The Insurance Hypothesis: The Case of KPMG's Audit Clients

Abstract

Although many have argued that independent audits implicitly provide clients with a form of insurance (the insurance hypothesis), there is limited empirical evidence to support the existence and magnitude of this function. In August of 2005, after months of intense negotiations and discussions, KPMG entered into a settlement agreement with the Department of Justice, which ended widespread speculation of an impending federal indictment against the audit firm and the notion that the firm would suffer the same fate as Arthur Andersen. We argue that the circumstances surrounding the settlement provide a natural setting to test the insurance value provided by auditors. Our results show that, while BIG 4 non-KPMG client firms earn insignificant abnormal returns, KPMG client firms earn significantly positive abnormal returns during the days surrounding news of the settlement. We show that these positive abnormal returns vary cross-sectionally, and are more pronounced for KPMG client firms in greater financial distress and for those subject to greater litigation risk.