A recommendation away from the consensus is more likely to be influential. Analyst characteristics such as a high ranking, more overall experience, or more experience on a particular firm all increase the likelihood of the analyst being influential.
Sometimes changes in analysts' recommendations appear to coincide with large changes in stock prices, while at other times seemingly analogous changes in recommendations appear to have no impact on market values. In When Are Analyst Recommendation Changes Influential?(NBER Working Paper No. 14971), co-authors Roger Loh and Rene Stulz attempt to determine under what conditions the change in an analyst's recommendation on a firm will influence the stock price of that firm.
This research draws on data from Thomson Financial's I/B/E/S U.S. Detail File, which contains stock recommendation ratings issued by individual analysts from 1993 to 2006. The authors find that 25 percent of the analysts never make a recommendation change that influences the price of the corresponding firm's stock. 'Influence' for this purpose is defined as coinciding with a statistically significant price movement, in the expected direction, in the underlying share price. Even among those analysts who do make recommendation changes that affect stock prices, not all of their changes are influential, only 20 percent are.
The authors ask what role analyst and firm characteristics play in determining whether recommendation changes have a statistically significant effect on share prices. They conclude that a recommendation away from the consensus is more likely to be influential. Analyst characteristics such as a high ranking, more overall experience, or more relative experience on a particular firm all increase the likelihood of the analyst being influential. In addition, when an analyst produces a recommendation change accompanied by an earnings forecast, there is more chance that the recommendation change will be influential.
If a firm is large, or a large number of analysts follow it, then a recommendation can more easily get lost in the crowd, and the probability of making an influential recommendation decreases. Similarly, if there is a strong consensus among analysts on one firm, or a high number of previous earnings forecasts, then the likelihood that a recommendation change will be influential is reduced. An analyst will have more impact on firms with lower turnover of shares and lower volatility of returns. Because analysts' reports are read mostly by an audience of institutional investors, analyst recommendations are more likely to be influential in firms with a higher institutional ownership.
When an analyst's recommendation is influential, analyst activity increases, as other analysts quickly revise their own ratings. Moreover, revisions of earnings forecasts are more extensive after an influential recommendation. The authors also observe an increase in share turnover and stock volatility in the aftermath of an influential recommendation.
Loh and Stulz are careful to select a sample of changes in analysts' recommendations that are not contaminated by the coincident release of other firm-specific news. They also examine how changes in the regulatory environment may affect the degree to which recommendation changes are influential. In particular, they find that both Reg FD -- the U.S. Securities and Exchange Commission's Regulation Fair Disclosure which was implemented in 2000 -- and the Global Analyst Settlement of 2003 increased the likelihood that a recommendation change would be influential.
-- Claire Brunel