The Timeliness of Restatement Disclosures and Financial Reporting Credibility

• Author: Kevin R. Smith,Mark Hirschey,Wendy M. Wilson
• Date: 01 September 2015
• Published date: 01 September 2015
• DOI: http://doi.org/10.1111/jbfa.12125

Journal of Business Finance & Accounting

Journal of Business Finance & Accounting, 42(7) & (8), 826–859, September/October 2015, 0306-686X

doi: 10.1111/jbfa.12125

The Timeliness of Restatement Disclosures

and Financial Reporting Credibility

MARK HIRSCHEY,KEVIN R. SMITH AND WENDY M. WILSON∗

Abstract: This study investigates whether prompt discovery and disclosure of earnings restate-

ments is associated with greater post-restatement financial reporting credibility. We measure

the timeliness of restatement detection by the length of time between the end of the misstated

period and the subsequent restatement announcement. We document that shorter detection

periods are significantly associated with high-quality corporate governance characteristics and

executive and/or auditor turnover, but not with characteristics of restatements. We also find

that firms with shorter detection periods exhibit a more moderate decline in the information

content of earnings following restatement announcements relative to firms with longer detec-

tion periods, and that detection period length has an incremental effect on the information

content of earnings relative to executive and/or auditor turnover alone. In addition, we find that

restatement disclosures are more timely following the implementation of the SOX-era reforms,

and that only firms with shorter detection periods experience more moderate post-restatement

declines in the information content of earnings following the implementation of the SOX-era

reforms. The results from this study suggest that the timeliness of restatement detection and

disclosure is associated with greater financial reporting credibility following restatements.

Keywords: accounting restatements, reporting credibility, misstatement detection, restatement

disclosures, corporate governance, information content of earnings

1. INTRODUCTION

This study examines whether the prompt detection and disclosure of earnings

restatements is associated with greater post-restatement financial reporting credi-

bility. Restatements are typically regarded as bad news events that often result in

a negative stock price reaction upon announcement (Palmrose et al., 2004) and

ongoing concerns about reporting credibility (Wilson, 2008; and Chen et al., 2014).

However, firms with higher quality corporate governance over the financial reporting

∗The first author was the Anderson W. Chandler Professor of Business at the University of Kansas. The

second author is at Utah Valley University. The third author is at Texas Christian University. The authors

thank Susan Scholz, Michael Ettredge, Bruce Johnson, Gerry Lobo, Andrew Leone, Zining Li, Ray Pfeiffer,

and seminar participants at the 2011 European Accounting Association Annual Meeting, the 2012 American

Accounting Association Annual Meeting, The University of Minnesota, Rice University and Texas Christian

University for helpful comments. We also appreciate the helpful suggestions provided by an anonymous

reviewer and the editor. The authors dedicate this paper in the memory of our co-author and friend Mark

Hirschey.

Address for correspondence: Wendy Wilson, TexasChristian University, Fort Worth, TX, USA.

e-mail: wendy.wilson@tcu.edu

C

2015 John Wiley & Sons Ltd 826

THE TIMELINESS OF RESTATEMENT DISCLOSURES 827

process are likely able to detect and disclose accounting issues sooner, which may

help to restore investor confidence in the firm’s operations and financial reports

following a restatement (Johnson, 2008). Regulators recognize the importance of

timely disclosures. For example, the Securities and Exchange Commission (SEC)

requires firms to announce restatements soon after management’s non-reliance

judgment because ‘timely disclosure of earnings restatements should raise investors’

expectations regarding the information that reporting companies are required to

make available to the markets’ (SEC, 2004).

We hypothesize that firms with better corporate governance over the financial

reporting process are able to detect and disclose accounting irregularities more

promptly, which, in turn, affects investors’ perceptions of post-restatement reporting

credibility. Prior research provides evidence that market reactions to post-restatement

earnings announcements are dampened, likely due to investors’ uncertainty regarding

whether earnings reports issued after accounting irregularities provide credible signals

of expected future cash flows (Wilson, 2008; and Chen et al., 2014). Prior research

also suggests that firms with better governance remediate material internal control

weaknesses (Goh, 2009) and provide details about the financial effects of earnings

restatements on a more timely basis (Schmidt and Wilkins, 2013). Building on these

results, we examine whether higher quality governance is associated with more timely

detection and disclosure of earnings restatements, and consequently whether more

timely disclosures result in relatively higher investor perceptions of post-restatement

reporting credibility.

