Does Real Earnings Management Matter in Default Prediction?

Abstract

This study examines the extent to which incorporating current-period and/or cumulative real activities earnings management in default models enhances their predictability. Aiming at Altman’s (1968) Z-score as well as Ohlson’s (1980) O-score predictors, such adjustments help mitigate the overestimation (underestimation) of survival probability for firms with aggressive (with conservative or less) current-period real earnings management. More remarkably, for financial distress detection models, we document significant effectiveness of adjusting for the cumulative earnings management over the previous three years. Consistently, false loan acceptance (rejection) rates for firms with upward (downward or no) earnings management are reduced with our modification on the scoring models.