Use Of Earnings Information For Stock Pricing In Different Market Cycles: The Effect Of Discretionary Accruals

Main Article Content

Alireza Daneshfar
Mohammad J. Saei

Keywords

Abstract

This study examines the association between stock prices and discretionary accruals in different stock market cycles and presents evidence about the discrepancy in prior research that investors were able to identify earnings management in some cases, but not in some other cases. We argue that investors’ reaction to the true nature of EPS changes may be different in different market cycles. We suggests that investors pay less attention to the nature of EPS changes in an optimistic cycle, and are more critical in neutral and pessimistic cycles. Therefore, investors are more likely to detect and count for any earnings management in a neutral or pessimistic cycle than in an optimistic cycle. Using the U.S. quarterly data from July 01, 1997 to June 29, 2001, three market cycles were identified: optimistic, neutral and pessimistic. The test results indicated that the association between discretionary accruals and abnormal stock returns were insignificant in the neutral market cycle, significant and positive in the optimistic cycle and significant and negative in the pessimistic cycle. These findings indicate that investors tend to ignore the income-increasing effect of discretionary accruals on EPS changes in an optimistic market. The finding suggests that a more delegate and technical analysis of EPS changes is required when earnings information is used for stock pricing. It also suggests that a consideration of market cycle effect on investors’ use of EPS could improve the earnings-based ratio analysis. The findings propose that researchers interested in investigating the association between stock prices and earnings management should control for the effect of the market cycle during which their samples are drawn.

Downloads

Download data is not yet available.
Abstract 170 | PDF Downloads 222

Most read articles by the same author(s)