The Accuracy and Incremental Information Content of Audit Reports in Predicting Bankruptcy Clive S. Lennox* 1. INTRODUCTION The efficiency of investment decisions depends on the accuracy of information available to investors. Auditing provides a means of ensuring that companies provide accurate information ± one role of an auditor being to warn investors when a company faces a significant possibility of bankruptcy. 1 However, in recent years there have been a number of well-publicised cases in which auditors failed to warn about impending bankruptcy and this has led to criticism of audit firms. 2 This paper attempts to evaluate and explain the accuracy and informativeness of audit reports in identifying failing companies ± its contribution to the existing literature is three-fold. First, it compares the accuracy of a bankruptcy model and audit reports in both estimation and hold-out samples (Koh, 1991). Secondly, the incremental information content of audit reports is evaluated taking into account a wider set of publicly available information than previously (Hopwood et al. (HMM), 1989). Finally, models of bankruptcy and audit reporting are compared to explain why audit reports were not very accurate signals of financial distress. Existing UK evidence indicates that audit reports are not accurate indicators of financial distress. Only 20±27% of failing Journal of Business Finance & Accounting, 26(5) & (6), June/July 1999, 0306-686X ß Blackwell Publishers Ltd. 1999, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA. 757 * The author is a Lecturer in Accounting and Economics at Bristol University and is grateful to Anindya Banerjee and Steve Bond for helpful comments and suggestions. (Paper received June 1998, revised and accepted November 1998) Address for correspondence: C. Lennox, Department of Economics, Bristol University, 8 Woodland Road, Bristol BS8 1TN, UK. e-mail: c.lennox@bristol.ac.uk