Banking supervision and external auditors: Theory and empirics

https://doi.org/10.1016/j.jfs.2019.100722Get rights and content

Abstract

This paper investigates the role of external auditors in banking sector supervision from a theoretical, institutional and empirical perspective. We first present a simple principal-agent framework that highlights the importance of several institutional characteristics in determining the optimal involvement of external auditors in supervision. We then construct a new index that captures the degree of involvement of external auditors in the oversight of the banking sector in 115 countries. Consistent with our theoretical arguments, we find that countries that increase the role of central banks in supervision are also more likely to involve auditors, suggesting that the added complexity of a supervisory function is likely to benefit from the expertise of an external auditor. Having experienced a financial crisis is also associated with a higher use of auditors, particularly among central banks with an increasing role in supervision, which suggests some reputational concerns of the supervisor. Finally, we show that higher audit quality is associated with an increased involvement of auditors in supervision.

Introduction

In the aftermath of the 2008 Global financial crisis, researchers and policymakers alike have pointed to the weakness of banking supervisory frameworks as one of the leading causes of the crisis (see Merrouche and Nier, 2010; Kupiec et al., 2017). An effective supervision technology ought to detect, well in advance, potential threats to the safety and soudness of the banking sector. Previous research has explored various aspects of banking regulation that can achieve such objectives, including bank capital requirements, regulatory treatments of non-performing loans and provisions or disclosure requirements.1 However, little attention has been directed towards another important aspect of financial sector oversight, i.e. the use of external auditors in the implementation of specific banking supervisory tasks.

The involvement of auditors, as private financial gatekeepers, can improve the credibility of the overall supervisory setting, as auditors generally have a beneficial influence on the behaviour of regulated firms, or banks in this case.2 For this reason, numerous initiatives on the global regulatory framework (such as the Basel Committee on Banking Supervision, 2008, 2014) have recommended a tight relationship between banking supervisors and external auditors to enable an effective information exchange. The supervisor can request external auditors to perform different kinds of tasks, at times going beyond the standard audit report. However, the involvement of private actors in implementing public tasks can carry risks. For example, the supervisor can incur reputational costs given that auditors are private firms with potentially close ties to the regulated financial institutions. Similarly, the expected quality of the auditing reports defines the relative confidence with which supervisors can trust the information that auditors release.3

In this paper, we present a simple principal-agent framework that highlights the costs and benefits of involving external auditors in banking supervision. We show that the optimal involvement of external auditors depends on several institutional and country-specific factors. These include: (i) the expected benefits in terms of financial sector oversight brought on by the auditor, (ii) the quality of the auditing function and (iii) the perceived costs of auditor involvement, such as reputational concerns. We then test this hypothesis empirically by taking a positive approach and investigating whether the actual level of auditor involvement in a broad set of countries is related to the country-specific institutional characteristics underlined in our model.

To this end, we first develop a new measure that captures the level of Auditors’ Involvement in Supervision, which we call the AIS Index. This new index is based on nine institutional characteristics of external auditing in the banking sector across the world, as surveyed by the World Bank in 2007 and 2012. We employ the AIS Index to systematically assess the state of banking supervision in a large sample of 115 countries.

We find that countries generally adopt high levels of auditor involvement in supervision, scoring on average 7.7 out of a maximum of 9 points, in 2007. Nonetheless, this involvement has increased even further in 2012. Countries generally require that external auditors cooperate closely with the supervisor, however less stringent requirements are in place regarding the statute and qualification of the auditors.

We then provide a series of proxies for the costs and benefits of involving auditors in supervision highlighted in our theoretical framework and investigate their relative importance in explaining the increased involvement of auditors in banking supervision observed between the two World Bank surveys. As such, our baseline model is an ordered logit model that looks at the changes in the AIS index between 2007 and 2012. Our analysis highlights several key determinants of auditors’ involvement in supervision. First, we show that countries characterized by lower levels of auditor involvement in 2007 are more likely to increase it, confirming the tendency towards an ever-higher collaboration between supervisors and external auditors. Next, we find that countries that have increased the role of their central banks in supervision from 2007 to 2012 are also more likely to have a higher involvement of auditors in supervision, suggesting that the added complexity of a supervisory function is likely to benefit from the expertise of an external auditor.

