Original Articles: 2018 Vol: 22 Issue: 3
Nana Yamfo Amoah, Rollins College
Isaac Bonaparte, Towson University
Muniratu Kelly, Morgan State University
Bilal Makkawi, Morgan State University
This study investigates whether accounting irregularity and improper revenue recognition are associated with the probability that an auditor will be sued for a defective audit of a client. We focus on accounting irregularity and improper revenue recognition given that they are accounting misstatement characteristics that could have severe valuation consequences on shareholder wealth and signal intentional misstatement. Using litigation data from the Stanford Securities Class Action Clearinghouse database and logistic regressions, we find that external auditors are more likely to be sued when audit clients improperly recognize revenues and the accounting misstatement is due to an accounting irregularity. We also find that external auditors are more likely to be sued by shareholders when the Securities and Exchange Commission (SEC) initiates an enforcement action through a court filing or an administrative proceeding. Our results suggest that revenue misstatement, accounting irregularity and SEC enforcement actions appear to be indicators of audit failure. Our results also imply that auditors could reduce the risk of litigation by averting accounting irregularity and improper revenue recognition in audited financial statements. The results should be of interest to regulators and other stakeholders who are concerned with improving audit quality.
Accounting Irregularity, Revenue Recognition, Audit Failure, Auditor Litigation.
Auditors, by the nature of their fiduciary relationship with investors, are required to design their audits to provide reasonable assurance that financial statements are not materially misstated. This duty of care of the auditor notwithstanding, many instances of audit failure have been recorded, which have resulted in huge losses in investor capital and lowered investor confidence in the capital markets (Chaney and Philipich, 2002; Lys and Watts, 1994; Stice 1991). Extant literature documents a high incidence of lawsuits against auditors, some of which have resulted in substantial auditor settlements (Choi and Pritchard, 2012; Fafatas, 2010; Ramseyer and Rasmusen, 2013; Stice, 1991; Public Accounting Report, 1985)1.
In this study, we examine whether improper revenue recognition and accounting irregularity are associated with the probability that an auditor will be sued for a defective audit of a client. We focus on improper revenue recognition and accounting irregularity as they are accounting misstatement characteristics that could have adverse valuation consequences on shareholder wealth and signal intentional misstatement. Revenues provide critical information to market participants in their decision-making process (Ou and Penman, 1989; Jegadeesh and Livnat, 2006; Stubben, 2010) and improper revenue recognition could trigger severe valuation consequences on the shareholders of an audit client (Hennes et al., 2008). Moreover, improper revenue recognition and accounting irregularity increase the likelihood that the financial statements of an audit client might have been intentionally misstated (Palmrose and Scholz, 2004). Accordingly, we examine whether improper revenue recognition and accounting irregularity are considered by shareholders as an indication of audit failure, thereby triggering a lawsuit against the auditor.
We use logistic regressions and litigation data from the Stanford Securities Class Action Clearinghouse database over the 10-year period after the Private Securities Litigation Reform Act of 1995 (PSLRA), that is, 1996 to 2005, to examine whether accounting irregularity and improper revenue recognition are associated with the probability of auditor litigation. We find that auditors are more likely to be sued when accounting misstatements are due to irregularities, and revenues are improperly recognized. Our study provides insights into the legal liability of external auditors given that:
1. Foster et al. (2007) argue that the PSLRA has not curtailed frivolous lawsuits against auditors.
2. Choi, Nelson, and Pritchard, (2009) and Pritchard (1999) note that lawsuits are settled for nuisance amounts after the PSLRA.
3. Many firms are unable to properly recognize revenues (Bloom and Schirm, 2001).
Given that misstatements due to accounting irregularity and improper revenue recognition could be perceived as sustaining a high inference of fraud, our finding that accounting irregularity and improper revenue recognition are associated with the probability of auditor litigation implies that shareholder lawsuits against auditors appear to have merit. Accordingly, our study contributes to the literature on the merits of auditor litigation in the decade after the passage of the PSLRA. Our paper also contributes to the audit quality literature by highlighting accounting misstatement characteristics that are considered by shareholders as indicators of audit failure. Finally, our results imply that auditors could reduce the risk of litigation by averting accounting irregularity and improper revenue recognition in audited financial statements.
