Elsevier

Journal of Corporate Finance

Volume 29, December 2014, Pages 122-142
Journal of Corporate Finance

Risk taking behavior of privatized banks

https://doi.org/10.1016/j.jcorpfin.2014.07.007Get rights and content

Highlights

  • We examine the risk taking behavior of privatized banks pre- and post-privatization.

  • Privatized banks experience a significant decrease in risk after privatization.

  • A higher fraction of the privatized banks' shares sold induces higher risk taking.

  • The relationship between private ownership and risk taking is somewhat U-shaped.

  • Risk taking by privatized banks is affected by the country's level of development.

Abstract

We examine the risk taking behavior of privatized banks prior to and after privatization and find that privatized banks experience a significant decrease in risk after privatization; however they continue to exhibit higher risk than their rivals. This finding is consistent with the assertion that following privatization and the removal of government guarantees and subsidies, privatized banks become more prudent. Since rival banks do not experience a significant change in risk taking, we attribute the reduction in risk experienced by the privatized banks to changes in the banks' ownership structure rather than to industry factors. Interestingly, we also find that a higher fraction of the privatized banks' shares sold beyond a certain intermediate level induces higher risk taking, as the privatized bank becomes more accountable to shareholders. The finding that the fraction of shares sold is positively related to risk taking, coupled with the result that the privatized banks had higher risk in the pre-privatization period than in the post-privatization period suggests a nonlinear relationship between government/private ownership of banks and risk taking. Results of further analysis are consistent with a somewhat U-shaped relationship between private ownership and risk taking. The risk taking behavior of newly privatized banks is also influenced by the country's level of development and degree of political risk. Our results are robust to different measures of risk.

Introduction

Most studies on bank privatization focus on the operating and stock market performance of newly privatized banks and generally document performance improvement for the privatized banks (see Megginson (2005) for a review of the literature). We extend this literature by examining the effects of privatization on risk taking by newly privatized banks in an environment of increased competition. As a result of the competitive nature of the newly privatized banks' operating environment, and the change in the firms' objective to profit maximization, we expect privatization to affect the banks' risk taking behavior. Understanding how privatization affects the risk taking behavior of banks is important because as the recent financial turmoil illustrates, bankers who take more risks could be rewarded with higher returns but their risk taking behavior can cause problems for the entire financial and economic systems. Given the wave of government bailout of banks around the world during the recent financial crisis (some of which were former state-owned banks), which led to an increased role of governments in these banks, understanding the risk taking behavior of privatized banks is all the more important.

The impact of privatization on the risk taking behavior of newly privatized banks is an interesting empirical question in itself. On one hand, it can be argued that privatization increases newly-privatized banks' risk taking. Since state-owned banks are not driven by principles of profit-maximization, because they are used to promote the government's economic and social agenda and to maximize social stability through the provision of employment and credit (Boycko et al., 1996, Shleifer and Vishny, 1994), and to ensure success of future elections, they are likely to pursue conservative investment strategies (Boubakri et al., 2013). In addition, in a less contestable market (pre-privatization period), state banks earn higher monopoly rents, and hence they will not jeopardize their position by taking excessive risk (Allen and Gale, 2000).1 However, once the government privatizes the banks, the banking sector becomes more contestable and the increase in competition that follows can raise the privatized banks' incentive to take on higher risk.2 Furthermore, following privatization and the attendant change in the objective of the firm to value maximization, privatized banks, now accountable to shareholders, could become more aggressive in their quest to create value for shareholders. In addition, liberalizing the banking sector and privatizing banks provide banks with greater freedom and more opportunities to take on risk (Gonzalez, 2005). It is therefore reasonable to surmise that risk taking by the privatized banks will increase in the post-privatization period.

On the other hand, there are good reasons to expect that government ownership of banks could induce higher risk taking in the pre-privatization period and that the banks' risk taking incentives could decrease after privatization. First, state-owned banks are used to implement governments' political agenda. Due to political exigencies, state banks tend to extend credit with little economic justification (sometimes at below market rates) to usually risky sectors such as agriculture and they are occasionally coerced into making economically questionable loans to friends of politicians (Faccio et al., 2006). In addition, the government guarantees enjoyed by state-owned banks and soft budget constraints can encourage managers to take higher risk, sometimes for personal gains. After privatization the banks may adopt more prudent lending practices and may shift lending away from state-owned enterprises (SOE) to more creditworthy clients who meet the banks' new and presumably prudent lending standards. Moreover, with the removal of government subsidies following privatization, the banks could become more cautious and less risky. Gropp et al. (2012) find that banks whose government guarantees were removed cut their credit risk and loan sizes. The foregoing argument suggests that privatization could have ambiguous influence on bank's risk taking incentives. Therefore, whether bank privatization leads to an increase or decrease in the risk taking behavior of newly privatized banks is an empirical question.