We first examine the relation between detection period length and firm, re-

statement, and corporate governance characteristics (Goh, 2009; Myers et al., 2013;

Schmidt and Wilkins, 2013; and Wang, 2013). We measure the length of restatement

detection periods for a sample of earnings restatements caused by accounting irregu-

larities, and define the restatement detection period as the number of days between

the end of the misstated period and the restatement announcement date.1The results

indicate a significant relation between higher quality governance and the speed by

which accounting irregularities are discovered. Specifically, firms with Big-N auditors,

an independent board of directors, greater analyst following, and that replace the

CEO, CFO, or auditor following restatement announcements have shorter detection

periods. On the other hand, we do not find evidence that restatement characteristics,

such as the magnitude or the sign of the restatement’s effect on previously reported

earnings, are significantly associated with detection period length.

Next, we examine cross-sectional differences in the information content of sub-

sequently reported earnings as measured by the short-window earnings response

coefficient (ERC). In our full sample analysis, we find that firms with shorter

restatement detection periods have a more moderate post-restatement decline in the

information content of earnings. In particular, the ERC is significantly lower for six

quarters following restatements with prompt disclosure, relative to a lower ERC for

1 We use a sample of irregularity-related restatements because it allows us to test whether there is a decline

in investor apprehension for a group of firms that are subject to credibility concerns. The use of this sample

also allows the comparison of our results with evidence from prior research (Chakravarthy et al., 2014; and

Chen et al., 2014). In addition, our sample holds the severity of the sample restatements constant, which

is important because the number of restatement announcements increased while the proportion of severe

restatements (relative to the total number of restatements) decreased over the sample period (Scholz, 2008;

and Plumlee and Yohn, 2010). Wediscuss robust inferences from tests conducted with an expanded sample

later.

C

2015 John Wiley & Sons Ltd

828 HIRSCHEY, SMITH AND WILSON

11 quarters following restatements with longer detection periods. We also find that

the difference-in-differences in the ERC for firms with shorter versus longer detection

periods is significantly positive for the seventh through the twelfth quarters following

the restatement. This evidence suggests that investors have more moderate ongoing

credibility concerns for firms with more timely restatement detection due to the

association between shorter detection periods and higher quality governance over

financial reporting.

As prior research finds evidence that replacement of the firm’s CEO, CFO

and/or auditors following accounting irregularities is interpreted by the market

as a reputation-building action (Chakravarthy et al., 2014), we use a difference-in-

differences research design to test whether the association between turnover and

detection period length captures a similar, or distinct, effect on the post-restatement

information content of earnings (Wilson, 2008; and Chen et al., 2014). We find

that firms with short restatement detection periods and executive and/or auditor

turnover have significantly greater ERCs between the fourth through the eighth

post-restatement quarters relative to firms that either do not experience turnover or

that have long detection periods. This evidence is consistent with the idea that the

speed by which managers and/or auditors discover and disclose restatements has an

incremental effect on investors’ assessments of post-restatement reporting credibility,

relative to executive and/or auditor turnover alone.

Finally, we analyze whether new corporate governance regulations designed to

increase the quality of auditing and financial reporting affect restatement detection

and the post-restatement information content of earnings. In 2002, the ‘SOX-era

reforms’ were enacted, which includes the Sarbanes-Oxley Act (SOX), Statement

on Auditing Standards (SAS) No. 99 (AICPA, 2002), and more stringent gover-

nance requirements for firms listed on the NYSE and NASDAQ stock exchanges.

We report evidence that the mean and median length of restatement detection

periods decreased after 2002, which is consistent with evidence from prior research

indicating increased auditor involvement and fraud detection following the SOX-era

reforms (Carpenter, 2007; and Dyck et al., 2010). We find that firms that announced

restatements prior to the regulatory changes exhibit a similar and long-lasting de-

cline in the information content of earnings, regardless of the length of detection

periods. However, in the post-reforms period, firms with shorter detection periods

have significantly higher ERCs for the fourth through the twelfth quarter following

restatements, compared to the ERCs of firms with longer restatement detection

periods.2This result is consistent with more subdued investor concern about reporting

credibility in the post-SOX era for firms that discover and disclose accounting

irregularities quickly.

This paper makes several contributions to the earnings restatements literature. First,

the results indicate a significant relation between higher quality corporate governance

and more timely restatement disclosures, which suggests that better monitoring

over the financial reporting process helps firms detect and disclose accounting

2 Given the number of changes in the regulatory environment during 2002, it is not possible to disentangle

the effects of specific SOX initiatives, guidance provided by SAS No. 99, and changes in NYSE and NASDAQ

corporate governance requirements (Leuz, 2007; Leuz and Wysocki, 2008; and Coates and Srinivasan, 2014).

Our SOX-era reforms time-series tests incorporate a pre/post research design centered on July 30, 2002,

the date that SOX was enacted. However, we report robust results when the research design incorporates

different regulatory enactment dates during the sample period.

C

2015 John Wiley & Sons Ltd