Document summary
Accounting governance, which is rooted within the principle of SRI or ESG investing, has become an increasingly integral component of the investment process. Accounting misstatements can cause severe share price underperformance and damage shareholder value. However, detection of corporate financial misstatements is typically a daunting task. Investors must sift through a multitude of financial statements and company behaviors in hopes of identifying accounting irregularities or anomalies. In this research piece, we discuss two interesting approaches to identify accounting fraud.
The first method is based on statistical analysis. In particular, we study the number distribution of a company’s financial statement data items, based on the Benford’s law. We find when a firm’s accounting data violates the typical distributional property of natural numbers, it suggests a higher probability of accounting irregularity and lower subsequent stock returns. In addition, using the AAER (Accounting and Auditing Enforcement Releases) database, we analyze the characteristics of companies exhibiting accounting irregularities. We further build a logistic regression model to predict accounting fraud. Our logit model can be quite useful to identify extreme underperformers and is uncorrelated with other quality factors.