While extant literature provides ample evidence on the operating performance of newly privatized banks, there is little evidence on bank's risk taking behavior, and the scant empirical evidence that exists is mixed. For example, Boubakri et al. (2005) suggest that ownership type affects the risk exposure of privatized banks. Jia (2009) also finds that privatization in China has provided incentives for banks to engage in prudent lending. Though not specifically on banks, Boubakri et al. (2013) find that state ownership is negatively related to corporate risk taking. Our work complements the study of Boubakri et al. (2013) by examining in a global context whether competitive pressures following privatization affect newly privatized banks' risk taking behavior.

Though our study is similar to that by Boubakri et al. (2013) in its focus on newly privatized firms, there are significant differences between the studies. First, Boubakri et al. focus on newly privatized non-financial firms and exclude privatized banks from their study because banks are heavily regulated and are therefore sensitive to regulatory burden (as argued by Faccio et al., 2011). We capitalize on this unique feature and examine the risk taking behavior of former state-owned banks when a key aspect of the regulatory burden (being state-owned banks) is lifted through privatization. Second, our focus on bank privatization and risk taking makes our study different from Boubakri et al.'s because public sector banks, unlike other state-owned enterprises, were used by governments for priority sector lending to other SOEs regardless of the risk. Thus political exigency makes the risk taking behavior of these banks in the pre- and post-privatization periods different from that of other (non-financial) newly privatized firms. Third, we examine the impact of privatization on bank risk taking, rather than the determinants of post-privatization risk taking (as done by Boubakri et al., 2013, pg. 3). The difference in objective is important because by focusing on the impact of privatization, we are able to employ a research design that allows us to control for industry-wide changes in risk taking which, for example, could result from liberalization of the financial sector. This research design ensures that the changes in risk taking that we observe for the privatized banks are not industry-wide phenomena but rather can be attributed to privatization.

Our results are summarized as follows. First, we find that prior to being privatized the state-owned banks exhibited higher risk (as measured by the z-score) than rival banks. Second, the privatized banks experienced a significant reduction in risk in the post-privatization period. Consistent with the findings based on the z-score, the privatized banks also experienced significant reduction in the volatility of return on equity (ROE) and volatility of return on asset (ROA), a decrease in the non-performing loans, and an improvement in solvency. These results are consistent with the hypothesis that privatized banks have become more prudent and less risky after privatization.

Having documented that privatized banks experienced a reduction in risk, we examine the possibility that the observed changes in risk documented for the privatized banks is an industry-wide phenomenon. We do so by examining the risk taking behavior of rival banks and find that unlike the newly privatized banks, which exhibit significant reduction in all the risk measures after privatization, rival banks do not experience significant changes in the z-score or any of the other risk measures except volatility of ROE. Even after controlling for industry -wide influences on risk taking by banks, the reduction in risk documented for the privatized banks remains significant. Despite the reduction in risk, however, newly privatized banks continue to exhibit higher risk than their rivals in the post-privatization period; however, the difference in risk in the post-privatization period is significantly less than that in the pre-privatization period.

Interestingly, we also find that a higher proportion of the bank's shares sold induces higher risk taking by the banks, as they become more accountable to shareholders. Consistent with this result, we observe that fully privatized banks on average exhibit higher risk taking than partially privatized banks. The finding of a positive relationship between fraction of shares sold and bank risk taking is similar to that of Boubakri et al. (2013) who find a positive (negative) relationship between foreign ownership (government ownership) and risk taking by privatized firms in the post-privatization period. The finding that fraction of shares sold is positively related to risk taking, combined with the finding that the privatized banks had higher risk in the pre-privatization period than in the post-privatization period points to a non-linear relationship between government/private ownership and risk taking. We further examine this nonlinear relationship between private ownership and risk taking by the privatized bank and find that the relationship is somewhat U-shaped. The U-shaped relationship remains even after we control for firm size, issue, and country characteristics.

The finding of a U-shaped relationship between private ownership and risk taking is new and it suggests that both full government ownership and full private ownership induce higher risk taking. However, the incentives to take high risks under full government ownership are different from those under full private ownership. For the former, political exigencies to lend to usually riskier segments of the market at concessionary rates, soft budget constraints and the availability of government subsidies enjoyed by state-owned banks and the possibility of bailout in case of financial difficulties can make public sector banks more risk loving. For fully privatized banks that are wholly accountable to their new owners, the pressure to create value for the shareholders can make the banks more aggressive, which in turn can induce higher risk taking. Our analysis suggests that an 80–20 public–private partnership (20% partial privatization) seems to be the optimal level of privatization that minimizes risk, as beyond that level of privatization, the risk increases with increase in private ownership of banks.

