Individual Stock Investing Increases Risk

The evidence

Hendrik Bessembinder contributed to our understanding of the risky nature of individual stocks with his study “Do Stocks Outperform Treasury Bills?” which appeared in the September 2018 issue of the Journal of Financial Economics. His study covered the period 1926 through 2015 and included all common stocks listed on the NYSE, Amex and NASDAQ exchanges. Returns are inclusive of dividends and thus assume investors reinvest dividends in the stocks that paid them.

Following is a summary of his findings:

• At the annual horizon, only 47.7% of returns are larger than the one-month Treasury rate.
• Even at the decade horizon, a minority of stocks outperform Treasury bills.
• From the beginning of sample, or first appearance in the data, through the end of sample or delisting, and including delisting returns when appropriate, just 42.1% of common stocks have a holding period return greater than one-month Treasury bills.
• A minority (49.2%) of common stocks have a positive lifetime holding period return, and the median lifetime return is -3.7%.
• Despite the existence of a small-cap premium (an annual average of 2.8%), just 37.4% of small stocks have holding period returns that exceed those of the one-month Treasury bill. In contrast, in the largest decile, 69.6% outperform the one-month Treasury bill.
• Reflective of the positive skewness in returns, only 599 stocks, just 2.3% of the total, have lifetime holding period returns that exceed the cross-sectional mean lifetime return.
• The median time that a stock is listed on the Center for Research in Security Prices’ (CRSP) database is just more than seven years.
• A single-stock strategy underperformed the value-weighted market in 96% of bootstrap simulations (a test that relies on random sampling with replacement) and underperformed the equal-weighted market in 99% of the simulations.
• The single-stock strategy outperformed the one-month Treasury bill in only 28% of the simulations.
• Only 3.8% of single-stock strategies produced a holding period return greater than the value-weighted market, and only 1.2% beat the equal-weighted market over the full 90-year horizon.

Bessembinder’s findings highlight the high degree of positive skewness (lottery-like distributions), and the riskiness, found in individual stock returns. He noted that the 86 top-performing stocks, less than one-third of 1% of the total, collectively accounted for more than half of the wealth creation. And the 1,000 top-performing stocks, less than 4% of the total, accounted for all of the wealth creation. The other 96% of stocks just matched the return of riskless one-month Treasury bills!

The implication is striking: While there has been a large equity risk premium available to investors, a large majority of stocks have negative risk premiums. This finding demonstrates just how great the uncompensated risk is that investors who buy individual stocks, or a small number of them, accept – a risk that can be diversified away without reducing expected returns.

Bessembinder concluded that his results help explain why active strategies, which tend to be poorly diversified, most often lead to underperformance. “At the same time,” he wrote, “the results potentially justify a focus on less-diversified portfolios by investors who particularly value the possibility of ‘lottery-like’ outcomes, despite the knowledge that the poorly diversified portfolio will most likely underperform.” Bessembinder showed the impact of the preference for lottery tickets with this finding: “Only 31.5% of monthly returns to stocks in the lowest share price decile exceed one-month Treasury bill rates, as compared to 59.1% of monthly returns to stocks in the highest share price decile.”

Bessembinder, with co-authors Te-Feng Chen, Goeun Choi and K.C. John Wei, extended his research to global markets in the July 2019 study “Do Global Stocks Outperform US Treasury Bills?” The authors examined compound return data for nearly 62,000 global common stocks traded on public markets over the period 1990 to 2018. Their key finding was: “While the cross-sectional mean stock return is indeed positive in all 42 countries we examine, returns to the majority [all but seven] of global stocks fall short of returns to one-month US Treasury bills over matched time horizons.” Specifically, they found: “Only 40.5% of global common stocks, including 43.7% of US stocks and 39.3% of non-US stocks, have full-sample buy-and-hold return that exceeds the accumulated return to one-month US Treasury Bills over matched time horizons.” And “In only two countries, Switzerland (50.1%) and Colombia (58.1%), did over half of the individual stocks outperform the value-weighted average stock return.” The skewness of returns was so extreme, they found that the best performing 811 firms (just 1.33% of the total) accounted for all net global wealth creation. Outside the U.S., the data was even more extreme, as less than 1% of non-U.S. firms accounted for all net stock market wealth creation.

Another interesting finding was that while “the most frequently observed annual returns are clustered in the vicinity of zero … the most frequently observed returns for both US and non-US stocks at the decade horizon (rounded to 5%) are -95% and -100%.”

