Geographical spillovers on the relation between risk-taking and market power in the US banking sector

https://doi.org/10.1016/j.najef.2018.05.005Get rights and content

Abstract

This paper investigates the relation between risk taking and market power in the US banking sector by introducing the effect of geographical spillovers caused by the transmission of risk taking among banks. For this purpose, we use spatial econometrics. Our results support a negative relation between risk taking and market power. The transmission of risk taking causes significant geographical spillovers, which increases the magnitude of the relation under analysis here. Moreover, the spillover effect is found to be stronger during the Subprime-crisis period when we noted an increased in the transmission of risk taking of banks.

Introduction

The banking sector possesses special characteristics compared to other sectors of the economy. Banks are important intermediaries that perform key functions to promote the good performance of the economy (e.g., Allen, 1990, Diamond, 1984, Williamson, 1986). Nevertheless, the banking sector is very fragile as well. Banks take deposits, which are highly liquid, and transform them into non-liquid assets, i.e. loans. In such an environment, banks cannot satisfy an unexpected increase in the withdrawal of their deposits, which can cause a bank failure (Chari and Jagannathan, 1988, Diamond and Dybvig, 1983, Jacklin and Bhattacharya, 1988). This fragility can be transmitted to other banks in the sector through interbank connections. Credit and portfolio linkages are two channels by which banks transmit risk to others. Interbank credits and bank portfolios introduce interdependence among banks given that a failure or disruption in one bank will reduce its payments to other banks (e.g., Allen and Carletti, 2006, Boyson et al., 2008, Cifuentes et al., 2005, Degryse and Nguyen, 2007).

The banking sector also provides mechanisms for mitigating the costs associated with transactions and market imperfections, promoting, in this way, key functions for the healthy performance of the economy, such as capital accumulation, technological innovation, and economic growth (Levine, 1997). Nevertheless, specific characteristics of the banking sector, such as asymmetric information and credit rationing, can have an effect on the level of market power of banks (e.g., Leland and Pyle, 1977, Stiglitz and Weiss, 1981). In this way, ensuring an optimal level of competition or market power becomes an objective for policy makers in charge of the banking sector as important as the level of risk taking of banks. Therefore, identifying the relation between risk taking and market power is of relevance for the policy design. A positive relation implies that relaxing regulation to promote competition will also decrease risk taking. Contrarily, a negative relation implies that policy makers have to decide which of these two variables is more important in order to focus their policies. Here, a controversy arises given that both theoretical and empirical evidence about this relation is not conclusive.

On the one hand, Chan et al., 1986, Besanko and Thakor, 1993, and Marquez (2002), among others, present theoretical evidence of a negative relation between risk taking and market power. This negative relation is supported empirically by Demsetz et al., 1996, Salas and Saurina, 2003, and Jiménez, López, and Saurina (2013), among others. On the other hand, Boyd and De Nicoló (2005) present theoretical evidence arguing for a positive relation between risk taking and market power. A positive relation is also supported empirically by Jayaratne and Strahan, 1998, Boyd et al., 2006), among others.

We introduce an additional factor affecting the relation between risk taking and market power; geographical spillovers, which is supported by two arguments. First, risk taking can be transmitted across banks. That is, increases in risk taking in one bank affect the risk taking of others because of the interbank connections. Second, risk taking can be strongly transmitted to banks that are near to each other. Geographical proximity is found to effect bank intermediation (Degryse and Ongena, 2005, Presbitero and Rabellotti, 2014, Zhao and Jones-Evans, 2016). Moreover, evidence shows that distance is a factor for the transmission of risk taking among banks (Aharony and Swary, 1996, Freixas et al., 2000). Our approach is in line with Pino and Sharma (2018) who find significant transmission of risk for banks that share similar characteristics and banks that operate in a same US state. However, we focus on geographical spillovers and, in particular, on the relation between risk taking and market power. To this end, we use geographical distance to capture the transmission of risk taking across banks allowing us to analyze how geographical spillovers affect the relation between risk taking and market power. To do so, we exploit the advantages of spatial econometrics. This procedure provides a way to include the transmission effect as an additional factor that affects the risk taking of banks. This effect has been studied mostly through the abnormal returns of the stock prices of banks (e.g., Aharony and Swary, 1983, Kanas, 2005, Madura and Tucker, 1991), or by simulations of the transmission of a bank’s failures (e.g., Ladley, 2013, Sachs, 2014, Zedda et al., 2014).

The objective of this paper is to study the impact of geographical spillovers on the relation between risk taking and market power. Our contribution is threefold. First, in spite of the fact that the transmission of risk taking is a documented issue (e.g., Akhigbe and Madura, 2001, Cornett et al., 2005, Iyer and Puri, 2012), it has not been included as a factor of the relation between risk taking and market power. In particular, an appropriate methodology, which connects different banks, is necessary to incorporate a transmission effect as an additional determinant of risk taking. We use spatial econometrics to deal with this issue. Second, spatial econometrics plays a central role for the purpose of this paper since this procedure allows decomposing the relation between risk taking and market power into direct and spillover effects where the latter is caused by the geographical transmission of risk taking. Thus, we are able to evaluate the significance and magnitude of the effect that is neglected when the transmission effect is not incorporated as an additional factor in the risk taking of banks. Finally, we investigate how the relation between risk taking and market power is affected by geographical spillovers during the crisis period. Analyze variations on the relation between risk taking and market power over time is of relevance to preserve the stability of the banking sector.

