These tests can provide misleading results without a control for earnings overstatement in the early years of their service.We predict that in the early years of their service, when the market is more uncertain about their ability, they have greater incentives to overstate earnings to favorably influence the market's perception.
For the sample period1992–2010,we show that, as expected, earnings over statement is greater in the early years than in the later years of CEOs’ service and that this over statement increases reported ROA by about 25% on average.
These results are robust to using different earnings management measures, specifically, discretionary accruals, abnormal discretionary expenses, such as R&D expense, and special items.
We also show that, as expected, the difference in the over statement of earnings between the early and the later years of CEOs’ service is less pronounced in firms with stronger monitoring, proxied by institutional ownership, analyst following, board independence, and audit committee independence. Finally, our findings have an implication for the tests of earnings over statement by departing CEO sin their final year.
These tests can provide misleading results without a control for earnings overstatement in the early years of their service.