We study whether analysts’ recommendations and the market’s reactions to recommendation changes are influenced by the structure of analysts’ research portfolios. We find that analysts maintain more positive recommendations for stocks that belong to the “core industry” in their research portfolios, and are more likely to upgrade these core stocks. Consequently, diversified analysts, who cover
... [Show full abstract] stocks from industries other than their core industry, make less optimistic recommendations and are less likely to upgrade their recommendations. We also find that the market’s reactions, captured by announcement returns, are stronger for recommendation changes for non-core stocks and recommendation changes made by diversified analysts. Finally, all these patterns are significantly less pronounced after the passage of Regulation Fair Disclosure, suggesting that the bias in recommendations is the result of strategic considerations and not self-selection.