We also investigate if having experienced a systemic banking crisis, as a proxy for higher reputational concerns of the supervisor, affects the likelihood of involving external auditors in supervision. We find that this is the case, but only in countries that also assigned more supervisory responsibility to their central bank, suggesting that the added complexity of a supervisory function is likely to benefit from the expertise of an external auditor. These results contribute to the literature that studies the role of financial crises in shaping supervisory and institutional settings (see Masciandaro et al., 2013, Masciandaro and Romelli, 2018 and Abascal and Gonzales, 2019, among others). Lastly, in line with our theoretical model, we also show that higher audit quality, proxied by a tight oversight of auditors’ actions, is associated with an increased involvement of auditors in banking sector supervision.

These results are robust to a series of sensitivity checks including various definitions of the AIS index, alternative empirical specifications and falsification strategies. To the best of our knowledge, this is the first paper to analyse the relationship between external auditing and banking supervision from a theoretical, institutional and empirical perspective. Our findings shed light on the similarities and differences in the supervisory setting around the world. Such knowledge can be particularly useful to assess the degree of convergence among national supervisory architectures in newly established communities of banking supervisors, as it is the case in the European Union (Masciandaro et al., 2011).

The remainder of the paper is structured as follows. Section 2 develops a theoretical framework to highlight the key factors that drive policymakers’ choices in involving auditors in banking supervisory activities. Section 3 proposes an institutional indicator of auditor involvement in supervision and presents some descriptive statistics. Section 4 presents the empirical strategy and results, while Section 5 concludes.

Section snippets

Optimal design of supervisory settings

In this section, we employ a simple principal-agent framework to study the optimal design of banking supervision involving external auditors. The supervisory technology is characterized by a mix of continuous monitoring and random auditing, where supervisors can delegate the auditing service to an independent audit agency in order to improve information quality and, at the same time, enforce supervisory settings.4

Institutional Framework

The theoretical framework presented in Section 2 predicts that external auditors’ involvement in supervision depends on well-identified structural country characteristics such as auditing quality, expected reputational benefits and expected costs. To test the empirical validity of these theoretical arguments, we first build a new measure that captures the degree of auditors’ involvement in supervision in a broad set of countries.

Empirics

The theoretical framework in Section 2 identified a series of institutional characteristics that might influence the choice of involving auditors in banking sector supervision. In this section, we test these predictions empirically by looking at the changes in the AIS index over time. Specifically, we assume that legislative reforms that require a higher level of auditor involvement in supervision (higher values of the AIS index) are related to the same set of factors that have been shown to

Conclusion

This paper studies the determinants of external auditors’ involvement in banking sector supervision from a theoretical and empirical perspective. First, in a simple principal-agent framework, we highlight the importance of several country-specific institutional characteristics in determining the optimal involvement of external auditors in supervision. The model suggests that one size does not fit all: each national policymaker must form expectations regarding the potential benefits and costs

Acknowledgements

We are particularly grateful to the Editor, Iftekhar Hasan, and the two anonymous referees for valuable suggestions. The usual disclaimer applies.

References (39)

  • D. Masciandaro et al.

    Exploring governance of the new European banking authority: a case for harmonization?

    J. Financ. Stab.

    (2011)
  • D. Masciandaro et al.

    The economic crisis: did supervisory architecture and governance matter?

    J. Financ. Stab.

    (2013)
  • D. Masciandaro et al.

    Central bankers as supervisors: do crises matter?

    Eur. J. Polit. Econ.

    (2018)
  • A. Abiad et al.

    Financial reform: what shakes it? What shapes it?

    Am. Econ. Rev.

    (2005)
  • A. Alesina et al.

    Bureaucrats or politicians? Part I: a single policy task

    Am. Econ. Rev.

    (2007)
  • K. Alvis

    Accounting credibility and liquidity constraints: evidence from reactions of small banks to monetary tightening

    Account. Rev.

    (2015)
  • J. Barth et al.

    Bank Regulation and Supervision in 180 Countries From 1999 to 2011, NBER Working Paper Series

    (2013)
  • Basel Committee on Banking Supervision

    External Audit Quality and Banking Supervision, Bank for International Settlements

    (2008)
  • Basel Committee on Banking Supervision

    External Audits of Banks, Bank for International Settlements

    (2014)
  • Cited by (0)

    View full text