We also examine how privatization interacts with the country's institutional characteristics in shaping the risk taking behavior of privatized banks and find that privatized banks in developed countries exhibit lower risk than those in developing countries. We also find that banks privatized through the stock market (SIP) in developed and non-common law countries exhibit more aggressive risk taking behavior than those in developing and common law countries, respectively. Our results are robust to the inclusion of industry factors and the use of different measures of risk.

Our paper contributes to the literature in a number of ways. First, most studies on bank privatization focus on operating and stock market performance. We extend the literature providing evidence on an important part of the privatization literature — the effect of privatization on risk taking by newly privatized banks. This aspect of privatization is important because decisions on the amount of risk taken by a bank are crucial to the bank's profitability and the survival of the banking system as a whole. To our knowledge, our work is the first to systematically examine changes in the risk taking behavior of newly privatized banks before and after privatization and to document a somewhat U-shaped relationship between private ownership of banks and risk taking. Boubakri et al. (2013) examine the role of state and foreign ownership in risk taking by a sample of newly (non-financial) privatized firms in the post-privatization period; our work focuses on the impact of bank privatizations and compares newly privatized banks' risk taking behavior before and after privatization to that of their rivals. This analysis allows us to control for industry-wide influences on risk taking. We also examine the impact of country, firm, and privatization characteristics such as the form of the divestiture (i.e., SIP or asset sales) on risk taking by newly privatized banks in order to gain further insights into the main drivers of risk taking by newly privatized banks.

Furthermore, we go beyond the traditional focus on earnings-based measures of risk by examining both market-based and bank-specific measures of risk, such as the z-score and idiosyncratic volatility. Thus, we offer greater insights into the risk taking incentives of privatized banks. For example, our study contributes to the extant literature by showing, among other findings, that the systematic risk of privatized banks is higher than that of rival banks. This finding highlights the importance of accounting for risk in the estimation of expected returns of privatized firms, and suggests that prior studies that use the market-adjusted model to estimate expected returns of privatized firms may have overestimated the excess returns.

The remainder of the paper is organized as follows. We develop our hypotheses in Section 2. Section 3 presents the data and measures of risk taking. In Section 4, we discuss the initial results of the effect of privatization and risk taking by the privatized banks relative to those of the rival banks. In Section 5, we focus on the privatized banks in greater detail and examine the effects of transaction- and firm-specific factors on the risk taking behavior of the sample firms. We present a series of robustness tests in 6 Robustness checks, 7 Summary and conclusion concludes the paper.

Section snippets

Bank privatization and risk taking: testable predictions

Extant literature identifies two theoretical views of privatization that relate to risk taking, namely the political and managerial incentive views. The political view suggests that politicians use SOEs to promote social objectives through the provision of employment and subsidies to supporters in return for political contributions and votes (Boycko et al., 1996, Shleifer and Vishny, 1994). Consequently, in order to ensure success of future elections, governments use SOEs to pursue conservative

Data description

Our test sample consists of 242 bank privatizations in 42 countries that occurred between 1988 and 2007.4 We obtain the list of privatized banks from Megginson et al. (2005). This source was augmented with more recent privatization transactions from the World Bank

Univariate tests: accounting risk measures

We first estimate a correlation matrix for the explanatory variables and present the results in Panel A of Table 2.The matrix shows that the banks in countries of low political risk and in developed countries exhibit lower risk (as measured by higher z-score) than those in developing countries. However banks in common law countries with better investor protection exhibit higher risk than those in civil law countries. Our sample of privatized banks become more stable after privatization,

Risk taking behavior of newly privatized banks: further analysis

Having documented that privatized banks experience a reduction in risk after privatization, we focus entirely on the newly privatized banks to gain further insights into the changing risk taking behavior of the former state-owned banks over the pre- and post-privatization periods. This analysis allows us to examine changes in risk taking by the privatized banks in greater detail and to gauge the effect of transaction and firm-specific factors on their risk taking behavior. We estimate the

Autocorrelation

By construction, both the volatility of ROA and the z-score suffer from autocorrelation, as we use a 5-year moving average to estimate the volatility of ROA as well as that of ROE. To reduce the impact of autocorrelation, we restrict our analysis to observations that are three years apart and hence are less affected by the issue of autocorrelation. The results of this restricted regression are presented in Table 9.12

Summary and conclusion

We examine the risk taking behavior of newly privatized banks and find that prior to privatization the banks were riskier than their rivals. Following privatization, however, the newly privatized banks experience a significant decrease in risk (as measured by the z-score, volatility of ROE and volatility of ROA as well as a decrease in the ratio of non-performing loans to total loans). Having documented evidence of a reduction in risk for the privatized banks, we explore the possibility that

Acknowledgments

We thank an anonymous referee, Arnold R. Cowan, William Megginson, and participants at the Northern Finance Association and SFM (Securities and Financial Markets) conference for their helpful comments and suggestions. This work was supported, in part, by grants from the Social Science and Humanities Research Council (SSHRC) of Canada, Grant number 410-2010-1737.

