The Importance of Analysts vs. Brokers in the Performance of Stock Recommendations

Abstract

We analyze who plays a more important role in the success of a stock recommendation: the analyst or the brokerage firm that employs the analyst. Using a Bayesian methodology that models abnormal performance as the output of a Cobb-Douglas production function with analyst and broker inputs, we find evidence that the brokerage firm drives the announcement effect while the skill of the analyst ultimately determines the long-run success of a recommendation. Top-performing analysts who switch brokerage firms are likely to maintain their strong track records. Similarly, analysts who perform poorly continue to do so regardless of their employers. Our results hold over the sub-sample periods that surround the sell-side analyst reforms of 2002.

Keywords: Analyst, Brokerage Firm, Recommendation, Bayesian Econometrics, Conflict of Interest

JEL Classification: G24, G29