Predicting Influential Recommendation Revisions

We find that 19% of sell-side analysts’ stock recommendation revisions move stock prices significantly on the event date. With a model that predicts influential revisions out-of-sample, we show that these predictions can be... [ view full abstract ]

We find that 19% of sell-side analysts’ stock recommendation revisions move stock prices significantly on the event date. With a model that predicts influential revisions out-of-sample, we show that these predictions can be used to form long-short one-month portfolios that earn an alpha of 26% per annum in the period between 1999 and 2013. This corresponds to more than five times the annualized returns of price momentum. Contrary to previous evidence, this strategy survives to substantial transaction costs. These findings show that recommendations are an important means by which analysts assimilate information into stock prices.