Corporate Governance and Banking Systemic Risk: A Test of the Bundling Hypothesis

Highlights

• For the systemic risk external and internal governance complement each other.

• Complementary effect of governance increases systemic risk contributions of banks.

• Banks have strategic flexibility in terms of configuring their governance structure.

Abstract

We provide new evidence that the systemic risk of large banks is higher when external and internal corporate governance mechanisms complement each other. Using a sample of large European banks from 2000 to 2016, we examine the relationship between various internal and external corporate governance mechanisms and the level of systemic risk. Specifically, we analyze how monitoring by institutional investors complements or substitutes various board-level governance mechanisms in determining the systemic risk of a bank. Our empirical findings show that external (institutional ownership) and internal (board level) governance mechanisms complement each other to determine the level of systemic risk of a sample of domestic systemically important banks. Our results are robust to alternative systemic risk measures and additional controls. We conclude that banks have strategic flexibility in terms of configuring their corporate governance structures to attain similar levels of systemic risk.

Introduction

“Most studies of board effectiveness exclude financial firms from their samples. As a result, we know very little about the effectiveness of banking firm governance.”

(Adams and Mehran, 2012, p. 243).

This paper examines whether the systemic risk of financial institutions is associated with shareholder-friendliness of corporate governance mechanisms. The recent global financial crisis can be partially attributed to the weaknesses of the corporate governance mechanisms of financial institutions (Basel Committee on Banking Supervision, 2010). Stronger corporate governance mechanisms can change the willingness of managers to take more risk (John, Litov & Yeung, 2008) but in financial institutions stronger corporate governance mechanisms can also result in excessive risk-taking (Erkens et al., 2012). For financial institutions, this excessive risk-taking can lead to undercapitalization. Therefore, this study investigates whether strong corporate governance mechanisms are related to the systemic risk contribution of financial institutions.

In contrast to previous studies see e.g., Pathan, 2009, Saunders et al., 1990 Laeven and Levine, 2009, John et al., 2008, that focus on idiosyncratic risks only, we pay attention to the systemic risk contribution of financial institutions.¹ Despite the acknowledgement of corporate governance (CG) as a tool to determine risk appetite and help a firm manage their risk portfolio (John et al., 2008), the role of CG in determining the systemic risk contribution of banks has received very little scholarly attention. This is surprising, as the board of directors ultimately determine the actions of the bank, which in-turn determine its risk outcomes (Forbes and Milliken, 1999). Instead, the literature primarily focuses on the estimation of systemic risk (Billio et al., 2012; Adrian and Brunnermeier, 2016; Acharya et al., 2017; Huang, De Haan and Scholtens, 2020). Consequently, the resulting metrics do not account for and/or investigate corporate governance as a driver of systemic risk contributions.

Furthermore, implicit and explicit government guarantees, highly leveraged capital structure, and too-big-to-fail (TBTF) policies encourage banks to take more risks (see e.g., Acharya et al., 2016, Abdelbadie and Salama, 2019). As a result, banks may not only increase their bank-specific risk but also create negative externalities for the financial system by increasing the aggregate level of systemic risk and undercapitalizing the system (De Haan and Vlahu, 2016, Brownlees and Engle, 2017). This warrants researchers to focus on the role of corporate governance in propagating or containing systemic risk contributions by banks. Yet, the literature in this direction remains insufficient and under-explored. The literature primarily explore the broad relations between overall strength of corporate governance structures and systemic risk (see e.g., Iqbal, Strobl & Vahamaa, 2015). However, there is limited evidence on the relationship between individual corporate governance mechanisms and banks’ systemic risk contributions. This study aims to contribute to the debate about the determinants of systemic risk by examining how internal and external corporate governance mechanisms relate to the systemic risk of European banks.

Our study is motivated by two considerations. First, firms employ several governance mechanisms simultaneously, in the form of governance bundles, which jointly determine outcomes (Rediker & Seth, 1995) and protect the interests of shareholders. Hence, we maintain that the level of a particular mechanism is ideally dependent on the levels of other mechanisms which are simultaneously in place in a particular bank. Our point of departure from the extant literature is as follows: we consider multiple corporate governance mechanisms by examining both the individual and interactive effects of various corporate governance mechanisms (governance bundles) on a bank’s systemic risk contribution. Since there is limited theory as to the most important board characteristics, the term ‘strong boards’ focuses on an ad hoc selection of board mechanisms which have been theoretically emphasized as effective in monitoring and aligning the interests of managers and shareholders; appropriate board size, board independence, female directorship and board meetings. Unlike Iqbal et al. (2015), the initial part of analysis investigates the individual rather than the effect of an indexed-measure of bank board mechanisms on systemic risk. We argue that using an index measure of board mechanisms undermines a deeper understanding of how a single bank board mechanism can influence the level of bank systemic risk.

