Institutional Consensus After Earnings Announcements: Information or Crowding?

Abstract

This paper examines information processing skills of institutional investors after earnings releases. If institutions correctly process earnings signals, their trades should push the price towards the new fundamental value. However, if they mechanically follow a positive-feedback strategy, Stein (2009) predicts that their crowding can lead to price overreaction. Splitting institutions by their investment horizon, we find that institutions with longer-term horizons are better at processing earnings signals, whereas crowding by shorter-term transient institutions can have destabilizing effects on stock prices. Overall, we reconcile previously mixed empirical evidence of institutional trading on price efficiency by conditioning on the length of the investment horizon.

Keywords: Institutional Consensus, Information Processing, Price Efficiency, Crowded Trading, Earnings Announcement

JEL Classification: G10, G14, G20