Brokerage Houses and their Stock Recommendations: Does Superior Performance Persist?

Abstract

This paper tests for the existence of performance persistence in brokerage house stock recommendations. For the period 1987-1996 we show that purchasing the current-year buy recommendations of the brokerage houses with the best prior performance earned an annualized geometric mean raw return of 18.6 percent, while purchasing the recommended stocks of the houses with the worst prior performance earned only 14.3 percent. After controlling for market risk, size, book-to-market, and price momentum effects, though, we find no significant difference, in general, between the abnormal returns of the best and worst brokerage houses. A series of supplementary tests confirm this result. The findings for brokerage house sell recommendations are even weaker. Overall, our tests provide no reliable evidence of performance persistence for brokerage house stock recommendations.