The rest of the study proceeds as follows: Section 2 provides the literature review and hypothesis development. Section 3 presents the methodology and Section 4 reports the empirical results of the study. The conclusion follows in Section 5.
Opportunistic behavior of managers is a major concern to researchers, investors, and regulators (Watts and Zimmerman, 1986). To help mitigate information asymmetry and managerial opportunistic behavior, firms hire independent auditors. Audit failures are more likely to occur when auditors do not exhibit competence, integrity, and objectivity in their duties2. Not only could the auditor be sued by shareholders for an alleged defective audit, but the Securities and Exchange Commission (SEC) and the Public Company Accounting Oversight Board (PCAOB) may also conduct investigations3 and issue appropriate sanctions (Chaney and Philipich, 2002)4. In a keynote address to participants attending the American Law Institute Conference on Accountants’ Liability in Washington, D.C., on September 22, 2016, Andrew Ceresney, Director of Enforcement at the SEC noted that the Division of Enforcement conducts investigations into issues pertaining to audit failures and auditor independence violations.
Fuerman (2012) reports a decline in auditor litigation in the post-SOX period, and according to the study, the decline in auditor litigation may be attributable to the strengthening of corporate governance mechanisms following the passage of SOX5. Notwithstanding the reported decline in auditor litigation post SOX, auditor culpability in matters pertaining to improper revenue recognition and accounting irregularity are likely to result in reputational and monetary penalties on the audit firm (Armstrong et al., 2010).
We first establish the likely relation between accounting irregularity and the propensity of the auditor to be sued in a securities class action. Investors react more negatively to misstatements that are due to accounting irregularity, and the SEC typically initiates investigations into misstatements due to accounting irregularity. Given the seriousness of an accounting irregularity, shareholders are more likely to allege that the financial statements were not properly audited if the auditor issued an unqualified opinion. Accordingly, we expect that the likelihood of a lawsuit against the auditor will be greater when misstatements are due to an accounting irregularity. We state our first hypothesis as follows:
H1: Ceteris paribus, there is a positive relation between accounting irregularity and probability of auditor litigation.
We next examine the relation between improper revenue recognition and auditor litigation. One of the key factors that impacts the quality of earnings is the reliability of the revenues recognition (Altamuro et al., 2005; Stubben, 2010). Improper revenue recognition is viewed negatively by market participants (Anderson and Yohn, 2002) and could be considered as an indicator of audit failure. Given that revenue provides critical information to market participants in their decision-making process (Ou and Penman, 1989; Jegadeesh and Livnat, 2006; Stubben, 2010) and investors may consider the audit as being defective, we expect that the likelihood of a lawsuit against the auditor will be greater when the audit client improperly recognized revenues. We thus state our second hypothesis as follows:
H2 Ceteris paribus, there is a positive relation between improper revenue recognition and auditor litigation.
Sample and Data Collection
We extract data on securities class action lawsuits from the Stanford Securities Class Action Clearinghouse database for the period 1996 to 2005. We supplement our data with lawsuit data from the Lexis Nexis database in line with Karpoff et al. (2013). We also obtain firm-specific data from the Compustat database. Our final sample consists of 148 litigation firms.
Hypotheses Testing Model
We examine the relation between the test variables (accounting irregularity and improper revenue recognition) and probability of auditor litigation using the logistic regression model stated in its general form as:
Auditor_Litigtn=f(Irreglar,SECINV, Imprevrec, Abninstrad, EquityIss, BTM, Levrge, LnTAssets).
The dependent variable is Auditor_Litigtn (Auditor Litigation) and it is equal to 1 if the Auditor is sued, and 0, otherwise. The test variables are Irreglar (Accounting Irregularity), and Imprevrec (Improper Revenue Recognition). The Irreglar (Irregularity) variable is a binary variable which is equal to 1 if the lawsuit against the auditor is triggered by an accounting irregularity, 0, otherwise. The Imprevrec (Improper Revenue Recognition) variable is a binary variable which is equal to 1 if the auditor litigation is triggered by improper revenue recognition, 0, otherwise.
