Detecting Earnings Management in Annual Reports: An Evolutionary Approach

The primary aim of this paper is to systematize the models used in literature to assess earnings management. By using an evolutionary approach, the paper discusses different model specifications useful to detect opportunistic behaviours in annual reports. A critical review of the literature leads us to conclude that earnings management is detectable by using specifications that can be conveniently classified in “first-generation” and “second-generation” models. The paper contributes to the literature and has implications in the practice, especially for investors. For scholars, it might be useful because it synthetizes the possible approaches that have been used over time by academics to detect earnings management. Investors might be aware of the models useful to detect the possible forms of misrepresentations in annual reports.