] A comparison of buy-side and sell-side analysts

There is very little research on the topic of buy-side analyst performance, and that which does exist yieldsmixed results. Weuse a large sample fromboth the buy-side and the sell-side and report several newresults. First, while the contemporaneous returns to portfolios based on sell-side recommendations are positive, the returns for buyside analysts, proxied by changes in institutional holdings, are negative. Second, the buy-side analysts' underperformance is accentuatedwhen they trade against sell-side analysts' recommendations. Third, abnormal returns positively relate to both the portfolio size and the portfolio turnover of buy-side analysts' institutions, suggesting that large institutions employ superior analysts and that superior analysts frequently change their recommendations. Abnormal returns are also positively related to buy-side portfolios with stocks that have higher analyst coverage, greater institutional holding, and lower earnings forecast dispersion. Fourth, there is substantial persistence in buy-side performance, but even the top decile performs poorly. These findings suggest that sell-side analysts still outperform buy-side analysts despite the severe conflicts of interest documented in the literature. Crown Copyright (C) 2015 Published by Elsevier Inc. All rights reserved.