Is the Decline in the Information Content of Earnings Following Restatements Short-Lived?

Abstract

Prior research finds that the decline in the information content of earnings after restatement announcements is short-lived and that the earnings response coefficient (ERC), the proxy for the information content of earnings, bounces back after three quarters. We re-examine the persistence of the drop in the ERC after restatement announcements using a more comprehensive and recent sample of restatements. We find that material restatement firms experience a significant decrease in the ERC over a prolonged period – close to three years after restatement announcements. In contrast, other restatement firms experience a decline in the ERC only for one quarter after restatement announcements. In cross-sectional analyses, we find that among material restatement firms, those that are subject to more credibility concerns and those that do not take prompt actions to improve reporting credibility are associated with a longer drop in the ERC than others. Lastly, we reconcile our findings with prior studies. Our analyses indicate that using a potentially more powerful proxy for material restatements and imposing less restrictive sampling requirement help increase the power of the tests to detect the long-run drop in the ERC.

Keywords: accounting restatements, information content of earnings, accounting irregularities

JEL Classification: G32, M40