Using debt-related events as explanatory factor of audit failures
Lee J. Yao*
Joseph A. Butt, S.J. College of Business
Loyola University New Orleans
Siew H. Chan
School of Accounting
Washington State University
Pullman, Washington, USA
Charlton College of Business
University of Massachusetts at Dartmouth
Key words: audit independence, event study, debt-related events, going-concern; audit failure
* All correspondence should be communicated to the first author.
The recent collapse of Enron, the world’s largest energy-trading company has put the accounting profession under scrutiny. Enron highlights the possible failure of the accounting profession. Arthur Andersen’s role in the Enron case has hurt the profession’s credibility and the public’s trust in it. This debacle has triggered wide-ranging debates and has subsequently raised many issues for the accounting profession, most important of all, the issue of auditor independence.
Internationally, the issue of auditor independence has been featured prominently in the reviews by the International Federation of Accountants and the European Commission, as well as the U.S. Securities and Exchange Commission (SEC). The Asia Pacific Economic Community (APEC) Task Force on Company Accounting and Financial Reporting have also identified auditor independence as a pressing issue throughout APEC member economies [Ministry of Finance, Singapore, 2001].
The auditor’s report plays a critical role in warning market participants of impending going-concern problems. Indeed, the term audit failure typically refers to cases in which auditors fail to issue going-concern opinions to clients that subsequently file for bankruptcy [Blacconiere and DeFond, 1997]; [Weil, 2001].
Since independent auditing is an essential feature of efficient capital markets, regulators have long been concerned about the potential threats to auditor independence. The implications for such threats are far-reaching and could affect many players in the capital markets.
Auditors are charged with the responsibility of evaluating the likelihood of going-concern of their clients. Although the evaluation is not intended to be a predictor of bankruptcy, financial statement users often treat the evaluation as an early warning of impending failure. Since regulators and users tend to treat an unmodified audit opinion as a “clean bill of health”, they do not expect the business to fail in the near future [Casterella et al 2000]. Research has shown that more often than not, auditors end up letting users down when it comes to predicting bankruptcy filings with audit opinions [Casterella et al 2000].
In the wake of Enron’s bankruptcy, auditors are blamed for not doing their jobs well. Although auditors assert that they are not responsible for predicting future events, it is very clear that their opinion decisions are evaluated, at least in part, based on events that occur after the audit report date.
The interesting and logical next step is to find out why auditors fail in their roles. Does it implicate the issue of auditor independence? At the time of auditing, was there contrary information that the auditors chose to ignore or were there mitigating factors that indicated that the firms would survive? With this knowledge in mind, auditors could prevent similar incidents like Enron from happening in the future and help to restore the public’s confidence in the accounting profession.
The objective of this study is to examine the issue of audit failure by studying why some bankrupt companies receive unmodified opinions while others do not. This study examines the use of a contrary information, the occurrence of negative debt-related events, in auditors’ decisions to issue going-concern opinions to firms, taking into consideration the financial stress level of the firms.
Previous research by Mutchler et al  has shown that contrary information and mitigating factors will affect the going-concern modification decision made by auditors for soon-to-be bankrupt companies. Contrary information questions the client’s continued existence, and mitigating factors partially offset contrary information.
Organization of Report
The remaining of the paper is organized as follows. Part 2 reviews prior research done on issues relevant to this paper. Studies on explanatory factors of audit failure are examined. They are existence of contrary information and mitigating factors, and lack of auditor independence. Part 3 provides discussions on the extension of prior research and the development of hypotheses for this paper. It explains the rationale for testing those hypotheses. The hypotheses are classified into five areas, each testing the different effects on the variables involved. In addition, the process of sample selection and events studies is elaborated. Part 4 presents the descriptive statistics and data analysis. In addition, correlation and independent samples t-tests are performed to provide further analysis. Part 5 summarizes the findings of this research, states the limitations and provides some suggestions for future studies.
