vishny, (1994, 1998) conclude that state firms are inefficient, due to the fact that governments deliberately transfer resources to supporters.

Cornet (2010) states that state-‐owned companies disregard social objectives and are extremely inefficient. The reason for this is that politicians often have goals that are in conflict with social welfare improvements, while having as main purpose political interests. This is also the case for government-‐owned banks. Shleifer and Vishny (1994) explain this by the fact that government-‐owned banks are run by political bureaucrats, who have a total power on the control rights, but not on the cash flow rights, as these need to go to tax-‐payers. The profits are distributed in the form of dividends to the state and do not accrue to the bureaucrats. Therefore maximizing profits are not extremely important for the bureaucrats, who own the state-‐owned banks. A bigger concern for them is to fund special interest groups, such as trade unions, who help the bureaucrats

(9)

win elections. This political interest of state-‐owned banks cause that their goals are in conflict with social welfare and maximization of profits. In this case the performance of state-‐owned banks would in general be inferior to privately owned banks (Shleifer and Vishny, 1994).

La Porta (2002) defines this as the “political view” of government participation in financial firms. Politicians who control banks would use their power over banks to win votes, instead of focusing on efficiency and maximizing profit. State-‐owned banks would give jobs or financial resources to friends and political supporters. So politicians would run state-‐owned banks not to finance resources to economically efficient users, but to maximize their own personal objectives (Sapienza 2004). According to Kane (2000) politicians try to preserve cheap subsidized loans for their parties and befriended sectors. These subsidized loans occur much more often at state-‐owned banks. This creates un-‐booked losses, which become visible only in the case of a banking crisis. As privatized banks main goal is to maximize profits, there is no incentive to give away these cheap loans.

Dinc (2010) gives reasons why the problems of political influence will be greater at government-‐owned banks than other government-‐owned firms. “First, the

asymmetric information between lending banks and outsiders about the quality of a specific loan makes it easy to disguise political motivation behind a loan. Second, revealing the costs of any politically motivated loan can be deferred until the loan maturity. Third, while a non-‐bank government-‐owned enterprise operates in a defined industry, which can limit the politicians’’ ability to transfer resources, banks operate across the whole economy, providing politicians with more opportunity to channel funds”. Funds can be channelled by providing loans more easily to large companies who are aligned with trade unions, so that these unions will support the political party of the bureaucrats (Dinc,2010).

Privatization is also supported by the ‘competition-‐stability’ view, which states that a high degree of competition creates a more stable financial system (Boyd and De Nicolo, 2005). In an economy with a high degree of state-‐owned banks, there is often a lower degree of competition. Too much market power of only a few banks will cause these banks to raise interest on loans, which will adversely select investors and firms with risky investments. This will have a negative impact on the stability of the banking system (Cipollini, 2012. Boyd and De Nicolo (2005) call this a moral hazard problem, as

(10)

borrowers confronted with higher interest costs, optimally choose a higher risk

portfolio. More competition would increase chances of solving the optimal contracting problem, which states that the actions of borrowers are unobservable, or observable at a cost. As privatization goes along with more competition, risks of financial distress will decrease.

Moreover, Figueira (2009) states that under the principal-‐agent theory managers in the private sector have greater incentives to aim at maximizing profit than managers under state-‐ownership. He states that this is due to the fact that the private capital market is far better in monitoring management behaviour than government departments.

O’Hara states that in the case of state-‐owned banks there is a lack of capital market discipline, which indicates that management in these banks experience a lower intensity of political pressure and may therefore operate less efficiently than privately owned banks. For example, when a bank is in a less competitive local market the bank may choose to take the gains of market power to spend more on staff or purposes, instead of aiming at higher profits. State-‐owned banks are typically more concentrated and face less competition (O’Hara, 1981). So did Berger and Hannan (1998) find that during the 1980s U.S. banks which operated in more concentrated markets faced lower cost efficiency.

2.2- What are the reasons for nationalizing banks?

Privately owned firms might underperform in terms of quality or neglect the interests of certain stakeholders, because their single-‐minded focus on profits. A politician, who should stand for social efficiency, could then improve efficiency, by controlling decisions of firms (Sappington & Stiglitz (1987). Sapienza (2005) expands this to the social view, based on the economic theory of institutions. State-‐owned firms are created to

overcome market failures, whenever the social benefits of state-‐owned firms exceed the costs. The social view implies that state-‐owned banks improve economic development and improve general welfare (Cole 2009). Moreover, La Porta (2002) argues that state-‐ owned banks ensure complete control of the government for the choice which projects should be financed, while regulation of private banks could only influence the financing of projects partly.

