Massimo Cecchi
This paper proposes an approach for improving the estimation of artificial smoothing, identifies important biases that affect previous research and proposes a more critical use of traditional models. The developed approach differs from the methods of previous studies because it first involves a more precise application of the concept of artificial smoothing. Second, the weight of individual artificial smoothing levers is identified. Third, we verify the existence of a smoothing effect derived from the matching principle. Our study of a sample of 4,426 Italian companies and 22,695 firm-year observations confirms the overall smoothing effect of adjusting entries. However, our analysis of individual levers shows that only the weight of the change in inventories is very strong. The paper demonstrates that this strong effect can also be determined only by the matching mechanism without any relation to managerial discretionary assessments. This finding suggests the need for a more critical use of models that measure artificial smoothing and the need to exclude these objective variations when analyzing the earning management estimate.