Many prior studies such as Altman and McGough [1974, 1982], Menon and Schwartz , Hopwood et al  and McKeown et al  had found that high proportion (more than 50%) of bankrupt firms did not receive a going-concern qualification (GCQ) in the year prior to their bankruptcies. It is questionable as to why such audit failure could have occurred. Could it be due to a lack of auditor independence? Or was the existence of contrary information at the time of auditing not strong enough to prompt GCQ or were there mitigating factors that led the auditors to conclude that the firms would survive?
Existence of Contrary Information and Mitigating Factors as Explanatory Factor of Audit Failure
Necessity to Consider Contrary Information and Mitigating Factors
Statement on Auditing Standards (SAS) No. 59 [AICPA, 1988] requires auditors to evaluate the prospects that a client will be able to continue in existence for a reasonable period of time as part of every engagement. In the evaluation, auditors are required to consider the contrary information and mitigating factors. Contrary information is information that questions the client’s continued existence [Mutchler et al 1997]. It increases the probability of bankruptcy and includes information such as debt or covenant defaults and substantial sale of assets to reduce debt [Sharma, 2001]. Mitigating factors are information that reduce the probability of bankruptcy and include financial and non-financial news and events [Sharma, 2001].
A number of prior researches examined whether the occurrence of negative debt-related events such as debt, covenant defaults, debt accommodations, had any association with the types of audit opinion issued. In essence, these studies tested whether the existence of contrary information would prompt auditors to issue GCQ. In addition, some studies tested if mitigating factors could explain the non-issuance of GCQ.
Contrary Information: Default Status and Restructuring
In Chen and Church , the default status of bankrupt firms in the year preceding bankruptcy was examined. The variable default included payment and covenant defaults together. The results showed that the relationship between default status and type of audit opinion received was not independent. The study found that about 77% of the firms receiving first-time GCQ were already in default or in the process of restructuring their debts, the purpose being to avoid subsequent default [Chen and Church, 1992].
Contrary Information: Payment and Covenant Default
In Mutchler et al , it was hypothesized that firms with payment or covenant default were more likely to receive qualified opinion. Unlike Chen and Church , the payment and covenant defaults were studied as two separate variables. The result supported the hypotheses.
Sharma  hypothesized that debt or covenant defaults would increase the likelihood of GCQ. Also it hypothesized that compared to companies that had never defaulted, those that had cured their problems were more likely to receive GCQ. The evidence showed that debt default and GCQ had a significant positive relationship. The hypothesis that firms with cured debt had a greater likelihood of being qualified was also supported.
Contrary Information and Mitigating Factors: News Reported
In Mutchler et al , the effects of news items reported as contrary information or mitigating factors were studied. They collected news items reported in the Wall Street Journal Index. The news found was classified as mild or extreme positive event or mild or extreme negative event. The news were further classified as “before” if they occurred in the time frame of two years prior bankruptcy to the last audit report date, and “after” if they belonged to the time frame of between the last audit report date and the bankruptcy date. The findings showed that only the “before” extreme negative events affected the opinion issued. The mitigating factors were not significant.
Auditors’ Independence as Explanatory Factor of Audit Failure
The debate over auditor independence is decades old. The public and regulators are concerned that with the fee dependency on the audit clients, auditors may compromise their independence. One consequence of compromise of auditor independence is audit failure. Many prior literatures undertook research in different ways to study whether auditor independence was sacrificed when the client was a huge source of revenue.
According to research from Massachusetts Institute of Technology, Michigan State University, and Stanford, auditors were more likely to condone earnings management when audit clients paid large non-audit fees [Anonymous, 2002]. From the result, it may indicate that the auditors had compromised their independence in exchange for the high fees. One possible implication of condoning earnings management is it gives a higher risk of corporate failure especially when the earnings management has been of a considerable scale. Since the auditors are not likely to issue GCQ to this group of firms, the result will thus be a higher probability of audit failure. One famous example of this would be Enron.