References (56)

  • R.C. Merton

    An analytic derivation of the cost of deposit insurance and loan guarantees

    J. Bank. Financ.

    (1977)
  • R. Morck et al.

    Management ownership and market valuation: an empirical analysis

    J. Financ. Econ.

    (1988)
  • I. Otchere

    Do privatized banks in middle and low-income countries perform better than rival banks? An intra-industry analysis of bank privatisation

    J. Bank. Financ.

    (2005)
  • I. Otchere

    Competitive and value effects of bank privatization in developed countries

    J. Bank. Financ.

    (2009)
  • S. Park

    Risk-taking behavior of banks under regulation

    J. Bank. Financ.

    (1997)
  • F. Allen et al.

    Financial contagion

    J. Polit. Econ.

    (2000)
  • M. Arellano et al.

    Some test specification for panel data: Monte Carlo evidence and an application to employment equations

    Rev. Econ. Stud.

    (1991)
  • A. Berger et al.

    Bank competition and financial stability

    J. Financ. Serv. Res.

    (2009)
  • M. Boycko et al.

    A theory of privatization

    Econ. J.

    (1996)
  • J. Boyd et al.

    The theory of bank risk-taking and competition revisited

    J. Financ.

    (2005)
  • Y.-S. Chan et al.

    Information reusability, competition and bank asset quality

    J. Bank. Financ.

    (1986)
  • H. Chang

    State-owned Enterprise Reform, United Nations, Department For Economic and Social Affairs (UNDESA) Policy Notes

    (2007)
  • A. Demirgüç-Kunt et al.

    Deposit insurance around the world: where does it work?

    J. Econ. Perspect.

    (2002)
  • A. Dyck

    Privatization in Eastern Germany: Management Selection and Economic Transition

    Am. Econ. Rev.

    (1997)
  • J. D'Souza et al.

    Does government ownership affect the cost of debt? Evidence from privatization

    Rev. Financ. Stud.

    (2011)
  • M. Faccio et al.

    Political Connections and Corporate Bailouts

    J. Financ.

    (2006)
  • M. Faccio et al.

    Large shareholder diversification and corporate risk- taking

    Rev. Financ. Stud.

    (2011)
  • M. Flannery et al.

    The effect of interest rate changes on the common stock returns of financial institutions

    J. Financ.

    (1984)
  • Cited by (33)

    • Onshore guarantees for offshore loans and bank risk-taking: Evidences from Taiwanese banks

      2022, Research in International Business and Finance
      Citation Excerpt :

      If the board size is larger than the median of all sample banks, then it is classified as a large board size bank; otherwise, it is classified as a small board size bank. Referring to Laeven and Levine (2009); Pathan (2009); Zhang et al. (2012), and Mohsni and Otchere (2014), this study employs control variables, including bank size (natural logarithm of a bank’s total assets), ROE (a bank’s return on equity), and revenue growth rate (the ratio of a bank’s revenue growth between consecutive years). Besides, two macroeconomic variables related to monetary policy and economic condition are also included, M1B growth rate, the annual growth rate of money supply, which is the aggregate amounts of currencies, check deposits, demand deposits, and demand saving deposits, represents the monetary policy; GDP growth rate, the annual growth rate of gross domestic product (GDP), is a measure of the country’s economic growth.

    • Did China's bank ownership reform improve credit allocation?

      2022, European Economic Review
      Citation Excerpt :

      Due to the inherent inefficiency of state ownership, many countries initiated bank reforms that fully or partially privatized SOBs. Many cross-country studies find that ownership reform improves bank performance, bank efficiency, and reduces risk exposure (Boubakri et al., 2005; Williams and Nguyen, 2005; Mohsni and Otchere, 2014). In addition, many single-country studies find similar results using the experience of China (Berger et al., 2009; Jiang et al., 2013), of Brazil (Beck et al., 2005a; Nakane and Weintraub, 2005), of Nigeria (Beck et al., 2005b), of Pakistan (di Patti and Hardy, 2005), of Argentina (Berger et al., 2005), of Egypt (Omran, 2007), and of Indonesia (Shaban and James, 2018).

    • Do negative interest rates affect bank risk-taking?

      2021, Journal of Empirical Finance
    View all citing articles on Scopus
    View full text