Herding, Information Aggregation and Momentum E¤ects (Job Market Paper)

摘要

Sell-side equity analysts at times have a tendency to herd toward the consensus estimate when making their quarterly earnings forecasts. I argue that such tendency to herd leads to inefficient aggregation of private information and consequently price momentum in stocks. I demonstrate that the Jegadeesh and Titman (1993) price momentum phenomenon is present among stocks only during those time periods when analysts who follow those stocks herd together. I nd that the herding tendency is stronger among smaller stocks, growth stocks, and stocks with higher share turnover ratio and more news media coverage. I provide diagnostics suggesting that my ndings are distinct from the earnings momentum effects, information uncertainty effects and liquidity risk already documented in the literature.