Accruals models are usually estimated cross-sectionally by year and by two-digit SIC code. This estimation approach, however, on the one hand is not applicable in small capital market countries, such as Italy, due to the lack of observations, and on the other hand relies on the implicit assumption of homogeneity of the accrual-generating processes within industries, which sometimes does not hold within two-digit SIC codes. We propose a new estimating technique, the mixture approach, which does not rely on industry classifications and allows researchers to estimate abnormal accruals for the vast majority of firms in small capital market countries. Our analyses suggest that accruals model tests are not less effective in Italy than in other countries and that the mixture approach provides the most powerful test for earnings management.
Keywords: Earnings management, accruals models, abnormal accruals, mixture model.