The Effectiveness of Accruals Manipulation Measures: International Evidence

ABSTRACT This study examines the ability of abnormal accruals measures to capture earnings manipulation in international settings. Variations in accrual recognition are common in international settings due to differences in reporting standards and institutional features. Moreover, data availability imposes significant constraints on sample sizes in international research examining accruals. Thus, we evaluate how well abnormal accruals measures capture earnings manipulation under alternative model specifications that maximize sample sizes. We demonstrate that sales growth can reliably replace employee growth in the estimation of abnormal accruals, thereby alleviating employee data constraints for non-U.S. firms. Our study shows that the Dechow, Hutton, Kim, and Sloan (2012) and modified Larson, Sloan, and Zha Giedt (2018) models are most consistent in capturing earnings manipulation across international samples. Additional analysis indicates that the effectiveness of accruals manipulation measures is consistent across commonly used databases: Worldscope, Compustat, and Osiris. JEL Classifications: M41; G14.