Does Competition Induce Hardworking or Herding? Evidence from a Natural Experiment of Broker Mergers

Abstract

Hong and Kacperczyk (2010) document that decreases in analyst competition due to broker mergers encourage analysts to please managers, leading to greater consensus optimism bias. We propose two addition effects of analyst competition. The hardworking hypothesis suggests that lower competition among analysts reduces their incentives to collect and analyze information, leading to less accurate forecasts and less informativeness on earnings or recommendation revisions. The herding hypothesis argues that lower competition reduces their career concerns and herding incentives. Analyst therefore place less (more) weight on common public (private) information, leading to more accurate forecasts and more informative forecast and recommendation revisions. We find evidence that is more consistent with the hardworking hypothesis than the herding hypothesis or the optimism-bias hypothesis.