Board size and its negative relation to bank risk is a common finding in the literature (Cheng, 2008, Pathan, 2009) due to potential free-riding problems, less cohesiveness, and high communication and coordination costs associated with larger boards (Jensen, 1993). Also, since an individual director’s incentive to acquire information and monitor managers is low on large boards, CEOs may find it easier to pursue their risk-averse preferences on (Jensen, 1993). Independent directors are believed to be better monitors of managers as independent directors value maintaining their reputation in the directorship market (Fama and Jensen, 1983, Bhagat and Black, 2002). Existing studies have empirically shown that firms with female representation on their boards lead to better firm performance (Adams and Ferreira, 2009; Lückerath-Rovers, 2013). Adams and Ferreira (2009) explain that women contribute to discussions and exchange of ideas from a diverse perspective which enhances the monitoring potential of the board of directors. Vafeas (1999) shows that years preceding better firm performance exhibit increased frequency in board meetings, suggesting that board meetings are an effective mechanism for monitoring executive behavior.

Similar to Pathan, 2009, Iqbal et al., 2015, we expect a strong board to effectively monitor managers so that they work for the shareholders and curtail the bank’s systemic risk contribution. In contrast, strong boards, especially in banks, can encourage managers to take excessive risk demanded by bank shareholders to maximize their wealth. This appetite for excessive risk-taking is further multiplied by the ‘moral hazard’ problem associated with the ‘too-big-to-fail’ phenomenon and deposit insurance schemes (Galai and Masulis, 1976, Saunders et al., 1990, Martinez Peria and Schmukler, 2001). Previous studies also argue that it is beneficial for banks to become large or achieve the status of too-big-to-fail to reap the benefits of implicit and explicit government funding subsidies (Brewer and Jagtiani, 2013). To achieve this status, banks may adopt riskier policies which can translate into greater systemic risk contributions. The recent global financial crisis demonstrated the negative side of systemic risk, where the interconnectedness of financial institutions resulted in the collapse of the financial system (Brunnermeier, Dong and Palia, 2020).

Second, existing literature advertises that the effect of governance mechanisms on bank risk is mainly dependent on the existing ownership structure (Choi and Hasan, 2005, Martín-Oliver et al., 2017). Empirically, Laeven and Levine (2009) show that the intended consequence of regulatory capital on risk-taking is attenuated when banks have large or concentrated ownership. In furtherance of this, the description of dispersed ownership with regard to the separation of ownership and control has been presumed to be universally applicable (Berle and Means, 1932). However, Fernández and Arrondo (2005) emphasize that managerial actions are mainly controlled by the board of directors and large shareholders in the European economy.² This suggests that direct control and monitoring by large (institutional) shareholders prevails as a fundamental and/or strong mechanism to increase managerial risk-taking. In this regard, Hoskisson et al., 2002, Connelly et al., 2010 show that institutional investors strongly influence a firm’s internal innovation and support risk-taking behavior. Ultimately, the key question which remains underexplored in the existing banking literature is whether the simultaneous existence of internal and external governance mechanisms limit or promote the systemic risk contribution of a bank. In our study we attempt to explore the role of institutional investors coupled with internal governance mechanisms on the systemic risk of banks.

To address this, we conduct our analysis using data from 38 European banks classified as Domestic Systemically Important Banks (D-SIBs) for the years 2000–2016. We use a market–based equity measure, absorption ratio (AR), proposed by Kritzman, Li, Page and Rigobon (2011) to proxy the systemic risk contribution by a bank. Our findings show that although strong boards have a varying effect on bank systemic risk, they synergistically promote prevailing bank systemic risk in the presence of external monitors (institutional shareholders). This evidence informs our conclusion that internal and external governance mechanisms mainly act as complements.

Our study makes several contributions to the literature. First, we extend the scope of bank risk by operationalizing a financial econometric estimation of systemic risk, namely the absorption ratio (AR). Second, most of the previous studies on corporate governance bundling have explored the interactive effects of multiple governance mechanisms without a single focus (see e.g., Rediker and Seth, 1995, Schepker and Oh, 2013, Zajac and Westphal, 1994). This arguably attenuates a deeper understanding of the role of a single mechanism conditioned on other mechanisms. In this sense, our study contributes by extending the theoretical boundary into how a single prevalent governance mechanism interacts with “strong board” mechanisms to determine the systemic risk contribution by banks. To the best of our knowledge, this is the first study to examine the bundling effect of bank governance mechanisms on systemic risk. Further, this study responds to the call of Schiehll et al. (2014) to further our understanding regarding the effect of corporate governance on organizational outcomes in the context of national/regional governance characteristics. By investigating the interactive effects of institutional control, which characterizes the ownership structure of European organizations, this study offers a relevant response to the call of Schiehll et al. (2014). Lastly, we acknowledge the recent agenda/calls towards a stakeholder approach to bank governance (BCBS, 2015, Schwarcz, 2017). Thus, although we build our arguments from the shareholder perspective, we appraise our findings and implications in light of the stakeholder perspective to inform how systemic risk could be managed/curtailed, especially from a practitioner and regulatory perspective.

The remainder of the paper is organized as follows. Section 2 reviews the relevant literature and states the empirical hypotheses. Section 3 describes the data, variables, and the empirical methodology, while Section 4 presents the results and discussion of our empirical tests. Section 5 concludes by discussing various implications of the findings and offering directions for future research.