Auditor Conservatism and Banks' Measurement Uncertainty during the Financial Crisis

• Author: Kevin Lam,Feng Chen,Wally Smieliauskas,Minlei Ye
• Date: 01 March 2016
• Published date: 01 March 2016
• DOI: http://doi.org/10.1111/ijau.12055

Auditor Conservatism and Banks’ Measurement Uncertainty during the

Financial Crisis

Feng Chen,1,2 Kevin Lam,3Wally Smieliauskas1and Minlei Ye1

1University of Toronto, Canada

2University of Missouri, USA

3Chinese University of Hong Kong

During the most recent financial crisis, the economic difficulties, along with potentially high uncertainties

associated with fair value estimates, increased the audit risks for bank auditors. We analyze a sample of US public

banks during the crisis (2008–09) and after the crisis (2010–11), and provide contrasting evidence concerning

auditors’ role in conservative financial reporting (proxiedby higher discretionary loan loss provisions). Specifically,

we document a significant positive association between discretionary loan loss provisions and the role of auditors

(proxied by higher amounts of audit fees and abnormal audit fees) during the crisis. However, the positive

association disappears following the financialcrisis. Similarly, we document that duringthe crisis, auditors are less

likely to issue unmodified audit opinions to banks that subsequently filed for bankruptcy. Thus, our evidence is

consistent with elevated auditor conservatism for the sample banks during the financial crisis. Our evidence also

suggests that bank auditors raise fees more to maintain accounting quality than to price protect from aggressive

accounting choices.

Key words: Auditor conservatism, discretionary loan loss provision, audit fees, asset measurement uncertainty

1. INTRODUCTION

This study investigates bank auditors’ influence on loan

loss provisions in light of higher perceived risks during

the recent financial crisis. We focus on the US banking

industry because banks were most adversely affected by

the financial crisis. In addition, we expect that a major

new source of risk for bank auditors would be the

extreme estimation uncertainty associated with fair value

estimates (e.g., Christensen, Glover & Wood, 2012).

The promulgation of FAS 157 (Fair Value

Measurements, now ASC 820; see FASB,2006) potentially

resulted in more volatility and discretion underlying fair

value measurements (Fiechter & Meyer, 2011; Riedl &

Serafeim, 2011; Milbradt, 2012). The financial crisis likely

exacerbated the estimation uncertainties related to

Level 3 fair valued investments, greatly increasing the

risks for auditors during the crisis period of 2008–09

(SEC, 2008; European Commission, 2010; PCAOB, 2010b).

Following the crisis, estimation uncertainties related to

Level 3 investments have decreased as a result of the

improvement in market conditions. Ettredge, Xu and Yi

(2014) focus on fair value estimates and audit fees, and

find that auditors charge more to audit banks with higher

fair value estimates. Our interest in this study is to

provide evidence about auditors’ role in conservative

financial reporting during and after the crisis.

Our conjecture is that if bank audit risks changed due

to changes in estimation uncertainty during the crisis,

then there should be an empirical relation between

discretionary loan loss provisions and audit fees (which

proxy for the role of auditors), and this relationship is

expected to shift over time during and after the financial

crisis. Thus, our analyses involve both the period during

the financial crisis (2008–09) and the period subsequent to

the financial crisis (2010–11).1

Few studies have investigated auditors’ responses to

the risk implications of the macroeconomic shocks

resulting from the financial crisis. Using the concept

of ‘auditor conservatism’ in the prior literature (e.g.,

Krishnan, 1994; DeFond & Subramanyam, 1998; Lu &

Sapra, 2009), we argue that auditors have incentives to

use more conservative financial reporting in at least some

estimates when they are exposed to greater risks in an

uncertain economic environment. During the financial

crisis, auditors maymitigate the risks by a combination of

increased audit effort, increased audit fees incorporating

the elevated litigation risk in their pricing, and increased

demand for more conservative accounting in at least

some accounting estimates (Watts, 2003; DeFond, Lim &

Zang, 2010). These qualitative issues take on increased

importance at a time of more reliance on accounting

estimates in financial reporting, and also highlight the

need to audit these estimates with high assurance

(European Commission, 2010, p. 6; IFAC, 2011a, p. 6;

IFAC, 2011b, p. 6).

In order to test for the conservative accounting

conjecture, our proxy for conservative financial reporting

is higher amounts of the discretionary component of loan

loss provisions (Kanagaretnam, Krishnan & Lobo, 2010).

Previous research has studied banks’ discretionary use of

loan loss provisions (DLLP) in earnings management

(e.g., Beaver & Engel, 1996; Ahmed, Takeda & Thomas,

1999; Beatty, Ke & Petroni, 2002) as loan loss provisions

are a major form of discretionary accruals in the banking

industry (Ryan, 2007). We first investigate the relation

between DLLP and Level 3 fair valueestimates during the

financial crisis. Most banks’ estimation uncertainties

were linked to banks’ securitization processes and the

increased use of Level 3 estimates in lieu of reliance on

the disorderly market prices during the crisis (Kothari &

Lester, 2012, p. 348). Level 3 fair value estimates are

based on unobservable inputs, giving rise to extreme,

unavoidable measurement uncertainties associated

with these fair value estimates in the period of crisis

(Christensen et al., 2012). Since higher litigation risk

Correspondence to: WallySmieliauskas, Rotman School of Management,

University of Toronto, Toronto, ON, M5S3E6, Canada. Email:

smieli@rotman.utoronto.ca

International Journal of Auditing doi:10.1111/ijau.12055

Int. J. Audit.

© 2015 John Wiley & Sons Ltd ISSN 1090-6738

20: 52 65 (2016)

–