Our control variables are: Abninstrad (Abnormal Insider Trading), EquityIss (Equity Issue), BTM (ratio of book value of equity to market value of equity), Levrge (Leverage), and LnTAssets (log of total assets). Abinstrad (Abnormal Insider Trading) variable is a binary variable which is equal to 1 if the lawsuit against the audit client alleges abnormal insider trading, and 0, otherwise. EquityIss (Equity Issue) is a binary variable which takes on the value 1 if the basis of the allegation is that the issuance of equity facilitated the opportunistic manipulation of financial statements, and, 0, otherwise. Other control variables are firm growth depicted by BTM (ratio of the book value of equity to the market value of equity), Levrge (Leverage), which is the ratio of total liabilities (TTL) to total assets (TTA), and the size of the audit client depicted by the log of the total assets of the audit client, LnTAssets (Log of Total Assets). Hall and Renner (1988) document that the auditors of growth firms are more likely to be sued when the internal controls of the growth firms are unable to cope with the growth rate of the audit client. We thus expect a positive relation between BTM (firm growth) and probability of auditor litigation. Using leverage as a proxy for financial distress, Lys and Watts (1994) document a positive association between probability of auditor litigation and leverage. Accordingly, we expect a positive relation between Levrge (Leverage) and probability of auditor litigation. We also expect a positive association between LnTAssets (Log of Total Assets) and Auditor_Litigtn (Auditor Litigation) based on Lys and Watts (1994) who report a positive relation between the size of the audit client and probability of auditor litigation.
Finally, we include in our model SECINV (SEC Investigation), which is a binary variable equal to 1 if the accounting misstatement triggers an investigation by the SEC, 0, otherwise. To avert multicollinearity issues due to the high correlation between Irreglar (Irregularity) and SECINV (SEC investigation), we do not introduce both variables at the same time in the model.
Panel A of Table 1 presents descriptive statistics of the continuous variables. We find that the mean BTM (ratio of the book value of equity to the market value of equity) of the firms contained in the sample is 0.0566 and the median is 0.0011. We also find that the mean levrge (leverage) is 0.5768 while the median is 0.5953. We also document that the mean of the size of the firms, LnTAssets (log of the total assets), is 7.8853 while the median is 7.7507.
Table 1 Descriptive Statistics |
||||
Panel A: Continuous Variables (N=148) | ||||
Variable | Mean | Median | Std. dev. | |
BTM | 0.0566 | 0.0011 | 0.1631 | |
Levrge | 0.5768 | 0.5953 | 0.2479 | |
LnTAssets | 7.8853 | 7.7507 | 2.0237 | |
Panel B: Binary Variables (N=148) | ||||
Number of firms | Percentage | Sample Size | ||
N | ||||
Auditor_litign | 28 | 18.92% | 148 | |
Abninstrad | 66 | 44.59% | 148 | |
Imprevrec | 65 | 43.92% | 148 | |
EquityIss | 59 | 39.86% | 148 | |
Irreglar | 61 | 41.22% | 148 | |
SECINV | 30 | 20.27% | 148 |
Panel B of Table 1 reports the descriptive statistics for the binary variables. We find that 18.92% of the audit firms were sued. Client firms that were sued for abnormal insider trading are 44.59% while 43.92% of client firms were sued for improperly recognizing revenues. We infer also from the analysis that 39.86% of lawsuits were related to equity issues, 41.22% of the lawsuits were related to accounting irregularity, and 20.27% of the misstatements triggered investigations by the SEC.
Table 1 presents the descriptive statistics of the continuous and binary variables. Panel A and Panel B report the descriptive statistics of the continuous and binary variables, respectively. BTM is the ratio of the book value of equity to the market value of equity. Levrge is the ratio of total liabilities (TTL) to total assets (TTA). LnTAssets is the log of total assets. Auditor_Litigtn is a binary variable which is equal to 1, if the audit firm is sued, 0, otherwise. Abninstrad is a binary variable equal to 1 if the suit against the audit client alleges abnormal insider trading, 0, otherwise. Imprevrec is a binary variable equal to 1 if the suit against the audit client is based on the improper recognition of revenue, 0, otherwise. EquityIss is a binary variable equal to 1 if the suit alleges that the misstatement was precipitated by executive equity incentives, 0, otherwise. Irreglar is a binary variable equal to 1, if the lawsuit against the audit client is based on alleged accounting irregularity, 0, otherwise. SECINV is a binary variable equal to 1 if the misstatement triggered SEC investigation, 0, otherwise.