We estimate the relation between risk taking and market power by using annual data of the balance sheet and income statements of US Commercial banks from 2001 to 2012. Our findings show significant geographical spillovers from 2003 to 2012, which presents a significant increase during the crisis period as is to be expected. The relation between risk taking and market power is found to be negative which increases in magnitude due to geographical spillovers, especially during the crisis period. Finally, our results show that the spillover effect increases with the number of interbank linkages in the geographical space which also increases the relation under analysis in this study.

This paper is organized as follows. Section 2 presents the methodological approach. Sections 3 Data and descriptive statistics, 4 Results and discussion discuss the data and the main results, respectively. Finally, in Section 5 we present the conclusions of this study.

Section snippets

Empirical approach

In order to measure the market power of a bank i, the Lerner index is computed here. The greater the Lerner index, the greater the market power of bank i. The Lerner index is a proxy of the degree of competition of bank i (e.g., Boyd et al., 2006, Yeyati and Micco, 2007). We use the Lerner index since this measure incorporates the heterogeneity among banks1 which is

Data and descriptive statistics

This paper uses information contained in the income statements and balance sheets of US commercial banks. Net income, total capital, and total assets are used to compute the Z-score. The average interest rate charged for loans, inter-banking interest rate, non-accrual loans, and total loans are used to compute the Lerner index. Finally, total cash, total operating expenses, total loans, and total assets are used to obtain the controls for the estimation in Eq. (7). The source of these variables

Results and discussion

Table 2 presents the LM statistic for testing spatial dependence across all the years studied here. The benchmark case presents no significant spatial dependence for all years from 2001 to 201210 (column 1 in Table 2). This

Conclusions

The relation between risk taking and market power is relevant for its policy-making implications. For instance, a negative relation implies a trade-off between competition and financial stability causing that policy makers have to decide which of these two variables is more important to focus their policies. This paper investigates the relation between risk taking and market power in the US banking sector by taking geographical spillovers into account. There is an extensive literature that

Acknowledgement

We would like to thank the two anonymous referees for their valuable comments which have improved the quality of this paper.

Gabriel Pino is an Assistant Professor of Economics at the Faculty of Business and Economics of the Universidad de Talca, Chile. He completed a PhD in Economics at the Southern Illinois University Carbondale in 2013. His current research interests include banking and regional economics.

References (58)

  • A.M. Oliver et al.

    Risk premium and market power in credit markets

    Economic Letters

    (2006)
  • V. Salas et al.

    Deregulation, market power, and risk behavior in spanish banks

    European Economic Review

    (2003)
  • L. Tonzer

    Cross-border interbank networks, banking risk and contagion

    Journal of Financial Stability

    (2015)
  • C. Upper et al.

    Estimating bilateral exposures in the german interbank market: Is there a danger of contagion?

    European Economic Review

    (2004)
  • S.D. Williamson

    Costly monitoring, financial intermediation, and equilibrium credit rating

    Journal of Monetary Economics

    (1986)
  • E.L. Yeyati et al.

    Concentration and foreign penetration in Latin American banking sectors: Impact on competition and risk

    Journal of Banking and Finance

    (2007)
  • J. Aharony et al.

    Contagion effects of bank failures: Evidence from capital markets

    Journal of Business

    (1983)
  • F. Allen et al.

    Financial contagion

    Journal of Political Economy

    (2000)
  • F. Allen et al.

    Systemic risk and regulation

  • L. Anselin

    Spatial econometrics: Methods and models

    (1988)
  • A. Babus

    The formation of financial networks: Discussion Paper 06-09

    (2007)
  • D. Besanko et al.

    Relationship banking, deposit insurance and bank portfolio choice

    (1993)
  • J.H. Boyd et al.

    The theory of bank risk taking and competition revisited

    Journal of Finance

    (2005)
  • J. Boyd et al.

    Banking risk taking and competition revisited: new theory and new evidence: Working Paper 06/297

    (2006)
  • N.M. Boyson et al.

    Hedge fund contagion and liquidity. Working Paper 14068

    (2008)
  • V.V. Chari et al.

    Banking panics, information, and rational expectations Equilibrium

    Journal of Finance

    (1988)
  • R. Cifuentes et al.

    Liquidity risk and contagion

    Journal of European Economic Association

    (2005)
  • H. Degryse et al.

    Interbank exposures: An empirical examination of systemic risk in the Belgian banking system

    International Journal of Central Baking

    (2007)
  • H. Degryse et al.

    Distance, lending relationships, and competition

    Journal of Finance

    (2005)
  • Cited by (2)

    • Rural banking spatial competition and stability

      2023, Economic Analysis and Policy
    • Carbon neutrality, bank lending, and credit risk: Evidence from the Eurozone

      2021, Journal of Environmental Management
      Citation Excerpt :

      The construction of these variables is explained in Table 2. In the lending industry, market power significantly influences the institutions' risk-taking behavior (Pino et al., 2019). As the market power stems from the size, we repeat our analysis on three size sorts.

    Gabriel Pino is an Assistant Professor of Economics at the Faculty of Business and Economics of the Universidad de Talca, Chile. He completed a PhD in Economics at the Southern Illinois University Carbondale in 2013. His current research interests include banking and regional economics.

    Rodrigo Herrera is an Associate Professor of Economics at the Faculty of Business and Economics of the Universidad de Talca, Chile. He completed a PhD in Economics at the Dresden University of Technology in 2009. His current research interests include energy economics, quantitative finance, and extreme value theory.

    Alejandro Rodríguez is an Assistant Professor of Economics at the Engineering Faculty of the Universidad de Talca, Chile. He completed a PhD in Business Administration and Quantitative Methods at the Universidad Carlos III. His current research interests include Econometrics, Financial Economics, and Computational Economics.

    View full text