Compared to What? Does Benford's Law Really Detect Corporate Fraud?

This paper criticizes a publication in the Review of Accounting Studies published in 2015 titled “Financial Statement Errors: Evidence from the Distributional Properties of Financial Statement Numbers” by Dan Amiram, Zahn Bozanic, and Ethan Rouen. The paper won the 2017 Deloitte Foundation Wildman Medal Award and was featured in the financial press including the Wall Street Journal. In the paper, the authors develop a measure, based on Benford’s law, to identify errors in financial statements. The authors claim that their measure can do two things: (1) predict material misstatements as identified by SEC Accounting and Auditing Enforcement Releases, and (2) serve as a leading indicator for these misstatements. I ask an additional question, “Compared to What?” Specifically, I compare their measure to (1) the F-Score, which is the parsimonious standard in the accounting literature, and (2) a naive screen on sales growth. The evidence in this paper shows that Benford’s law does not measure up to the prediction of material misstatements nor does it meaningfully serve as a leading indicator relative to these measures. This topic is important to SEC regulators and industry professionals, particularly as a similar measure that cites the award-winning paper is currently being offered commercially through AuditAnalytics.