Panel A of Table 2 provides information on the number of securities lawsuits against audit clients over the sample period 1996-2005. We document from the distribution that the highest number of class action lawsuits (28) against audit clients occurred in fiscal year 2002 representing 18.92% of the sued firms. The second and the third highest number of lawsuits occurred in 1997 and 1999 fiscal years, respectively, with 19 firms representing 12.84% of the sample and 17 firms representing 11.49% of the sample firms. The table also reveals that the lowest number of lawsuits (6) was recorded in 1996 representing 4.05% of the sample firms.
Table 2 Distribution Of The Event Year And Industry Classification For Shareholder Litigation |
|||||||||||||
Panel A. Distribution of shareholder litigation by fiscal year (N=148) | |||||||||||||
1996 | 1997 | 1998 | 1999 | 2000 | 2001 | 2002 | 2003 | 2004 | 2005 | Total | |||
Securities Lawsuits | 6 | 19 | 15 | 17 | 15 | 16 | 28 | 13 | 11 | 8 | 148 | ||
Panel B: Industry Classification of Sample Firms (N=148) | |||||||||||||
Industry | Two-Digit SIC Code |
Number of Firms |
Percentage | ||||||||||
Metal Mining | 10 | 2 | 1.35 | ||||||||||
Oil and Gas Extraction | 13 | 2 | 1.35 | ||||||||||
Bldg Cnstr-Gen Contr,Op Bldr | 15 | 1 | 0.68 | ||||||||||
Heavy Construction, Except Building | 16 | 1 | 0.68 | ||||||||||
Food Products | 20 | 4 | 2.70 | ||||||||||
Textile Mill Products | 22 | 1 | 0.68 | ||||||||||
Apparel and other Textile Products | 23 | 4 | 2.70 | ||||||||||
Chemical and Allied Products | 28 | 11 | 7.43 | ||||||||||
Stone, Clay, and Glass Products | 32 | 1 | 0.68 | ||||||||||
Primary Metal | 33 | 2 | 1.35 | ||||||||||
Industrial, Communication, Machinery and Computer Equipment | 35 | 10 | 6.76 | ||||||||||
Electronic and Other Electric Equipment | 36 | 13 | 8.78 | ||||||||||
Transportation Equipment | 37 | 3 | 2.03 | ||||||||||
Instruments and Related Products | 38 | 4 | 2.70 | ||||||||||
Miscellaneous Manufacturing | 39 | 1 | 0.68 | ||||||||||
Railroad Transportation | 40 | 1 | 0.68 | ||||||||||
Water Transportation | 44 | 1 | 0.68 | ||||||||||
Transportation by Air | 45 | 1 | 0.68 | ||||||||||
Communications | 48 | 6 | 4.05 | ||||||||||
Electric, Gas and Sanitary Services | 49 | 15 | 10.14 | ||||||||||
Wholesale ? Durable Goods | 50 | 2 | 1.35 | ||||||||||
Wholesale ? Nondurable Goods | 51 | 3 | 2.03 | ||||||||||
General Merchandise Stores | 53 | 2 | 1.35 | ||||||||||
Food Stores | 54 | 1 | 0.68 | ||||||||||
Apparel and Accessory Stores | 56 | 1 | 0.68 | ||||||||||
Furniture and Home Furnishings Stores | 57 | 1 | 0.68 | ||||||||||
Miscellaneous Retail | 59 | 6 | 4.05 | ||||||||||
Depository Institutions | 60 | 6 | 4.05 | ||||||||||
Nondepository Credit Institution | 61 | 1 | 0.68 | ||||||||||
Security and Commodity Brokers | 62 | 1 | 0.68 | ||||||||||
Insurance Carriers | 63 | 6 | 4.05 | ||||||||||
Insurance Agents, Brokers and Service | 64 | 2 | 1.35 | ||||||||||
Holding, Other Invest Offices | 67 | 1 | 0.68 | ||||||||||
Personal Services | 72 | 1 | 0.68 | ||||||||||
Business Services | 73 | 19 | 12.84 | ||||||||||
Motion Pictures | 78 | 1 | 0.68 | ||||||||||
Amusement and Recreation Services | 79 | 1 | 0.68 | ||||||||||
Health Services | 80 | 3 | 2.03 | ||||||||||
Educational Services | 82 | 1 | 0.68 | ||||||||||
Engr,Acc,Resh,Mgmt,Related Services | 87 | 2 | 1.35 | ||||||||||
Nonclassifiable Establishment | 99 | 3 | 2.03 | ||||||||||
Total | - | 148 | 100 |
Panel B of Table 2 presents the securities class actions categorized by industry groupings. We use the 2-digit SIC code to categorize the firms into the designated industries. We infer from the table that the Business Services sector (2-digit SIC Code=73) had the most lawsuits during the sample period. We document that 19 (12.84%) firms within the Business Services sector were sued. The Electric, Gas, and Sanitary Services sector (2-digit SIC Code=49) followed with 15 (10.14%) sued firms. The Electronic and other Electric Equipment sector (2-digit SIC Code=36) ranked third with 13 (8.78%) sued firms.
