Unpacking analyst forecast bias: The role of optimism and sequence in shaping earnings predictions

Abstract

Earnings forecasts by financial analysts are critical to guiding investment decisions and corporate valuations. This study examines how forecast sequence (disaggregation vs. aggregation) interacts with initial optimism (presence vs. absence) to shape the accuracy of earnings predictions. A 2×2 between-subjects experimental design was employed, involving 97 professional financial analysts from leading U.S.-based brokerage firms with extensive experience in equity research. These analysts, representative of the target population making critical market forecasts, were tasked with predicting the annual earnings per share (EPS) of a hypothetical global hospitality firm, Firm X, listed on the New York Stock Exchange. The sample was chosen to ensure high external validity by mirroring real-world practices and decision contexts in financial forecasting. Initial optimism was manipulated using "strong-buy"and "neutral"stock recommendations, while forecast sequence was adjusted by requiring updates either after each management announcement (disaggregation) or collectively (aggregation). Results demonstrate that disaggregation amplifies optimistic bias in the presence of initial optimism, resulting in inflated earnings forecasts. This effect is attributed to confirmation bias. In contrast, no significant differences in forecasts were observed between sequences in the absence of initial optimism. These findings offer practical insights into mitigating cognitive biases in financial analysis, emphasizing the dualedged role of disaggregation. Future research may extend these findings across diverse industries and forecasting contexts to further refine strategies for enhancing decisionmaking accuracy and investor trust.