By LARRY SWEDROE
A large body of evidence demonstrates that the ability of active managers to generate alpha (outperform risk-adjusted benchmarks) has been rapidly declining. For example, Mike Sebastian and Sudhakar Attaluri, authors of the 2014 study Conviction in Equity Investing, found that the percentage of skilled managers was 20 percent in 1993 but had fallen to just 1.6 percent by 2011. This result matches closely the findings of the 2010 study Luck versus Skill in the Cross-Section of Mutual Fund Returns. The authors, Eugene Fama and Kenneth French, found that only managers in the 98th and 99th percentiles showed evidence of statistically significant skill.
In our 2015 book, The Incredible Shrinking Alpha, Andrew Berkin and I explain how four major themes which have been conspiring against the ability to generate alpha. These are academic research converting what was once alpha into beta (or loading on a common factor), the shrinking supply of victims that can be exploited, tougher competition, and the increasing supply of capital chasing alpha.
Lubos Pastor, Robert Stambaugh and Lucian Taylor, authors of the 2015 paper Scale and Skill in Active Management, provided a good example of why the hurdles to generating alpha have been increasing even though today’s managers are more skilled. “We find that the average fund’s skill has increased substantially over time, from -5 basis points (bp) per month in 1979 to +13 bp per month in 2011.”
However, they also found that the higher skill level has not translated into better performance. They reconcile the upward trend in skill with no trend in performance. They noted: “Growing industry size makes it harder for fund managers to outperform despite their improving skill. The active management industry today is bigger and more competitive than it was 30 years ago, so it takes more skill just to keep up with the rest of the pack.”
Sung Jun Park and Ki Young Park provide further evidence that the hurdles to generating alpha are increasing with their September 2019 study, Is the Predictive Value of Analysts’ Recommendations in Decline?. Their data sample covered 719,830 analyst recommendations from 1994 to 2017. They constructed various portfolios based on levels and changes in analyst recommendations and examined how the value of those recommendations in predicting the abnormal stock returns has changed over time.
They began by discussing prior research, including the 2001 study Can Investors Profit from the Prophets? Security Analyst Recommendations and Stock Returns and the 2010 study Ratings Changes, Ratings Levels, and the Predictive Value of Analysts’ Recommendations. They had found that a strategy of purchasing stocks with most favorable consensus recommendations and shorting stocks with the least favorable ones yielded abnormal gross returns.
In addition to examining the value of analyst recommendations, they also sought to determine the impact of the implementation in 2002 of NASD (National Association of Securities Dealers) Rule 2711, mandating that sell-side analysts should disclose the distribution of their security recommendations. They found that before the implementation, “the shares of strong buy, buy, and hold recommendations are around 30% each, but right after the implementation, the share of neutral recommendations goes up to around 45%, while the shares of both strong buy and buy recommendations fall to around 20%. The share of strong buy recommendations even decreases to less than 20% from 2010, while that of buy recommendations increases to around 30%.
Interestingly, the shares of strong sell and sell recommendations are quite small and close to zero before the implementation, but they rise after the implementation, although they are still less than 10% each.” This led them to conclude that “in terms of stock returns, the Rule 2711 has mitigated the analysts’ conflicts of interest in stock recommendations.” They also found that “the positive relationship between stock market performance and analyst recommendation becomes stronger, especially for buy-related ones.”
The value of recommendations
Park and Park examined the trends of abnormal returns of five portfolios formed on analyst recommendations. The first set of portfolios is based on the level of recommendations: strong buy, buy, hold, sell, and strong sell. The second set is constructed based on ratings change: upgrade, reiterate/initiate, and downgrades. The third set is based on the magnitude of upgrade and downgrade: single or double upgrades, and single or double downgrades. The fourth set reflects both the level and magnitude of change. The set consists of single (double) upgrades to buy or strong buy, and single (double) downgrades to sell or strong sell. The fifth set considers the long-short portfolios. To determine if returns were abnormal, they used the Carhart four-factor (beta, size, value and momentum) model.
The following is a summary of their findings:
- Over the full period, the buy and sell recommendations in the four long or short portfolios show statistically significant (at the 5 percent confidence level) abnormal returns. However, once transactions costs are considered, the abnormal returns are close to zero and are no longer statistically significant.
- The pattern of abnormal returns begins to decline in the early 2000s.
- In the 2010 through 2017 period, the alphas for the four buy and sell portfolios are all negative (even before transactions costs), though not statistically significant. However, the alphas for the long-short portfolios were positive and statistically significant, though their statistical significance declines once transaction costs are considered.
- The decline in the predictive value of recommendations results from the loss of its value in buy- and upgrade-related recommendations, while sell- and downgrade-related ones are still associated with the negative abnormal returns. In other words, buy recommendations and upgraded recommendations (such as from buy to strong buy) no longer have any predictive value, while sell recommendations and downgrades still maintain value (at least before transactions costs).
The difference in outcomes between the buy and sell portfolios is likely explained by the fact that while it is easy to correct undervaluation (you simply buy the stock), it is more difficult and expensive to correct overvaluation because of the need and expense of shorting, as well as the risk of unlimited losses. In addition, the charters of many institutional investors prohibit shorting. These limits to arbitrage allow anomalies to persist.
The evidence makes clear that despite the fact that today’s security analysts/money managers are far more skilled than their predecessors, there are forces at work that are making the market ever more efficient, increasing the hurdles to generate statistically significant alpha. The evidence is so strong that despite the implementation of NASD Rule 2711, which reduces the agency biases inherent in analyst recommendations, the predictive value of the buy and upgrade recommendations have moved from positive to negative. On the other hand, the evidence shows that the sell and downgrade recommendations still have predictive value, though limits to arbitrage (including transaction costs) might limit the ability to exploit their value.
The bottom line is that the publication of research findings such as those discussed above are fuelling the trend toward passive/ index investing. One result of that trend is that since it’s the poorest performers (and therefore the weakest competition) who are most likely to abandon the game of active management, the pool of victims needed to exploit in order to generate alpha (which is a negative-sum game after expenses) is persistently shrinking, and the average level of skill is rising. And since in “games” of skill it’s not the absolute level of skill that matters but the relative skill level of the competition, the hurdles to generating alpha are persistently rising.
LARRY SWEDROE is Chief Research Officer at Buckingham Strategic Wealth and the author of 17 books on investing, including Think, Act, and Invest Like Warren Buffett.
If you’re interested in reading more of his work, here are his other most recent articles for TEBI: