سرمایه گذاری با تعداد دفعات بالا ، پیشنهاد تحلیلگر تغییرات برای سرمایه گذار عادی

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 known to identify superior recommendations. Although much of the frequently revising analysts' advantage follows events proxied by abnormally high returns or trading volume, it does not appear to derive from more public events such as earnings announcements. Further, these analysts outperform their counterparts even over the short-run, suggesting that this is not simply a “quantity over quality” phenomenon. In summary, our results imply that the superior profitability of frequently revising analysts emanates at least partly from their ability to generate private information using their superior skill. Overall, the ordinary investor is better off following the advice of analysts who revise their recommendations more frequently.

One of the oldest questions in finance is whether investment professionals, a.k.a. “experts,” can predict or impact the market. Much of the literature suggests that when analysts speak, investors listen.2Dimson and Marsh (1984) find that analysts' predictions of future stock returns are exaggerated, but directionally correct on average. Elton et al. (1986) find that abnormal returns congruent with the nature of the recommendation exist for up to two months after the recommendation is made, while Womack (1996) finds significant excess returns for up to six months after recommendation downgrades. Similarly, Barber et al., 2001, Barber et al., 2003 and Jegadeesh et al., 2004 find that abnormal returns increase with a stock's level or change in average recommendation. These empirical findings are consistent with Grossman and Stiglitz (1980) that information gatherers bring new information to the market to justify their compensation. Most of the earlier studies have focused on the aggregate performance of recommendations and recommendation revisions. More recently, there is a host of papers providing evidence on cross-sectional differences in recommendations' profitability. One reason for these differences is that some analysts may have better skill or better access to private information. Loh and Mian, 2006 and Ertimur et al., 2007 find that analysts who issue more accurate forecasts also make more profitable recommendations. Mikhail et al., 2004 and Li, 2005 document that the past performance of an analyst's recommendations tends to persist in the future. Fang and Yasuda (2011) find that top-ranked All-American analysts, who tend to be more experienced than lower-ranked All-American and other analysts, recommend better investments as measured by subsequent risk-adjusted returns. Jegadeesh and Kim (2006) show that analysts who make bolder recommendations than the consensus have a greater price impact. Loh and Stulz (2011) argue that star analysts and more experienced analysts tend to issue more influential stock recommendations than do other analysts. However, Emery and Li (2009) suggest that analyst status can be likened to a “popularity contest” from which subsequent recommendations do not yield positive abnormal profits — and in some cases even yield negative profits. Another reason for the cross-sectional differences in analysts' recommendations is that some analysts may have distorted incentives. For example, Barber et al. (2007) find that analysts employed by investment banks provide less profitable buy recommendations than analysts employed by independent research firms. Recommendation profitability can also differ based on corporate events. For example, Bradley et al., 2008 and Loh and Stulz, 2011 indicate that recommendations issued concurrently with companies' earnings announcements are more influential. Finally, there is evidence that recommendations accompanied by earnings forecasts are more profitable (e.g. Kecskes et al., 2010), and that other items included in analysts' reports, such as price targets and qualitative analysis, also impact the profitability of recommendations (e.g. Asquith et al., 2005). In this paper we investigate the relative performance of analysts' recommendation changes based on how frequently they revise a typical recommendation. There are at least three reasons to believe that analysts who differ by the frequency of revision may have differential profitability for investors. Analysts compare their own assessment of a firm's fundamental value with the market price and revise their recommendation when they see a substantial change relative to their outstanding recommendation. It is possible that some analysts have better knowledge of the industry, more timely access to management, suppliers, and customers, or alternatively are able to process publicly available information more efficiently than others. We expect that analysts with an advantage in information or skill are able to identify mis-valuation in stock prices more often, and thus revise their recommendations more frequently. Therefore, trading on their recommendations will be more profitable to investors in terms of exploiting short-term mispricing. Frequent recommendations can also reflect more timely communication with investors, an aspect of analyst performance deemed important by Institutional Investor survey participants (Leone and Wu, 2007). Alternatively, one could argue that analysts who revise their recommendations more often are acting on noise, perhaps owing to overconfidence (Barber and Odean, 2000 and Odean, 1998). Accordingly the analysts who revise recommendations less frequently are the ones who actually uncover new information. Furthermore these analysts may be more cautious and perhaps use multiple valuation screens before making their recommendation revisions. By this reasoning, investors may be better off following the recommendations of analysts with less frequent revisions. Additionally, it is possible that the type of information the analyst has determines how frequently he or she changes recommendations. If this is the case, analysts with short-term information will more frequently change recommendations while analysts with long-term information will not change recommendations as frequently. Therefore investors will be better off in the short-run following the recommendations of analysts with more frequent revisions. However, the analysts with less frequent revisions may provide more value to investors in the long-run. Our study examines whether the frequency of recommendation revisions relates to their profitability. We attempt to generate a profitable trading strategy for investors by focusing on performance differences at the analyst level, similar in spirit to Li, 2005, Barber et al., 2007 and Fang and Yasuda, 2011. The findings herein could potentially help the ordinary investor, who has limited time and resources and is likely to follow just one or a few analysts, identify superior performers from the universe of all sell-side analysts based on a simple measurable yardstick of frequency of recommendation changes in the recent past. Our study also contributes to the academic literature by providing information about the sources of competitive advantage in sell-side equity research. Finally, our work complements some of the findings in the buy-side literature. Yan and Zhang (2009) find that the positive alphas associated with institutional investors are concentrated in those institutions whose focus is on short-term, rather than long-term, profits. Although the work environment of sell-side analysts substantially differs from that of buy-side managers in terms of regulations, bases for evaluation and compensation, employers, and responsibilities (Schipper, 1991), to the ordinary investor the primary question remains unchanged: which type of analyst yields more profitable recommendations? We then explore some of the potential reasons for why frequently revising analysts yield greater returns. The source of this higher profitability could, for example, be an enhanced ability to interpret publicly available information. However, we find no statistical difference between frequently and infrequently revising analysts in the percentage of recommendation changes that occur around earnings announcements, and the difference in profitability between the two groups is not concentrated in those revisions that are made during earnings announcement periods. Additionally, it is possible that frequently revising analysts respond more quickly to other information events proxied by abnormally large stock returns or trading volume (in the spirit of Leone and Wu, 2007). Indeed, we find that frequently revising analysts respond more quickly both to large stock price changes and above-average turnover than do infrequently revising analysts, and those responses yield much higher profits. Finally, we examine whether the underperformance of infrequently revising analysts is driven simply by their relative lack of revisions. We compare the short-run profitability of the two analyst groups by excluding from our portfolios all outstanding recommendations older than two months, and find very similar results to our “long-run” analysis. Thus it appears that those analysts who most frequently revise their recommendations do better even in the short-run; they do not outperform their counterparts simply on the basis of quantity. Overall, our results suggest that at least part of the superior profitability of frequently revising analysts' recommendations derives from a better knowledge of the industry or companies they cover or from more timely access to management, suppliers, and customers. The remainder of the paper is organized as follows. Section 2 describes the data and methods used. Section 3 documents the main results, followed by robustness checks in Section 4. Section 5 investigates the source of the profitability of recommendations made by analysts with more frequent revisions. Section 6 concludes the paper.

