The Cost of Capital Effects of Accruals Quality before and after Restatements-A Revisit of Francis ET AL. (2005) in a Restatement Setting

Abstract

This study revisits Francis, Lafond, Olsson and Schipper (2005) in a restatement context by examining the cost of capital effects of accruals quality several years before versus after restatements. Francis et al. report a positive association of poorer accruals quality with higher cost of capital and a dominant effect of the innate over discretionary accruals quality in explaining the cost of capital. They claim that their results are unconditional on external indicators of extremely poor quality of earnings such as restatements. Using the U.S. General Accounting Office's (2002) restatement database, we find no support in the restatement setting for Francis et al.'s two major results. In addition, our results show a long-term decrease in the average accruals quality and a corresponding increase in cost of capital from before to after restatements. These results confirm Hribar and Jenkins's (2004) conjecture about a possible decrease in earnings quality following restatements, and support their finding of an increase in implied cost of capital after restatements. Moreover, we report no significant difference in average discretionary accruals quality between the pre- and post-restatement periods. This finding is in line with Moore and Pfeiffer's (2009) finding of no significant difference in abnormal accruals before and after restatements.