Identifying unexpected accruals: a comparison of current approaches

While prior research, as noted in our paper, often uses various accrual prediction models to detect earnings management, not much is known about the accuracy, both relative and absolute, associated with these models. Our paper investigates the accuracy of six di€erent accrual prediction models, and o€ers the following ®ndings. Only the Kang ± Sivaramakrishnan (1995) model performs moderately well. The remaining ®ve models provide little ability to predict total accruals: they are less accurate than a naõve model which predicts that total accruals equal ÿ5% of the total assets (TA) for all ®rms and years. Conventional R values from a regression of actual accruals on predicted accruals are less than zero for a substantial majority of ®rms for these ®ve models. These low R values in the prediction period contrast sharply with the much higher R values that are obtained within the estimation period. Similar performance is observed when predicting current accruals alone. However, the relative rankings of the di€erent models are altered somewhat: the Jones (1991) model is then the only model that exhibits some predictive ability. Ó 2000 Elsevier Science Ltd. All rights reserved.