(11)

According to Dinc (2005) it is a common view that government ownership of banks facilitates financing of certain investments that private banks would not finance. This is especially important for projects that could help economic development. La Porta (2002) even introduces the development view, which states that especially in emerging markets due to bank ownership the government can more easily finance investments which create financial and economic development.

There is also a “light” version of the social view, which is called the agency view. The agency view also shares the idea that state-‐owned banks in essence try to create maximum social welfare. Conversely, bank ownership of the state increases the chances of corruption and misallocation. Sapienza (2004) compares the choice between state-‐ owned and privately owned firms as “ a trade-‐off between internal and allocative

efficiency”. State-‐owned banks would allocate resources more efficiently, as state-‐owned banks move resources to socially profitable activities. Conversely, state-‐owned banks would internally be less efficient than private banks, as public managers would exert less effort or use resources for personal benefits, which could lead to corruption (Sapienza, 2004).

In the 1930s the main view was that to preserve stability, competition needed to be restrained. The basic idea was that when banks earn monopoly rents, banks become conservative. The reason for this is that they see their monopoly position as a valuable asset, so they want to lower their risks of bankruptcy (Boyd, 2005). This theory is supported by Hellmal et al (2000), who find that an implementation of deposit interest rate ceilings decreases risks of bank failure. Allen and Gale (2004) describe the

“competition-‐fragility view”, which argues that bank competition causes banks to finance risky projects to increase their profit margins. This results in a higher probability of financial distress. In the case of government-‐ownership of banks competition will be much lower, so these problems are less likely to arise.

Lastly, the assumption that managers in privately owned banks are monitored better by their shareholders than managers in state-‐owned banks is rejected by

Altunbas (2001). He states that effective monitoring is not dependent on an ownership form and not weakened in a market, where it is not possible to trade ownership rights.

IV. Chapter 3

(12)

3.1- Comparing Nationalized and Private banks

Government ownership of banks for many years has been a frequently investigated topic in economics. In general, there are three different ways, in which papers judge

government ownership of banks. The first and most investigated way is to consider bank performance, based on economic growth, profits and efficiency. The second group

examines the difference in lending behaviour between private and government-‐owned banks, due to elections or other political incentives. The third group investigates what the effect is of bank ownership during a crisis (Cornett, 2010). In this chapter the results of all these three different judgements on the effect of bank ownership will be discussed.

3.2- Bank Performance based on economic growth, profits and efficiency

3.2.1 Effect on economic growth and the financial system

There are a lot of articles about the difference in performance between private and government-‐owned banks. In general most articles found that economies with a lot of private banks come with more economic growth, profits and efficiency (La Porta et al, 2002., Cole, 2009., Boyd and De Nicolo, 2005.). Government ownership of banks today is still common around the world, though it did decrease the last 40 years. The percentage of state-‐owned banks is especially high in countries, which are poor, have lots of

government interventions and underdeveloped financial systems. Government ownership is also higher in countries that have less political rights, or are less democratic. However, this does not implicitly mean that state-‐owned banks

underperform relative to private banks (La Porta, 2002). A lot of studies were done to try to find if there is a difference in performance.

La Porta (2002) did a big investigation by analyzing government ownership of large banks, measured in terms of assets, in 92 countries and compares them with the private banks in these countries. For each country, the then largest commercial and the then largest development banks (state-‐owned) are included. The countries are

(13)

investigated by running a regression over the 92 countries over a period from 1970 to 1995.

La Porta (2002) finds that the higher the degree of government ownership of banks, the slower the development of the financial system, the lower the economic growth and especially, the lower the growth of productivity. These negative effects are not weaker in less developed countries. This supports the political view of the effects of government ownership of banks. So government ownership of banks go along with slower financial and economic development, also in poorer countries. These results are consistent with Barth, Caprio and Levine (1999), who find that government ownership of banks is higher in countries with less developed financial systems. Still, these results do not prove causality between performance and ownership, as also other reasons can be given for the higher economic growth of developed countries. Cross-‐country

regressions suggest plausible relationship, but it does not provide evidence.

Cole (2009) uses a policy experiment of India in 1980 to evaluate the effect of bank ownership on financial and economic development. In 1980 the government of India nationalized all private banks with a deposit base above Rs. 2 billion. Comparable and smaller banks were kept private. The nationalization of the banks came along with strict policy rules, which were identical for public and private banks. In this way

differences in performances between banks can be accounted on bank ownership, instead of characteristics of the banks, or different stages of regulation. The

nationalization induced variation in the share of credit issued by public banks and therefore causal effects of bank nationalization on the economy could be interpreted.

