اعتبار حسابرس و مدیریت سود: شواهدی بین المللی از صنعت بانکداری

We examine the relation between auditor reputation and earnings management in banks using a sample of banks from 29 countries. In particular, we examine the implications of two aspects of auditor reputation, auditor type and auditor industry specialization, for earnings management in banks. We find that both auditor type and auditor industry specialization moderate benchmark-beating (loss-avoidance and just-meeting-or-beating prior year’s earnings) behavior in banks. In addition, we find that once auditor type and auditor industry specialization are included in the same tests, only auditor industry specialization has a significant impact on constraining benchmark-beating behavior. In separate tests related to income-increasing abnormal loan loss provisions, we find that both auditor type and auditor expertise constrain income-increasing earnings management. Again, in joint tests, only auditor industry expertise has a significant impact on constraining income-increasing earnings management.

We examine the effect of auditor reputation on bank earnings management using an international sample of banks. Banks operate in a highly regulated environment in that they are monitored by Central Banks and other regulatory agencies (such as deposit insurance corporations). Consequently, auditor reputation may not be as important in constraining income-increasing earnings management in banks. On the contrary, if our finding establishes a negative association between auditor reputation and income-increasing earnings management, then auditor reputation likely is even more relevant for firms in other industries that are not subject to such direct regulatory scrutiny. To our knowledge no other study has examined how auditor reputation is related to earnings management in the international or the US banking industry. Our main prediction is that auditor reputation (auditor type and auditor industry specialization) is negatively related to earnings management in banks even after controlling for several previously identified international institutional factors and bank monitoring factors. Evidence of a negative relation between auditor reputation and earnings management may not be surprising for US banks because, in a high-litigation environment such as the US, high-reputation auditors have an incentive to maintain a high level of earnings quality to protect their reputation and legal exposure (Francis and Wang, 2008). Whether such a relation exists across different legal and institutional environments clearly is of interest. We are able to address this issue by analyzing an international sample of banks. Prior research in banking has examined the relation between international institutional factors and bank monitoring variables and earnings management (Shen and Chih, 2005 and Fonseca and González, 2008). Shen and Chih (2005) using earnings benchmark tests document that most banks manage their earnings. They also show that stronger investor protection and greater transparency in accounting disclosure reduce a bank’s incentives to manage earnings. Fonseca and González (2008) focus on factors influencing income smoothing through loan loss provisions, the major bank accrual. They find that income smoothing is lower in jurisdictions with greater bank regulation and supervision. Interestingly, neither of these papers addresses the impact of auditing, an important external monitoring mechanism, on earnings management or income smoothing. In the auditing literature, Becker et al. (1998) report lower earnings management in industrial firms for clients of Big 5 auditors. Krishnan (2003) finds that firms audited by industry specialists report lower discretionary accruals, a commonly used proxy for earnings management in industrial firms. And, Francis and Wang (2008), using an international sample, report that earnings quality is higher for firms that use Big 4 auditors, but their result holds only for regimes with strong investor protection. A notable deficiency of these studies is that they exclude firms in banking and financial services. In addition, auditing banks is more complex than auditing industrial firms. In its May 2006 report on large firm Public Company Accounting Oversight Board (PCAOB) inspection deficiency analysis, the American Institute of Certified Public Accountants’ (AICPA, 2006) Center for Public Company Audit Firms finds that banks’ loan loss allowance ranks number one among the various deficiencies found by inspectors. This indicates that auditing the loan loss allowance and the related loan loss provision are challenging tasks for auditors in general. High-reputation auditors have incentives to provide high-quality audits to avoid jeopardizing their reputation capital. Thus, auditor reputation is potentially important in assessing the adequacy of loan losses and mitigating earnings management incentives of bank managers. In this study, we extend the research on benefits of auditor reputation to the banking industry. Specifically, we examine two aspects of auditor reputation. First, we investigate the implications of auditor type (Big 5 vs. non-Big 5 auditors) for constraining income-increasing earnings management in banks. A large body of empirical research documents that higher audit quality is associated with Big 5 auditors. 1 Relative to non-Big 5 auditors, Big 5 auditors have greater expertise, resources, and more importantly, market-based incentives (e.g., mitigating the risk of litigation and protecting their reputation capital) to constrain the tendency of their audit clients to engage in aggressive reporting. Consequently, we predict that earnings management will be lower for banks audited by Big 5 auditors. Second, we examine the effect of auditor industry specialization on reducing earnings management in banks. Auditors who are specialists in the banking industry can better assess the adequacy of the loan loss provisions than non-specialist auditors. Prior research documents that auditor industry specialization enhances financial reporting quality and mitigates fraudulent financial reporting (Johnson et al., 1991, Carcello and Nagy, 2004, Krishnan, 2003 and Krishnan, 2005). We measure auditor industry specialization/expertise by an auditor’s industry market share.2 We employ three traditional proxies of earnings management, managing earnings to avoid losses, managing earnings to just-meet-or-beat the prior year’s earnings, and an accrual-based proxy (abnormal loan loss provisions), to test the extent of income-increasing earnings management through bank loan loss provisions. By using three different tests (accruals- and non-accruals-based tests), we strengthen the validity and robustness of our results. Our loss-avoidance/just-meeting-or-beating prior year’s earnings tests closely resemble the methodology used by Beatty et al., 2002 and Altamuro and Beatty, 2010. Our proxy for abnormal loan loss provisions is based on prior banking research on loan loss provisions (Wahlen, 1994 and Kanagaretnam et al., 2004). We use an international bank sample from the BankScope database representing 29 countries over the period 1993 to 2006 to test our hypotheses. We find in separate tests that both auditor type and auditor expertise moderate benchmark-beating (loss-avoidance and just-meeting-or-beating prior year’s earnings) behavior in banks. However, we find that once auditor type and industry expertise are included in the same tests, only auditor industry expertise has a significant impact on constraining benchmark-beating behavior. In separate tests related to income-increasing abnormal loan loss provisions, we find that both auditor type and auditor expertise constrain income-increasing earnings management. Again, in joint tests, only auditor industry expertise has a significant impact on constraining income-increasing earnings management. Overall we find that audit specialists constrain income-increasing earnings management in banks. Our results are robust to several sensitivity tests including alternate classification of audit specialists, controlling for self-selection, and different bin-widths for benchmark tests. The rest of this paper is organized as follows. The next section develops the hypotheses. Section 3 explains the empirical models used for tests of earnings management. Section 4 describes the sample selection process. Section 5 discusses the results and Section 6 concludes the study.

