Earnings Management and Accounting Income Aggregation

Abstract

Quarterly earnings allow aggregation into annual earnings in four different ways. Fiscal year reported earnings is one of these four possible measures of annual earnings, the others being earnings for years ending at the first, second and third fiscal quarters. We provide evidence on earnings management in fiscal year earnings relative to these three alternative measures of firms' annual earnings. We confirm prior findings in Burgstahler and Dichev (1997) of discontinuities around zero and around prior year earnings in histograms of fiscal year earnings. Subsequent research questions whether these discontinuities are evidence of earnings management or whether they are attributable to biases induced by taxes, scaling and sample selection. Using the histograms of our alternative annual earnings measures, we offer additional evidence in this debate. We also find evidence of earnings management in broader intervals around thresholds. We believe that our research design is better suited to test for earnings management in these broader intervals than those used in prior studies. We also compare the statistical properties of fiscal year earnings to annual earnings starting with the fiscal year quarters two, three and four. We find that the variance and kurtosis of earnings are higher for fiscal year earnings while skewness of earnings is lower at the fiscal year. These results are more consistent with earnings management than with the effects induced by 'settling up' in fourth quarter earnings. Overall, this study contributes to the literature on the prevalence, effects of and factors associated with earnings management.