A test of the determinants of auditor tendency to issue going-concern audit reports to nonfailing companies.

Abstract

The purpose of this study is to develop and test a model to explain why auditors issue going-concern reports to companies that subsequently do not fail. Given the auditor's access to records and unique interaction with management and legal council during the audit, it seems plausible that the auditor' s going concern report can serve as a useful indicator or "signal" of the company's potential inability to maintain itself as a going-concern. However, before one can even discuss the usefulness of the going-concern audit report, it must be shown to be reliable. Developing a model to explain why the auditor's signal of a client's going-concern status subsequently proves to be unreliable should help financial statement users better assess the information value or usefulness of the going-concern audit report. In this study the likelihood that the auditor has sent a reliable signal of the company's subsequent economic status to financial statement users was hypothesized to be a function of the client and auditor related factors which affect the auditor's judgements and/or reporting decisions. The tests showed that the company's estimated probability of bankruptcy, an indicator of ambiguity and complexity in the auditor's decision-making environment, was a significant determinant of signaling reliability. In addition, audit technology was found to be significantly related to signaling reliability, with more structured auditing firms issuing reports which appear to be more reliable than the audit reports issued by less structured auditing firms. Finally, subsequent mitigating actions or events which are reported between the date of the audit report and the subsequent financial statements are significantly negatively associated with signaling reliability of the going-concern audit report. The remaining client and auditor-related factors--audit/client tenure, auditor's industry concentration ratio, and client size relative to the auditor's total client base are not significantly associated with the report's signaling reliability. However, the results suggest that these factors should be further examined.