McKeown et al  hypothesized that smaller companies were more likely to be qualified, ceteris paribus. In the model for forming qualified opinion, the variable size was already incorporated in the variable probability of bankruptcy, yet it was found to be a significant variable in the model. Thus, the fact that the variable size was significant indicated that size was a factor in the auditors’ opinion decisions above and beyond the fact that there might be a relationship between bankruptcy and size [McKeown et al 1991]. This provided evidence that with a larger client size, auditor independence was more likely to be compromised.
Mutchler et al  investigated whether the use of contrary information and mitigating factors could explain the relationship between size and GCQ found in McKeown et al . The result showed that the presence of contrary information and mitigating factors did not explain the relationship. Once again, suggesting that auditors may compromise independence in the presence of larger clients.
In another study by Rose-Green et al , the auditor independence issue was investigated by examining the associations between audit and non-audit fees paid to a firm’s auditor. The research showed that audit firms collecting high audit fees from their client firms might be willing to reduce those fees in exchange for non-audit services contract. This implied that the provision of non-audit services influenced the auditors on audit-related decisions. Although in this case, the effect was on pricing of audit fees, the fees charged might have some impact on the audit effectiveness, since the resources allocated for the audit might be affected.
Contrary to the findings of the above-mentioned literatures, the following researches found no association between fee dependency and auditor independence. Reynolds and Francis  hypothesized that large clients created economic dependence, which was defined as a client’s size relative to the size of the audit office, which might cause auditors to compromise their independence and report more favorably to retain valuable clients. However, no support for the hypothesis was found. Evidence did not show that economic dependence caused the Big Five auditors to report more favorably for larger clients.
A study similar to the above, Craswell et al  investigated whether fee dependency affected the auditors’ exercise of independent judgment represented by the propensity to issue qualified opinions. The result found no support for the claim that the level of fee dependency would affect the auditors’ propensity to issue unqualified audit opinions.
DeFond et al  examined for correlation between non-audit services fees and auditor independence, whereby auditor independence was indicated by the auditor’s willingness to issue a GCQ. In addition to non-audit services fees, the correlation between total fees and auditor independence was also tested. Neither the non-audit services fee nor the total fees was found to have significant association with the auditors’ propensity to issue GCQ.
DEVELOPMENT OF HYPOTHESES AND RESEARCH METHOD
This study examines the issue of audit failure, investigating the possible explanations as to why auditors fail to issue GCQ to firms that displayed signs of financial distress (FD) prior to their bankruptcies. This could entail issues relating to the lack of independence. However, as numerous studies investigating audit fees as proxy of lack of independence have been done, this issue will not be explored in this study. Instead, effects of negative debt-related events and the debt magnitude of these events will be studied to further explore if these factors could possibly provide explanations for the non-issuance of GCQ.
The first part of this chapter develops the hypotheses that test the objectives of this study. The second part deals with the sample selection.
Effect of Financial Stress Level of Firms on Auditors’ Opinions
This study examines the effect of financial stress level of firms on auditors’ opinions. Firms in FD should be issued GCQ as SAS No. 59 [AICPA, 1988] requires auditors to evaluate the prospects of a client being able to continue in existence for a reasonable period of time as part of every engagement. Raghunandan and Rama  concluded that auditors were more likely to modify the opinions of distressed firms after the issuance of SAS No. 59. However, Carcello et al  did not find a difference in auditors’ inclination to issue GCQ after SAS No. 59.
On the other hand, Chen and Church  found evidence of auditors issuing GCQ when certain signs of FD were present. In view of the inconsistency in results previously obtained, this study examines the following hypothesis:
H1: Auditors tend to issue GCQ for bankrupt firms in FD prior to bankruptcy.
McKeown et al  proposed an identical hypothesis. The only difference lies in the proxy used for FD. In their study, the probability of bankruptcy was used as the proxy. In the above hypothesis, two variables, FD and GCQ are tested for any positive correlation between them. If auditors had issued GCQ for firms in FD and they subsequently went bankrupt, it shows that the auditors had done a good job in evaluating the going-concern probability of the firms.