Cole’s (2009) results showed that the nationalization first increased the quantity, but substantially lowered the quality of financial intermediation. Though more money was lent to agricultural borrowers, agricultural outcome did not improve. The bank nationalization may have even slowed the growth of employment in more developed sectors of trade and services. No effect on financial development was found nor

mentioned. Cole (2009) concludes that one shortcoming of his paper is that due to the fact that banks were nationalized according to their deposit base, the nationalized banks are significantly larger than the private (control) banks. As only the effects of

government ownership of banks on the larger banks could be evaluated, it is not sure if similar results would be generated if also smaller banks would be nationalized.

(14)

3.2.2 Effects on profits, efficiency and risks of the banks themselves

Ianotta (2007) investigated the effect of ownership structure on performance and risk in the European banking industry during a period from 1999 to 2004. In this article the effect of three types of bank ownership are measured, namely privately owned banks, government-‐owned banks and mutual banks. I will only elaborate on the results of the differences between the first two ownership types. The performance and risk of a sample of 181 large banks from 15 European countries are evaluated. A bank is

classified as large if it has total assets of at least € 10 billion. This information is based on the income statements, balance sheets and ownership information data from the investigated banks. When testing for systematic performance differences in ownership, Ianotta (2007) controls for bank characteristics, country and year effects, and specific macroeconomic growth differentials. The results show that private banks are more profitable than government-‐owned banks. These higher profits come from higher net returns on the assets, and not from a superior cost efficiency. In addition, government-‐ owned banks turn out to be riskier, as these banks have a poorer loan quality and higher insolvency risks than private banks. The paper came with curious results for the

banking activity of government-‐owned banks, as a larger share of their funding comes from the wholesale interbank and capital markets, their liquidity is higher and their investments in loans is lower compared to private banks (Ianotta 2007).

The general view in economic literature is, according to Boyd (2005), that, in case of increased competition, banks rationally choose more risky portfolios. Private banks, which are subject to more competition, would in this view be riskier. Boyd (2005) doubts this view and shows with his paper the opposite, namely that banks become more risky as the markets become more concentrated. This study focuses on the asset side of the balance sheet, which was not done before. As competition decreases, banks charge higher loan rates in the loan market, which led to higher bankruptcy risks for bank borrowers. So banks would use their increased market power to raise loan rates and borrowers, who are confronted with increased funding costs, would optimally choose riskier projects. There is assumed that borrowers totally determine project risk, conditional on the loan set by banks. In general the study suggests that government-‐ owned banks, which are subject to more market power and less competition, are riskier than private banks (Boyd, 2005).

(15)

Cornett (2010) investigated performance differences between state-‐owned and privately owned banks before, during and after the Asian financial crisis. The study examined year-‐end financial statement data from 1989 until 2004 for 16 Far East countries. This period includes the Asian financial crisis, which started in July 1997. The drop of the Thai baht relative to the U.S. dollar, was followed by devaluation of al lot of Asian currencies. A bank is accounted as state-‐owned if the government holds at least 20% of the shares.

Cornet’s (2010) results show that on average state-‐owned banks were less profitable, held less core capital and had greater credit risk than private banks during the period of 1989-‐2000. The reason for this was that state-‐owned banks had un-‐booked losses, which could no longer be kept secret during the crisis. Another result was also found, namely that during the years after the Asian crisis, the period of 2001-‐2004, state-‐owned banks recovered relatively quicker than private banks. This resulted that in the year 2004, cash flow returns, core capital and nonperforming loans were not

significantly different anymore from private banks. According to Cornett (2010) state-‐ owned banks could catch up due to the increase in globalization, which created an increase in financial competition. This has lead to a more sustainable banking policy that disciplines inefficient regulators.

Examples of countries with mixed results

Complementary to the last finding of Cornet (2010), there are also examples of individual countries where government-‐owned banks performed on average equally well, or even performed better in total than private banks.

Bhattacharyya (1997) tried to measure the variation in performance of Indian commercial banks, during the early years of financial liberalization in India from 1986 to 1991. Before this period most of the banking sector was under government protection and ownership since 1980. In 1969 already 14 major banks were nationalized and in 1980 a further 6 major banks were nationalized. This was due to the decision of the Indian government to nationalize all private banks with a deposit base above Rs. 2 billion . From 1986 to 1991, in response to changing international financial markets and as an effort to make the Indian banking system internationally more competitive

(16)

privatization took place and private banks were allowed to expand without the fear of nationalization. Efficiency of 70 Indian commercial banks was measured during this liberalization process, which revealed the differences in performance between privately owned and government-‐owned banks. Performance was associated with technical efficiency; which is the ability to transform multiple resources into multiple financial services. A bank is technical efficient when it is generating maximum output from the minimal amount of capital. Bhattacharyya (1997) calculates technical efficiencies by using data envelop analysis (DEA), a method often used when performance differences in relation to bank ownership are calculated. Variation in efficiency of banks is explained by decomposing variation into a systematic and a random component. From the

systematic component differences in efficiency are attributed to differences of bank ownership.

