Do Analysts’ EPS Forecasts Obey Benford’s Law? An Empirical Analysis

Abstract

Benford’s law gives the expected frequencies of digits in tabulated data. In this study, I investigate the extent to which a sample of analysts’ earnings per share (EPS) forecasts obey Benford’s law. I conduct Benford’s law’s second digit and last-two digits tests on a sample of analyst EPS forecasts of S&P 500 firms from 1998 to 2018. Overall, I find that analysts’ EPS forecasts obey Benford’s law’s second digit test but do not obey the last-two digits test. These findings suggest that while analysts do not engage in number invention, they do engage in rounding when making EPS forecasts.