Determinants of Bias in Management Earnings Forecasts

Abstract

This study investigates the effect of five firm characteristics on bias in management earnings forecasts. These characteristics are hypothesized to have different effects on bias. Abnormal earnings growth, industry competitiveness, and disclosure-related legal liability are expected to lead to forecast pessimism, while financial distress and external financing are hypothesized to lead to forecast optimism. Using a sample of 242 management earnings forecasts made by U.S. firms during the first three fiscal quarters of the 1990-95 time period, the results show that these forecasts are on average optimistic. In both univariate and multivariate analysis, abnormal earnings growth and financial distress are significant in the hypothesized direction. The multivariate model is able to explain 25.6% of the variation in the bias measure. These findings are robust across most sub-samples and are also invariant to different proxies used to measure the various constructs.