Differential Informativeness of Accrual Measures to Analysts’ Forecast Accuracy

This paper evaluates whether analysts incorporate formal measures of earnings quality into their earnings forecasts. It examines whether the accrual ratio and abnormal accruals, measured with the Modified Jones (1991) Model of discretionary accruals, differentially inform analysts’ earnings forecasts. It uses the accuracy of analysts’ forecast as a context in which to evaluate how well analysts incorporate effects of the information contained in accrual ratio and abnormal accruals. The results indicate that the accrual ratio is negatively related to the absolute value of analysts’ forecast errors while the Modified Jones (1991) Model of discretionary accruals have virtually no economic effect on analysts’ forecast error. The insignificant effect of discretionary accruals on analysts’ forecast may be attributed to analysts having already incorporated the information therein in their earnings forecasts, effect of the accrual anomaly having been largely arbitraged away by market participants or both. This paper contributes to the research on analysts’ earnings forecast and earnings quality and helps bridge the gap between practice and theory by demonstrating the differential impact of discretionary accruals (favored by academics) and the accrual ratio (favored by analysts) on analysts’ forecast accuracy. This study informs researchers and policy makers interested in better understanding how analysts affects the financial markets including how they may have learned from previously documented market anomalies such as the accrual anomaly. This is important as ultimately, efficient economy-wide capital allocation decisions are based partly on outputs of analysts’ forecasting processes