Information Aggregation, Herding, and Momentum Effects

Abstract

This paper studies how market's aggregation of public and private information may be related to momentum effects. Consistent with the information externality view of herding behavior, I characterize the overweighting of public information as informational herding. Using three distinctive types of news: past return news, quarterly earning announcement news, and corporate issued guidelines news, I provide empirical evidence suggesting that the momentum effect is closely related to informational herding. First, various momentum effects almost entirely concentrate in the portfolios where the market excessively relies on public information and poorly aggregates private information. Second, there are different future price movements among stocks sharing similar prior news but differing in whether market herds on the news. Third, information uncertainty about fundamental values is probably only a necessary but not sufficient condition for the momentum effect. That is, it is not uncertainty per se that leads to the momentum effects, but rather the particular way the information is weighted and gets aggregated into the market that matters.