An Empirical Model of Stock Analysts' Recommendations: Market Fundamentals, Conflicts of Interest, and Peer Effects

Abstract

In this paper we develop an empirical model of equity analyst recommendations for firms in the NASDAQ 100 during 1998-2003. In the model we allow recommendations to depend on publicly observed information, measures of an analyst's beliefs about a stock's future earnings, investment banking activity, and peer group effects which determine industry norms. To address the reflection problem, we propose a new approach to identification and estimation of models with peer effects suggested by recent work on estimating games. Our empirical results suggest that recommendations depend most heavily on publicly observable information about the stocks and on industry norms. In most of our specifications, the existence of an investment banking deal does not have a statistically significant relationship with analysts' stock recommendations.