Model for studying earning manipulation in Indian market

Abstract

The key objective of our research is to come up with the metric or application of the previously established metrics on the financial statements of Indian Listed Companies dataset to identify the possible earnings manipulation and investigating the reasons for the same. We applied the Beneish Model to identify the potential manipulator companies on the dataset of ~ 1,000 listed companies in India and then analysed some of the manipulators to check the efficiency of the model. We then applied non-linear models to identify the outliers on the key financial statement ratios which can highlight the possible instances of earnings management. Hence, the context is to identify the possibility of earnings management based on the plausible relationships between financial statement variables and their movements over the period. Our work is the first to extend the idea of non-linear models to Beneish based features in the current literature. Particularly, we consider two non-linear models: DBSCAN and Isolation Forests.