How Do Restatements Begin? Evidence of Earnings Management Preceding Restated Financial Reports

Abstract

Most earnings restatements are blamed on error, or misunderstanding of GAAP, but suspicion persists that many of these restatements are instead due to intentional earnings management. We analyze balance sheet bloat, or unusually high levels of working capital account balances, for evidence of sustained, income-increasing earnings management prior to initial non-GAAP financial reports. We establish a pattern of systematically increasing balance sheet bloat for firms later issuing clearly fraudulent financial reports. Next, we compare bloat for apparently non-fraud restatements to fraud and control samples. We find non-fraud restatement companies‟ bloat is higher than control companies for two years preceding the initial misstated financial report. But, these firms accumulate less balance sheet bloat than companies with restatements clearly involving fraud. This suggests meaningful, but not pervasive, earnings management underlying even apparently non-fraudulent restatements. We extend our analysis to discretionary accruals and real activity earnings management.