The results showed that government-‐owned banks were more efficient than privately owned banks. Government-‐owned banks utilized their resources better to deliver financial services to their customers. However, government-‐owned banks underperformed privately owned banks, when looking at profitability. So the

investigation of the Indian banking sector gave mixed results on the effect of ownership of banks on performance (Bhattacharyya, 1997).

Figueira (2009) investigated the difference in performance between state-‐owned and private-‐owned banks in Latin America in 2001. During the decade before 2001 in the Latin banking industry liberalization was common to most countries, which included privatization of banks. Private and government-‐owned banks were compared using cross-‐sectional data on financial and economic performance ratios, which provide insights into the results of reforms of ownership during the 1990s. The results showed that there was no evidence that private banks in Latin America in 2001 performed better than government-‐owned banks. The study even showed that differences in performance were more related to the national regulations and economic environment in which banks operate, instead of ownership.

An interesting exception

As followed from the results of the different articles above, most empirical

(17)

less efficiently. According to Altunbas (2001) this empirical evidence was especially found in non-‐financial firms and non-‐competitive markets. However Altunbas (2001) criticizes this view when applying it to financial firms and tries to show that public financial banks in a competitive market do not necessarily need to underperform in comparison with private financial firms. He investigated separate cost and alternative profit frontiers (stochastic frontier analysis) for three different owner types in the German banking period between 1989-‐1996: privately owned banks, state-‐owned banks and mutual cooperative banks. The sample consists of 1195 private banks observations, 2858 public bank observations and 3486 mutual cooperative bank observations. The breakdown of these banks is helped by the detailed classification by the Deutsche Bundesbank. We focus only on the differences between public and state-‐owned banks. Inefficiency differences are measured using the stochastic-‐frontier approach and the distribution-‐free approach. The stochastic-‐frontier approach classifies a bank as inefficient if its costs are higher, or profits lower, than predicted by an efficient bank producing the same input/output combination and where the difference cannot be explained by statistical noise. The cost or alternative profit frontier is derived by

estimating a cost or alternative profit function with a composite error term, the sum of a two-‐sided error term representing random fluctuations in cost and profit and a one-‐ sided positive error term representing efficiency (Altunbas, 2001).

Altunbas’ results show that there is little to suggest that privately owned banks are more efficient than state-‐owned banks in the German banking world. Both

ownership forms benefit from economies of scale and it seems that larger banks seem to realize greater economies of scale than smaller banks. The inefficiency measures show that public banks tend to have slight cost and profit advantages compared to the private banking counterparts. This may be explained by lower cost of funds. The findings show that there are no agency problems for public banks operating in the German banking market. This German example shows that public banks in a market economy with a stable and democratic government and a good governance system can be more profitable.

3.3- Difference in Lending Behaviour, due to elections or political invectives

(18)

Some articles have found evidence that state-‐owned banks have different lending behaviours than private banks. As politicians choose managers of state-‐owned banks, differences in lending behaviour may be due to elections or other political incentives (Sapienza 2002 and Dinc 2005).

Bank ownership and lending behaviour in Italy

According to Sapienza (2002) looking at bank performance is not the appropriate way to value the differences between state-‐owned and privately owned banks. As most empirical evidence shows that government ownership is less profitable, it is not clear if this is because state-‐owned banks maximize broader social objectives, or that they have lower incentives to maximize profit or because they inefficiently give in to political wishes. Instead of focusing on bank performance, Sapienza (2002) focuses on the lending behaviour of banks in correlation with ownership structure. The data used to determine differences in lending behaviour include information on the balance sheets and income statements of more than 37 000 companies in Italy. This data comes from “Centrale der Bilanci”, an institution created to provide its members, which are mainly banks, with economic and financial information for screening Italian companies. The information in this database has proven to be very precise in predicting the chances of success for a company. As both state-‐owned and privately owned banks use this information, any difference in credit policy is more likely to reflect differences in the objectives of the banks, instead of differences in the evaluation skills.