Several studies document that the positive excess return subsequent to analyst recommendations (e.g. Womack, 1996, Logue and Tuttle, 1973 and Dimson and Marsh, 1984) is driven by analysts with certain characteristics. Consistent with the empirical literature, we find positive “alphas” subsequent to analyst recommendations, but we also find that the analysts who most frequently change their recommendations are the ones with the highest excess returns. We conduct several robustness tests and show that our results are not subsumed by characteristics that are known to predict performance, such as analyst independence, experience, star status, past profitability, the number of stocks covered, or whether a recommendation is accompanied by an earnings forecast. Thus it appears that analysts in our first quintile are simply able to identify mispriced stocks more often. These robustness results suggest that the frequency of recommendation revisions captures incremental information ignored by current factors known to help identify superior analyst recommendations. Additionally, we conduct further analysis to explore the source of frequently revising analysts' abnormal return advantage. We find that most, if not all, of this advantage is concentrated in recommendations that do not occur after earnings announcements. We then look to see whether frequently revising analysts respond more quickly to other types of information events proxied by large stock price changes and trading volume. We find that first-quintile analysts not only are significantly quicker to respond to this type of investor behavior, but that their profits relative to the fifth quintile are also much higher. Last, we shorten the holding period for all stocks in our portfolios to a maximum of two months and find that the results remain largely unchanged. Our results suggest that analysts who revise their recommendations more frequently are more responsive to the market and do not derive their advantage from firms' earnings reports. A significant part of their overall advantage lies in those revisions that follow unusual market activity. Additionally, their higher abnormal returns do not appear to be a result of simply having made more revisions. Our results are consistent with the hypothesis that at least part of the superior profitability of frequently revising analysts could be attributed to their superior skill in uncovering private information. We conclude that investors are better off, both in the short-run and in the long-run, following the advice of analysts who make revisions more frequently.