Implications of mistakes in financial statements for short- and long-run returns

My first dissertation essay examines the question of whether markets overvalue (undervalue) firms that materially overstate (understate) earnings in the period preceding public announcements of mistakes in reported earnings. The analysis is based on a hand collected sample of firms that announced correction of material mistakes in financial statements during 1997-2002. I find that at the beginning of the period when a mistake is made but not yet corrected, investors rely on published incorrect earnings, especially those due to fraud. As the restatement date approaches, investors start penalizing firms that have overstated net income and seem to guess the sign of the restatement, but not the magnitude. This paper also extends the analysis of the impact of restatement announcement on the market by examining post-restatement stock performance and its determinants. My results show that more highly levered firms and firms that do not disclose the full impact of restatement underperform relative to control sample three years after the restatements announcement. My second dissertation essay investigates the extent to which market reactions to restatement announcements are explained by litigation risk. I model the relationship between restatement announcement returns and litigation risk as a simultaneous equations system. I find that firms with higher litigation risk have much larger negative market reaction to restatement announcement, controlling for other determinants of market reaction. A ten percent increase in the likelihood of litigation decreases announcement period cumulative abnormal return by approximately 1.55%. This finding suggests that when assessing the impact of restatements, investors consider expected future costs of litigation.