ارزش توصیه تحلیلی: شواهد بین المللی

We evaluate the value of analysts’ recommendations in the G7 countries. Stock prices react significantly to recommendation revisions in all countries except Italy. We find the largest price reactions around recommendation revisions and the largest post-revision price drift in the US. Neither differences in the timing of recommendation revisions relative to earnings announcements nor differences in industry coverage explain the superior performance of the US analysts’ recommendations. Tests within a subsample of ADRs indicate that the most likely explanation for the superior performance is that the US analysts are more skilled at identifying mispriced stocks than their foreign counterparts.

Stock analysts form an integral part of capital market operations, providing stock related research for their brokerage clients who use their research reports, earnings forecasts and stock recommendations for investment decisions. Recently, analysts’ roles as investment advisors have come under intense scrutiny. Media reports and some anecdotal evidence suggest that analysts compromise their objectivity and issue positive recommendations to curry favor with potential investment banking clients of their employers, rather than for the benefit of the investors who rely on their advice. A recent Fortune article, for instance, claims that the investment advice offered by analysts is “so dishonest and fraught with conflicts of interest that it has become worthless.’’1 Analyst recommendations do contain an element of bias towards being favorable. Earlier papers document that analysts rarely issue sell or strong sell recommendations. Jegadeesh et al. (2004) report that the average analyst rating over the 1985 to 1999 period is close to a buy recommendation and sell or strong sell recommendations make up less than five percent of all recommendations. Furthermore, Lin and McNichols (1998) and Michaely and Womack (1999) find that the analysts employed by the lead underwriters for new equity issues issue more favorable recommendations for those stocks than the other analysts who follow them.2 Extant literature finds that despite inherent biases analyst recommendations do add value. Stickel (1995) and Womack (1996) document that recommendation upgrades tend to outperform downgrades, and Barber et al. (2001), Jegadeesh et al. (2004), Boni and Womack(2003) and Green (2006) find that the stocks with the most favorable recommendations outperform the stocks with the least favorable recommendations. These findings indicate that investors can benefit from analysts’ recommendations if they consider the relative levels of recommendations across stocks, or if they pay attention to changes in recommendations. How much value can analysts potentially add? This is certainly a hard question to answer because there is no natural benchmark. In a perfectly efficient market, analysts would not be able to add any value because market prices already would reflect any information analysts might have. If analysts possess unique skills in collecting and analyzing value-relevant information, however, then they can add value, but its magnitude can only be empirically determined. Several papers have examined the value of analyst recommendations in the US to shed light on this issue. The recent controversy about the US analysts’ objectivity and the debate about the usefulness of their recommendations, however, indicate that our understanding of how much value analysts can add is far from complete. An examination of the value of analyst recommendations in other developed countries will give us a more comprehensive picture of the extent to which the unique skills of analysts are useful for investors. This paper examines analyst recommendations in the Group of Seven (G7) industrialized countries and evaluates their value for investors. While the value that analysts add has been extensively studied in the US, that value in the other countries has received scant attention in the literature. This paper fills that gap. The results here shed light on the importance of the role that analysts play in various markets. We first examine the distribution of analyst recommendation levels in each country in our sample. The literature generally views the level of recommendations as one indication of the conflict of interest analysts face (e.g., Lin and McNichols, 1998; Michaely and Womack 1999). We find that the frequencies of sell and strong sell recommendations are far fewer than the frequencies of the buys and strong buys in all countries. The frequency of sell recommendations is the lowest in the US. In fact, during our sample period, sell recommendations are about four to five times as frequent in the other countries as in the US. These results support the general notion that the analysts in the US face the largest conflicts of interest. Therefore, if conflicts of interest were a dominant factor in determining the value of analysts’ forecasts, then we would expect the value of analyst recommendations to be the lowest in the US. Our next tests examine the value of analyst recommendations in each country. We use two approaches to assess the value of recommendations. In the first approach, we examine the event time performance of analysts’ recommendation upgrades and downgrades. Specifically, we examine the returns of upgraded and downgraded stocks from the revision dates to six months after the revisions. This approach is similar to that in Womack (1996), and it allows us to determine the extent of mispricing that analysts can detect in various countries. The second approach analyzes the performance of calendar-time trading strategies that buy stocks with recommendation upgrades and sell stocks with recommendation downgrades. The calendar-time strategies are implementable in practice, and hence their performances provide an indication of the extent to which investors can profit by using analyst recommendation changes as inputs in their trading strategy. Earlier papers by Barber et al. (2001) and Jegadeesh et al. (2004) examine calendar time trading strategies based on the level of analysts’ recommendations. Specifically, the strategies buy the most favorably recommended quintile of stocks and sell the least favorably recommended quintile of stocks. Barber at al. (2001) and Jegadeesh et al. (2004) find that this strategy yields significant profits. Jegadeesh et al. (2004), however, show that recommendation levels-based strategy profits largely from the price and earnings momentum, which are correlated with recommendation levels, rather than from the recommendations per se.3Jegadeesh et al. (2004) also examine the performance of trading strategies based on analysts’ recommendation revisions and they find that the profitability of revisions-based strategies is not driven by momentum or any other stock characteristics. Since the incremental information content of recommendation revisions is stronger than that of recommendation levels, we examine the profitability of revisions-based calendar-time trading strategies. We examine the performance of various trading strategies, with different holding periods, and different levels of delays between the time when the revisions are made and the time when the stocks enter the portfolios. In both the event time analysis and the calendar time strategies, we find that the US analysts add the most value. The values analysts add in the other countries are about the same. In a sample of American Depositary Receipts (ADRs) that are covered by both US analysts and foreign analysts, we find larger price reactions following recommendation revisions by US analysts than by foreign analysts. Since both US analysts and foreign analysts follow the same set of stocks in this experiment, the higher value added by the US analysts cannot be due to any differences in efficiency across markets. We also find that revisions issued by large brokerages are related to larger price reactions, especially in the US. These findings indicate that the US analysts are more skilled than the analysts in other developed countries at identifying mispriced stocks, which more than offsets the effect of potential conflicts of interest. Most likely, the higher compensation that the analysts in the US receive attracts more skilled analysts, who in turn deliver more informative recommendations. An alternative possibility is that US analysts add more value because they time their revisions close to earnings announcements. Since analysts tend to upgrade stocks with positive earnings surprises and downgrade stocks with negative earnings surprises, a part of the value that the US analysts add could come from the post-earnings announcement drift, rather than from any value that they add through their skills.4 To test this alternative hypothesis, we exclude all recommendations that were issued around earnings announcement dates and find similar results. In additional tests, we also find that the differences in value of analyst recommendations across countries are not due to any inter-country differences in industry composition. Additionally, we examine the value of analysts’ recommendations for ADRs that are followed by both US analysts and by analysts in the ADR's domestic country. We find that the US analysts’ recommendations add more value than the recommendations of their foreign counterparts for this subsample of stocks as well. These findings indicate that the most likely explanation for the US analysts’ superiority is that they are more skilled than their foreign counterparts. We also examine trading volumes around recommendation revisions. We expect that the level of attention that investors pay to analyst recommendations would be related to the value they add. We use changes in trading volume around recommendation revisions as a measure of attention that analysts get in various countries. We find that the US stocks experience the largest increase in trading volume around recommendation revisions, which is consistent with our findings that analysts in the US add the most value. The rest of the paper is organized as follows. Section 2 describes our data sources and our sample. Section 3 compares the level of recommendations across the countries we examine. Section 4 presents the event-time analysis of the performance of analyst recommendations. Section 5 investigates the source of the superior performance of US analysts. Section 6 examines the cross-sectional relation between the brokerage size and analysts’ recommendations. Section 7 examines the pattern of trading volume around recommendation revisions. Section 8 evaluates the profitability of calendar-time trading strategies and Section 9 presents our conclusions.