Table 2 presents the distribution of the securities lawsuits across the sample period and the industry distribution of the sample firms. Panel A presents the distribution of the lawsuits across the sample period (1996-2005) and Panel B presents the industry distribution of the sample firms.
In Table 3, we provide results of the logistic regressions of the probability of auditor litigation. In model 1, we exclude SECINV (SEC investigation) from the regression. In Model 2, SECINV (SEC investigation) enters the model in place of Irreglar (irregularity). This procedure is employed due to the high correlation between the Irreglar (irregularity) variable and the SECINV (SEC investigation) variable. In both models (1 and 2), we use the same control variables: Abninstrad (Abnormal Insider Trading), EquityIss (Equity Issue), BTM (Ratio of book value of equity to market value of equity), Levrge (Leverage), and LnTAssets (log of total assets). Consistent with our first hypothesis, we find a significantly positive relation between accounting irregularity and probability of auditor litigation. The variable irreglar (irregularity) is positive and significant at the 1% level (coefficient: 1.4979; p<0.001). Also in model 1, we document a positive relation between the variable Imprevrec (Improper revenue recognition) and probability of auditor litigation. Imprevrec (Improper revenue recognition) is positive and significant at the 10% level (coefficient: 0.9913; p<0.010) suggesting that the auditor is more likely to be sued when the audit client improperly recognizes revenues. The positive relation between improper revenue recognition and probability of auditor litigation is consistent with our second hypothesis. In Model 2 of Table 3, SECINV (SEC Investigation) is positive and significant at the 5% level (coefficient: 1.1310; p<0.05). The significantly positive relation (coefficient: 1.1310; p<0.05) between SECINV (SEC Investigation) and probability of auditor litigation implies that SEC investigation may be perceived by shareholders as a credible signal of audit failure, which in turn increases the probability of auditor litigation. In Model 2, we also document a positive and significant relation between the Imprevrec (Improper revenue recognition) and probability of auditor litigation. Imprevrec (Improper revenue recognition) is positive and significant at the 5% level (coefficient: 1.1206; p<0.05), which is consistent with hypothesis H2.
Table 3 Logistic Regression Of Probability Of Auditor Litigation |
|||
Variable | Pred Sign |
1 | 2 |
Intercept | + | -5.1054a (15.72) |
-4.1585a (11.21) |
Irreglar | + | 1.4979a (8.60) |
- |
SECINV | + | - | 1.1310b (4.57) |
Imprevrec | + | 0.9913c (3.81) |
1.1206b (5.29) |
Abninstrad | + | 0.2741 (0.30) |
0.1966 (0.16) |
EquityIss | ? | 0.5511 (1.22) |
0.4115 (0.73) |
BTM | + | -0.6062 (0.17) |
-0.1820 (0.01) |
Levrge | + | 2.0877 (2.5711) |
1.8745 (2.29) |
LnTAssets | + | 0.0810 (0.2862) |
0.0444 (0.0866) |
Model ChiSq (P-value) N |
NA | 25.51a (<0.01) 148 |
20.56a (<0.01) 148 |
Table 3 presents logistic regressions of the probability of auditor litigation and other variables over the period 1996-2005. Irreglar equals 1 if the shareholder lawsuit against the audit client is based on alleged accounting irregularity, 0, otherwise. SECINV is a binary variable equal to 1 if the misstatement triggered SEC investigation, 0, otherwise. Imprevrec equals 1 if the lawsuit against the audit client is based on improper revenue recognition, 0, otherwise. Abninstrad is a binary variable equal to 1 if the suit against the audit client alleges abnormal insider trading, 0, otherwise. EquityIss is equal to 1 if the lawsuit against the audit client alleges that the misstatement was related to executive share-based compensation, 0, otherwise. BTM is the ratio of the book value of equity to the market value of equity, Levrge is the ratio of total liabilities to Total Assets, and the LnTAssets is the log of the Total Assets. Predictions for the signs of the logit coefficients are presented in parentheses next to the variable name. Chi-square test statistics are in parentheses below the logit coefficients for each variable. Statistical significance at the 1%, 5% and 10% levels is denoted by (respectively) a, b and c. Likelihood ratio test statistics for all the models correspond to p-values less than 0.01.