Another important finding was on the extreme degree of skewness in individual stock returns. The skewness of a normal distribution is zero. For global stocks the skewness at annual horizons is in excess of 18, and at the decade horizon in excess of 68! Reflecting the positive skewness, the authors noted, “Full-sample median returns were just 1.27% at the annual horizon, and ‑2.16% at the decade horizon.”

Let’s examine some further evidence on the odds of picking a stock that outperforms a broad market index.

Further evidence

In the decade of the 1990s, 22% of the 2,397 stocks still in existence at the end of the period actually had negative absolute returns. And this does not include survivorship bias – a significant percentage of stocks delist every year and thus would not have been considered in the data. If they had been, the 22% figure would have been much higher. This occurred during a decade in which the S&P 500 had an annualized return of 18.2%. Another interesting statistic relates to the companies that comprise the S&P 500. In their 2001 book, Creative Destruction, authors Richard Foster and Sarah Kaplan noted that of the 500 companies originally in the index in 1957, only 74 remained in 1998, and only 12 of those outperformed the index itself. This means that approximately 85% of the companies in the index in 1957 were eventually replaced or merged out of existence.

We also find startling results from the study by Longboard Asset Management, The Capitalism Distribution, covering the period 1983 through 2006 and the top 3,000 stocks. The authors found that while the Russell 3000 Index provided an annualized return of 12.8% and a cumulative return of 1,694%:

• The median annualized return was just 5.1%, 7.7 percentage points below the return of the market.
• The average (mean) annualized return was -1.1%.
• 39% of stocks lost money (even before inflation) during the period.
• 19% of stocks lost at least 75% of their value (again, before considering inflation)!
• 64% of stocks underperformed the Russell 3000 Index.
• Just 25% of stocks were responsible for all the market’s gains.

We’ll now look at the findings from a recent study by Vanguard.

Vanguard weighs in

Chris Tidmore, Francis M. Kinniry Jr., Giulio Renzi-Ricci and Edoardo Cilla of Vanguard’s research team examined the issue of concentrating stock risk in a small number of stocks in their February 2019 study, “How to Increase the Odds of Owning the Few Stocks that Drive Returns.” They began by noting: “Many investors and financial professionals believe that broadly diversified portfolios are inferior to concentrated portfolios made up of a manager’s ‘best ideas.’” Their dataset encompassed the universe of stocks of the Russell 3000 Index from January 1987 through December 2017. They assumed that a hypothetical investor would invest her/his capital over the entire data sample period. They created time series of portfolio performance as follows:

1. At the beginning of each quarterly rebalancing period, randomly select stocks to form portfolios holding n stocks, where n equals 1, 5, 10, 15, 30, 50, 100, 200, or 500. Each stock must exist at the beginning of the quarter and has an equal probability of selection. Each portfolio is created independently (so the same stock could appear in more than one portfolio).

2. After selecting the stocks, construct the portfolio by equally weighting them. Then compare each portfolio’s performance with that of the equal-weighted benchmark and check whether the portfolio outperformed or underperformed the benchmark.

3. For each portfolio, compute the excess return and tracking error versus the benchmark.

4. For each portfolio size, conduct the above three steps 10,000 times and compute average risk and return summary statistics.

Following is a summary of their findings:

• The simulated portfolios’ probability of outperforming the benchmark rises monotonically with an increase in holdings – from about 11% with one stock to about 48% with 500 stocks. While holding more stocks increases the odds of beating the Russell 3000, even at 500 stocks the odds are less than 50%.

• The dispersion of the average conditional excess return shrinks with an increase in holdings. For example, with a single stock portfolio, the average outperformance was about 4%, but average underperformance was much larger at about -12%. By the time we reach even 15 stocks, the average outperformance was only 1.5%, while the average underperformance was -2.0%. At 500 stocks, both the average outperformance and underperformance was just ± 0.3%.

• Increasing the number of holdings decreases the average tracking error.

The authors noted that a highly concentrated active fund with only 30 holdings must have enough skill to overcome the expected negative performance gap of -0.4% they found in their simulations. But that gap isn’t the only hurdle. In addition, they must have enough skill to overcome not only the fund’s higher expense ratio (relative to a low-cost index fund) but also their generally higher trading costs.

Vanguard also examined the performance of actively managed funds over the same period and found that “Given any specified level of sector or factor concentration, a fund’s performance improves with a higher number of holdings.”

They concluded: “We find that higher numbers of randomly selected portfolio holdings are associated with increased chances of outperforming the benchmark, higher average excess returns, less-negative excess returns, and lower tracking error. These findings make a compelling case for broadly diversified active portfolios.”

