Does Competition Improve Analysts' Forecast Informativeness?

This paper studies the effect of competition on analysts' forecast informativeness. I show that the impact of competition on forecast informativeness is ambiguous in general, and identify the necessary and sufficient conditions under which more intense competition can make forecasts less informative. If prior uncertainty is sufficiently high, analysts' payoff structure is convex in their forecast accuracy rankings, and their private signals are conditionally correlated, then competition impairs information revelation. When analysts know that their research may be conditionally correlated with their competitors' research, disclosing honest estimates based on private information makes their forecasts less likely to stand out. Therefore, they tend to differentiate themselves from each other by understating their private signals, revealing less information in equilibrium. One possible fix is to soften the tournament-like aspect of analysts' reward structure; for example, by reducing the convexity of analyst compensation