Given the importance of banking to national and global economies, there is surprisingly little evidence on the implications of auditing for banks’ earnings quality. Banks are very different from industrial firms, and given the recent, heightened concern with the quality of banks’ reported earnings following the meltdown in this sector, a study of the effect of auditor reputation on earnings management in the banking industry is of considerable interest to regulators and investors. We examine the relation between auditor reputation and earnings management in banks using a sample of banks from 29 countries. We hypothesize that high-reputation auditors will constrain income-increasing earnings management in banks. In particular, we examine the implications of two aspects of auditor reputation, auditor type (Big 5 vs. non-Big 5 auditors) and auditor industry specialization for earnings management in banks. We find in separate tests that both auditor type and auditor expertise constrain earnings benchmark-beating (loss-avoidance and just-meeting-or-beating prior year’s earnings) behavior in banks. However, we find that once auditor type and industry expertise are included in the same tests, only auditor industry expertise has a significant impact on constraining benchmark-beating behavior. In tests related to income-increasing abnormal loan loss provisions, we find in separate tests that both auditor type and auditor expertise constrain income-increasing earnings management. Again, in joint tests, only auditor industry expertise has a significant impact on constraining income-increasing earnings management. Our results show that even in a highly regulated industry such as banking, auditor reputation has an important role in constraining income-increasing earnings management. Moreover, in an international banking context, our study can be regarded as documenting an important external monitoring mechanism in addition to previously identified international institutional factors and bank monitoring factors that constrains earnings management in banks.