This paper examines analyst recommendations in the G7 countries and evaluates the value of these recommendations. We find that stock prices react significantly to recommendation revisions on the day of recommendation and on the following day in all of these countries except Italy. Stock prices continue to drift up for upgrades and down for downgrades over the next two to six months. We find the largest price drifts in the US, followed by Japan. These are the two largest markets in the world, and our evidence indicates that the value of analyst recommendations is the largest in these countries. Additionally, we examine the value of analysts’ recommendations for ADRs that are followed by both US analysts and foreign analysts. US analysts’ recommendations add more value than the recommendations by analysts in other countries for this subsample of stocks as well. These findings indicate that the US analysts add more value because of their skills, rather than because the US market is less informationally efficient than the other markets. We test an alternative possibility that US analysts time their revisions close to earnings announcements and find no support for this view. We also examine the performance of calendar-time trading strategies that buy upgraded stocks and sell downgraded stocks. We consider several strategies with different holding periods, and with different delays between the time when revisions are made and the time when the stocks enter the portfolios. The equal-weighted strategies with no delays and with holding periods of one month are profitable in all countries, except Italy, before transaction costs. The value-weighted strategies, however, are profitable only for a one-month holding period strategy that is initiated without a delay. All these strategies are more profitable for small firms than for large firms. Among the G7 countries, we find that the trading strategies are the most profitable in the US, across all horizons. Our evidence indicates that the market is not semi-strong form efficient in the countries in our sample, despite the fact that these countries have the most developed stock markets in the world. Although the profits from these trading strategies are probably smaller than reasonable transaction cost bounds considered by Barber et al. (2001), it is possible that investors can profitably use recommendation revisions in combination with other signals in their trading strategies. For instance, Jegadeesh et al. (2004) show that in the US, recommendation revisions in combination with several momentum and value signals earn significant profits. Such strategies may work in the other countries as well. However, the US results may not carry over to other countries because the analysts outside the US add less value. We leave the analysis of the incremental value of recommendations in other countries for future research. Our comparative analysis of the value of analyst recommendation in developed countries also provides a basis for assessing the likely benefits to investors of the recent regulatory settlement with brokerage firms. This settlement includes provisions to restrict links between the investment banking and the stock research departments, and to promote independent research. The settlement will likely promote more ethical practices, and make analysts less reluctant to issue sell recommendations than they have been in the past. In fact, we do find that sell recommendations are more common in the post-settlement period.29 The recommendations levels also exhibit more cross-sectional variation in 2003 than in the earlier years. Neither the frequency of sell recommendations nor the cross-sectional variability, however, need translate into more value for investors. During our sample period, the US had the lowest frequency of sell recommendation among all countries we examine. Also, the issue of analysts’ conflict of interest has not been a major public concern outside the US. However, analyst recommendations in the US provide the most value for investors. It seems likely that the US investors were able to see through potential conflicts of interest, even prior to the settlement. Furthermore, the evidence that analysts in all of the most developed countries add only a modest amount of value through their recommendations suggests that these markets are fairly efficient, and it is unlikely that analysts can routinely uncover larger mispricings. Therefore, we do not see any reason to expect that the recommendations of the US analysts will perform any better in the future, even with the current efforts to remove their conflicting incentives, than they have in the past. In fact, we would caution limiting analysts’ role in investment banking business. Brokerage houses pay analysts for the value they add to the firm. If analysts’ role in the investment banking arena were restricted, then their compensations would likely decline in the future.30 As a consequence, tighter restrictions on analysts’ activities could lead to a decline in the skill level of analysts, and thereby adversely affect the value of their recommendations.