Catch Me If You Can : Improving the Scope and Accuracy of Fraud Prediction

We propose a parsimonious metric – the Adjusted Benford score (AB-score) – to improve the detection of financial misstatements. Based on Benford's Law, which predicts the leading-digit distribution of naturally occurring numbers, the AB-score estimates a firm-year's likelihood of financial statement manipulation, compared to its peers and controlling for time-series trends. The AB-score's biggest advantage is coverage: It can be computed for about 60% more firm-years than the leading accounting-based metric (the F-score) without sacrificing accuracy. Notably, it can be computed for financial firms, which are often excluded from financial misconduct research due to data availability issues. For firm-years with all data available, combining the AB-score and F-score variables into one model yields higher accuracy in predicting misstatements. Our metric performs well out-of-sample as well as in-sample, across different misstatement databases, and for a set of notorious financial frauds. It should be especially useful to regulators and industry professionals