Asymmetric Timeliness of Earnings Recognition and Overestimation of Accruals Quality

Abstract

This study examines the relationship between the asymmetric timeliness of earnings recognition and accruals quality. Ball and Shivakumar (2006) provide evidence that the relationship between accruals and cash flows is distorted under the linear assumption. This study estimates new measures of accruals quality after including the asymmetric timeliness of earnings recognition in Dechow and Dichev’s (2002) model and compares these with Dechow and Dichev’s accruals quality. I decompose Dechow and Dichev’s accruals quality into (1) one component after controlling for asymmetric earnings timeliness, which is unexplained by asymmetric timeliness recognition, and (2) another component that is explained incrementally by asymmetric timeliness recognition beyond three consecutive-period operating cash flows. I examine the relationship between decomposed accruals quality and earnings forecast accuracy, cost of debts, and cost of equity. Empirical findings show accruals quality significantly improves after incorporating asymmetric timeliness of earnings recognition. The accruals quality component after incorporating asymmetric timeliness shows significant relationships with earnings forecast accuracy and cost of equity, while the other component does not. This study shows that accruals quality improves when estimated by allowing a non-linear relationship with asymmetric timeliness. This measure enables users to make better decisions when evaluating firms’ operating or financing activities. Key words: Accruals Quality; Asymmetric Timeliness of Earnings Recognition; Cost of Debts; Cost of Equity; Earnings Forecast Accuracy JEL Classification: G10, M40