Sapienza (2002) looks at the individual loan contracts of the two types of banks and compares the interest rates charged to two different companies with the same score, where one borrows from an state-‐owned bank and the other from a privately owned bank. The main result is that when controlled for firm and bank characteristics, state-‐owned banks charge interest rates that are 44 basic points lower than interest rates charged by comparable privately owned banks. First is tested whether these results can be explained by the social view. As earlier mentioned Sapienza (2002) explains that according to the social view state-‐owned firms are created to overcome market failures, whenever the social benefits of state-‐owned firms exceed the costs. In this view state-‐owned banks improve economic development and improve general welfare. The hypothesis, that state-‐owned banks lend to firms for which borrowing from

(19)

private banks is too difficult or too expensive, is rejected. When the sample is restricted to firms that borrow both from state-‐owned and privately owned banks still a lower interest rate of 44 basic points is found at state-‐owned banks. Second, companies located in the south of Italy benefit more from state-‐owned companies than companies located in the north, which can be due to the fact that political patronage is more common in the south (supporting the political view), or because of the fact that the south is poorer and the government wants to channel funds to this depressed area (supporting the social view). As earlier mentioned according to the political view,

politicians who control banks would use their power over banks to win votes, instead of focusing on efficiency and maximizing profit. State-‐owned banks would give jobs or financial resources to friends and political supporters. As state-‐owned banks tend to favour especially large firms, the social view is rejected.

Finally, to give more evidence for the political view, Sapienza (2002) examined the relationship between interest rates, political affiliation of the bank and electoral results. The relationship between objectives of politicians and the lending behaviour of banks is investigated by collecting data on the political preferences of the top executives of the state-‐owned banks. The results show that the lending behaviour is indeed affected by the electoral results of the party affiliated with the bank. The party affiliation of the top executives of state owned banks come with a lower interest rate given by state-‐ owned banks in the provinces, where this associated party is stronger. Overall these results support the political view of state-‐ownership of banks and so suggest that state-‐ ownership goes along with political arbitrage (Sapienza, 2002).

Effect of bank ownership on lending behaviour of different countries

Dinc (2005) also investigated the political influences on government-‐owned banks in lending behaviour. In comparison with Sapienza (2002), this article considers many different countries; so that its results have more diverse support.

Rather than looking at the differences in interest rates charged, which followed from Sapienza (2002), he looked at the absolute increases in lending behaviour due to

political influences. Dinc (2005) wanted to know if the actions of banks are motivated by political concerns, given that politicians control state-‐owned banks. During elections politicians especially might use their power to use state-‐owned banks for political

(20)

purposes. If so, these state-‐owned banks would increase their lending in election years, so this is investigated. Actions of state-‐owned banks and private banks around elections are compared in major emerging markets over the period 1994-‐2000.

To isolate the effect of difference in ownership from macroeconomic factors and bank specific factors, three major issues need to be considered. First, an event needed to be identified, in which politicians may use state-‐owned banks for their own political goals. General elections were used as an event that could trigger politicians to influence state-‐owned banks to increase their chances of re-‐election. Second, institutional

differences across countries are controlled for by taking a firm-‐level, instead of a

country-‐level analysis. As banks are compared with each other within the same country, it is possible to control for many institutional differences across countries. Lastly,

previously documented differences in efficiency must be accounted for, so that political actions of state-‐owned banks can be distinguished from other differences between state-‐ owned and privately owned banks. This is done by comparing changes in actions of state-‐owned banks compared to private banks around election years, relative to other years.

The cross-‐country analysis strengthens the results, as elections occur in different years in different countries. This prevents a correlation pattern between elections and some other time-‐events in the world. The sample contains 43 countries, who all have free or partially-‐ free elections, so therefore countries as China, Egypt and Indonesia were left out of the sample. The ten largest banks in each country were examined, with the requirement that the country contains both state-‐owned and privately owned banks (so the U.S. was left out, Dinc, 2005).

The results show that state-‐owned banks increase their lending in election years relative to private banks. State-‐owned banks increase their lending with about 11% compared to their average lending amount during the election-‐year. In this way politicians were able to distribute rents to their supporters. Also the share of

government securities on the balance sheet is 50% greater in government-‐owned banks in emerging markets than in private banks. These facts indicate that political incentives influence the actions taken by state-‐owned banks and cannot be attributed to other differences between private banks and state-‐owned banks (Dinc, 2005). However, the results fail to find an identical election-‐year increase in developed economies. According to Dinc (2005) this can be due to several factors. State-‐owned banks in developed

The choice between private banks and state-owned banks

University of Amsterdam -­‐ Amsterdam Business