We examine the likelihood that an auditor will be sued in a securities class action lawsuit initiated by shareholders against an audit client and ascertain the probability that the accounting misstatement characteristics that triggered the lawsuit will result in an action being brought against the auditor. We find a higher propensity that the auditor will be sued when the accounting misstatement is triggered by an accounting irregularity. We also find that the auditor is more likely to be sued when revenues are improperly recognized by the audit client. Our results support initiatives by the Financial Accounting Standards Board (FASB) and the Securities and Exchange Commission to improve revenue recognition. The FASB’s new revenue recognition standard, effective from December 15, 2017, is the result of a joint effort with the International Accounting Standards Board (IASB) and it addresses revenue recognition issues that have adversely impacted the capital markets and shareholder wealth. Under the new revenue recognition rules, reporting firms are required to disclose contract details, contract terms and conditions, progress reports, accounting judgments, and accounting estimates employed in recognizing revenues. The rationale for these disclosure requirements is to enhance users’ understanding of the nature of revenue transactions, the contract amount, the timing and any uncertainties regarding the revenues and the cash flows to enable them to make informed decisions.
Given that the effectiveness of a client firm’s corporate governance could impact audit quality, our results support the strengthening of corporate governance mechanisms such as audit committees to enhance audit quality. Our results also support initiatives by the PCAOB to improve audit quality such as the use of audit quality indicators to guide audit firms, data analytics and technologies in audits, and randomization of audit quality inspections. Randomization of audit quality inspections could result in the identification of more audit deficiencies and greater remediation of deficiencies. Our results highlight the importance of auditor compliance with quality control standards to reduce the risk of litigation. As an audit firm’s quality control standards are expected to provide reasonable assurance that audits are performed in conformity with auditing standards, greater compliance with quality control policies and procedures could limit audit failures and the risk of litigation. Our results also imply that auditors could reduce the risk of litigation by averting accounting irregularity and improper revenue recognition in audited financial statements. Taken together, our findings support increased efforts to enhance revenue recognition and audit quality and should be of interest to shareholders, the SEC and PCAOB, auditors and other stakeholders who are concerned with deterring fraud and improving earnings quality and audit quality.
1. In 2005, KPMG paid $22.5 million to settle the lawsuit against the audit firm by the Securities and Exchange Commission (SEC) in relation to its audit of XEROX. In 2008, Ernst and Young agreed to pay over $300 million to settle the class action lawsuit initiated by shareholders of Cendant Corporation in relation to defects in its audit of the company’s financial statements.
2. In a keynote address to the American Law Institute Conference on Accountants’ Liability on September 22, 2016, Andrew Ceresney, Director, Division of Enforcement at the Securities and Exchange Commission noted that auditors are critical gatekeepers in ensuring that managers issue credible and reliable financial statements.
3. The Enforcement Division of the Securities and Exchange Commission is tasked with the responsibility of conducting investigations to ascertain the likelihood of infractions relating to federal securities laws.
4. Restatement of Fannie Mae’s 2001 to 2004 financial statements reduced earnings by about $6.3 billion. Following the restatement, KPMG and Fannie Mae were sued by investors for issuing misleading financial statements. In 2013, Fannie Mae and KPMG settled the class action for $153 million.
5. Elaine Hardwood and Laura Simmons of Cornerstone Research attribute the decline in auditor litigation risk in the post-SOX period to the corporate governance provisions of SOX.