We find that analysts who frequently revise their stock recommendations outperform those who do not. This result holds for portfolios formed on the basis of favorable changes in recommendations as well as unfavorable changes. The frequency of revision captures information incremental to factors...