Summary

The results from the studies we have examined highlight the important role of portfolio diversification. Diversification has been said to be the only free lunch in investing. Unfortunately, most investors fail to use the full buffet available to them. That is likely because they don’t recognize that poorly diversified portfolios are likely to underperform because they omit the relatively few stocks that generate large positive returns.

The results also explain why active portfolio strategies, which tend to be poorly diversified, most often underperform their benchmarks. Underperformance is typically attributed to transaction costs, fees and/or behavioral biases that amount to negative skill. The results here show that underperformance can be anticipated more often than not for active managers with poorly diversified portfolios, even in the absence of costs, fees or inadequate skill.

Takeaways

The studies we have examined are important in the debate regarding active stock selection versus the passive holding of broadly diversified stock indexes. Their results clearly demonstrate that the wealth created by stock market investing is largely attributable to outstanding outcomes from a relatively few stocks. The evidence, such as from the annual SPIVA Active Versus Passive Scorecards, demonstrates that very few active fund managers are able to identify the few big winners ahead of time, making active management the loser’s game. That begs the question: Why do investors play the loser’s game, taking on idiosyncratic, diversifiable, uncompensated risks in the face of all the evidence? The field of behavioral finance provides us with many answers. For individual investors, behavioral problems compound the problem.

Behavioral errors compound the problem

Compounding the problem of poorly diversified portfolios is that we know from a series of papers by Brad Barber and Terrance Odean that individual investors are poor stock-pickers. For example, they found that the stocks both men and women buy trail the market after they buy them, and the stocks they sell outperform after they are sold. Yet, they persist in the effort. Why? The following is a brief list of some of the reasons:

• The majority of investors have not studied financial economics, read financial economic journals or read books on modern portfolio theory. Thus, they don’t have an understanding of how many stocks are required to build a truly diversified portfolio. Similarly, they don’t have an understanding of the difference between compensated and uncompensated risk. The result is that most investors hold portfolios with assets concentrated in relatively few holdings.

• Richard Thaler of the University of Chicago and Robert Shiller, two Nobel Prize-winning economics professors, noted that “Individual investors and money managers persist in their belief that they are endowed with more and better information than others, and that they can profit by picking stocks.” This insight helps explain why individual investors don’t diversify: They believe they can pick stocks that will outperform the market. Overconfidence is an all-too-human trait.

• People make investment decisions based on what they believe is important, or what economists call “value relevant” information. They virtually never consider that others, with far more resources than they have, almost certainly have the same information. Thus, that information must already be “baked into” prices. Mark Rubenstein, a professor of applied investments at the University of California, Berkeley, put it this way: “One of the lessons of modern financial economics is that an investor must take care to consider the vast amount of information already impounded in a price before making a bet based on information.” Legendary investor Bernard Baruch put it more succinctly, stating, “Something that everyone knows isn’t worth knowing.” The failure to understand this leads to a false sense of confidence, which in turn leads to a lack of diversification.

• Investors have the false perception that by limiting the number of stocks they hold, they can manage their risks better.

• Investors gain a false sense of control over the outcomes by being involved in the process. They fail to understand that it is the portfolio’s asset allocation that determines risk, not who is controlling the switch.

• Investors confuse the familiar with the safe. They believe that because they are familiar with a company, it must be a safer investment than one with which they are unfamiliar. This leads them to concentrate their holdings in a few companies with which they are familiar, essentially Peter Lynch’s advice to “buy what you know.” Unfortunately, a study found that the returns of local stocks investors purchased badly lagged the returns of local stocks sold.

If you have made any of these mistakes, you should do what all smart people do: Once they have learned that a behavior is a mistake, they correct their behavior.

Conclusion

Individual stock ownership offers both the hope of great returns (finding the next Google, for instance) and the potential for disastrous results (ending up with the next Lehman). Concentrated portfolios not only provide the possibility of outperformance, they increase the possibility of large outperformance. However, they also increase the odds of extreme losses by even larger amounts. The more concentrated the portfolio, the greater the risk. And the market doesn’t compensate investors with higher expected returns for taking risks that are easily diversified away.

Because investors are not compensated for taking the risk that their result will be below the benchmark, the rational strategy is not to buy individual stocks – just like at the roulette wheels, the craps tables and the slot machines in Las Vegas casinos, the surest way to win a loser’s game is to not play. Unfortunately, the evidence is that the average investor, while being risk averse, doesn’t act that way. In a triumph of hope over wisdom and experience, they fail to diversify.

Larry Swedroe is the director of research for The BAM Alliance, a community of more than 130 independent registered investment advisors throughout the country.

More Active Management Topics >