Earnings Management? Erroneous Inferences Based on Earnings Frequency Distribution: A Reply to Durtschi and Easton (2009)

Abstract

Durtschi and Easton [2009, DE2009] identify methodological issues with a large literature that attributed “discontinuities” in the earnings distribution around the zero profit line to earnings management. To address concerns raised by DE2009, we avoid inferences from the year-end distribution of deflated earnings, and focus, instead, on un-deflated earnings and EPS during the fourth quarter using the approach of Kerstein and Rai (2007). We find: 1 abnormal shifts of 1, 2, or 3 intervals away from small losses into the smallest profit interval 2) firms that are barely profitable after three quarters remaining so at an abnormal rate; and 3) firms avoiding and/or shifting away from the smallest loss interval at a higher rate than any of the 20 intervals surrounding the zero profit line. In addition, we find evidence that contradicts one of the main assertions made by DE2009, i.e., that for firms near the zero profit line after three quarters, the integral method of accounting severely dampens the income during the fourth quarter. Overall, our findings are consistent with earnings management to avoid smallest losses and achieve smallest profits.