Effect of Occurrence of Debt News on Auditors’ Opinions
Foster et al ’s study on usefulness of debt defaults and GCQ in bankruptcy risk assessment suggested that loan default/accommodation and loan covenant violation were both significant explanatory variables of bankruptcy. In addition, these events captured the information contained in auditors’ GCQ. Chen and Church , Mutchler et al  and Sharma  had concluded in their respective studies that the occurrence of debt-related events in a firm would increase the likelihood of it receiving a GCQ.
The occurrence of negative debt news prior to the firm’s bankruptcy may provide some insights to the firm’s going-concern problems. Further to prior studies in this area, debt news is classified into seven categories in this study (refer to Table 3.5). This study attempts to find out the impact of negative debt news on auditors’ decisions to issue GCQ, thus leading to the following hypothesis:
H2: The occurrence of negative debt-related events in the year prior to the bankruptcy prompted auditors to issue GCQ.
The hypothesis is further broken down into each category of the negative debt-related events as shown in Table 3.1. Each hypothesis is tested for any associative relationship between the occurrence of a certain type of debt news and the issuance of GCQ.
Table 3.1: Hypotheses Relating to Occurrence of Debt News and Auditors’ opinions
H3: The occurrence of covenant violation news prompted the issuance of GCQ.
H4: The occurrence of credit facility news prompted the issuance of GCQ.
H5: The occurrence of debt news prompted the issuance of GCQ.
H6: The occurrence of notes-related news prompted the issuance of GCQ.
H7: The occurrence of default news prompted the issuance of GCQ.
H8: The occurrence of restructuring news prompted the issuance of GCQ.
H9: The occurrence of waiver of debt news prompted the issuance of GCQ.
Occurrence of Debt News and Financial Stress Level of Firms
Prior FD studies [Giroux and Wiggins, 1984]; [Lau, 1987]; [Ward, 1994, 1995]; [Ward and Foster, 1996] researched on distress events other than bankruptcy experienced by stressed firms and concluded debt loan default/accommodation as one of the most severe FD events.
This study builds on these prior literatures to examine for correlation between the occurrence of negative debt-related events and financial stress level of firms. It is expected that FD firms would experience a higher incidence of debt-related events compared to non-financial distress (NFD) firms. This leads to the following hypothesis:
H10: The occurrence of negative debt-related events is positively correlated with FD.
This hypothesis is further tested for correlations between FD and each of the seven categories of debt news as shown in Table 3.2. This is to find out whether there is any significant correlation between FD and occurrence of a certain type of debt news.
TABLE 3.2: Hypotheses Relating to occurrence of Debt News and Financial Stress Level of Firms
H11: The occurrence of covenant violation news is positively correlated with FD.
H12: The occurrence of credit facility news is positively correlated with FD.
H13: The occurrence of debt news is positively correlated with FD.
H14: The occurrence of notes-related news is positively correlated with FD.
H15: The occurrence of default news is positively correlated with FD.
H16: The occurrence of restructuring news is positively correlated with FD.
H17: The occurrence of waiver of debt news is positively correlated with FD.
Magnitude of Debt on Financial Stress Level of Firms
This section further examines the magnitude of debts in the debt-related events on the financial stress level of the sample firms. This is an area that prior FD studies have not explored.
Besides considering the debt amount, which is taken to be the aggregate of all the debt amounts in the debt-related events taken into account within the time period, the magnitude of debt ratios are also examined. These debt ratios include ratio of the debt amount to the total sales, total assets, net profit and total liabilities in the year of the news occurrence. These ratios are included in this study as they are commonly used as performance indicators by both management and financial institutions when evaluating the liquidity and viability of the company or clients. Table 3.3 shows the 5 hypotheses that are tested in this section. The debt amounts and ratios of FD and NFD firms are tested for differences in mean.