Less liquid assets are not usually more volatile, so they are not riskier by that measure. 18 Less liquid investments may involve the risk, however, that one cannot sell (before the investment matures, if it ever does) without suffering delays or being forced to make substantial price concessions. Thus, liquidity is very popular and likely to remain so. All else being equal, high liquidity is more expensive than low liquidity. Low liquidity lowers valuations but raises expected returns.
Strategies that involve buying stocks that are too popular should have lower returns than strategies that buy less popular stocks. By its nature, a strategy that focuses on buying the less popular stocks will be contrarian. Investors will have to go against the crowd.
The liquidity effect illustrates that once we go beyond the risk–return paradigm, we can better understand the various anomalies and premiums found in the marketplace. Investors can have reasons to dislike an asset other than it being too risky. They can also have reasons to like an asset other than it having low risk. We associate liking with popularity and disliking with unpopularity.
Now consider the generally accepted premiums in asset prices. 19 These premiums are considered systematic; that is, they are more or less permanent in nature. Each of the premiums is associated with something that investors do not like. Once they are discovered, many anomalies are no longer priced. Premiums remain priced, however, because investors might not like characteristics even if they understand that they are paying for them. In such a case, these premiums continue to exist in equilibrium.
Size Premium. Banz (1981) identified the size or small-cap premium . Like the equity premium, the size premium may be directly related to risk because, on average, small-cap companies are less stable and have more volatile stock prices than large-cap companies. In addition to having higher risk, small-cap stocks are usually less liquid, are less well covered by analysts (have higher research costs), and have lower investment capacity than large-cap stocks. If one thinks of the amount invested as a relative popularity vote, then by definition, small-cap stocks are unpopular relative to large-cap stocks.
Following this logic, mega-cap companies are more popular than large-cap companies and so on until small-cap companies are more popular than microcap companies. In Figure 2.3 , we reproduce an exhibit from Ibbotson (2018) that shows a monotonic relationship between size decile and realized return: The decile of largest companies (#1) has the lowest arithmetic average annual realized return and systematic risk (beta), and as we move along the size decile spectrum from #1 to #10, we see average return and systematic risk steadily increase.
Figure 2.3. Security Market Line and Scatterplot of Arithmetic Average Total Return of the CRSP/NYSE/NASDAQ Size Deciles, 1926–2017
Note: CRSP is the Center for Research in Security Prices.
Source: Ibbotson (2018) .
______________________________
Value Premium. One of the original and most significant cracks in the CAPM and risk–return paradigm came with the publication of work by Basu (1977) that documented the “value effect.” Basu found that portfolios of value stocks (stocks with low price–earnings ratios) dramatically outperform portfolios of growth stocks (stocks with high price–earnings ratios), with less risk. 20 Thus, calling a return premium a risk premium is not necessarily correct. Subsequent authors have documented the value premium in various time periods and markets and by using various measures of value.
Academicians continue to debate potential explanations for the value premium. Lakonishok, Shleifer, and Vishny (1994) argued that the value premium (or “glam[orous] stock discount”) is the result of systematic suboptimal behavior by typical investors springing from their consistent overconfidence about, and overestimation of, the future growth of earnings and cash flows. Anginer and Statman (2010) offered evidence that the names of some companies elicit a positive reaction, leading investors to wrongly associate them with higher expected future returns.
Most of these explanations are consistent with the popularity perspective: If growth companies are popular and value companies are unpopular, for whatever reason, then growth stocks will be relatively higher priced relative to value stocks. If this popularity is somewhat permanent, we can think of these as premiums rather than as mispricing.
Liquidity Premium. Liquidity does not require a behavioral explanation, but as noted previously, it is consistent with popularity. Rational investors want more liquidity, so investors with longer horizons who do not seek liquidity in the short run earn a liquidity premium. Liquidity has a number of potential measures that are likely to capture somewhat different characteristics. Using bid–ask spreads to proxy liquidity, Amihud and Mendelson (1986) documented that less liquid stocks outperform more liquid stocks. Haugen and Baker (1996) and Datar, Naik, and Radcliffe (1998) showed that low-turnover stocks earn higher future returns than high-turnover stocks.
Ibbotson et al. (2013) demonstrated that investing in the less liquid securities could result in a premium that is at least comparable to the size and value premiums. Building on stock-level liquidity premiums, Idzorek et al. (2012) , as noted earlier, found that after they controlled for other characteristics, mutual funds that held the less liquid stocks outperformed, net of fees. From a popularity perspective, all else being equal, investors clearly prefer greater liquidity. However, there is a tight relationship between liquidity and size. Again, securities with the unpopular characteristic have outperformed those with the popular characteristic.
Severe Downside-Risk Premium. The severe downside-risk premium is consistent with the idea of popularity. Kraus and Litzenberger (1976) demonstrated that because investors do not like stocks with severe downside risk (negative coskewness), these stocks have higher returns. This phenomenon is directly related to prospect theory posited by Kahneman and Tversky (1979) . In this application of the theory, losses hurt far more than equivalent gains help, particularly when assessed relative to an investor’s starting position. Harvey and Siddique (2000) showed that coskewness (one measure of severe downside risk) has earned a significant premium, averaging approximately 3.6% per year for US stocks.
Low-Volatility and Low-Beta Anomalies. Clarke, de Silva, and Thorley (2011) , Baker and Haugen (2012) , Ang (2014) , Blitz and van Vliet (2007) , Frazzini and Pedersen (2014) , and others have contributed to the growing empirical documentation of the low-volatility and low-beta anomalies. Haugen and Heins (1975) first documented that high risk (volatility and systematic risk) does not necessarily lead to high returns within the universe of equities.
Possibly, restrictions on leverage or a general aversion to leverage, as put forth initially by Black (1972) and more recently by Asness et al. (2012) , reverse the expected relationship between risk and return. 21 Also, Baker, Bradley, and Wurgler (2011) suggested that institutional active equity fund managers intentionally seek a beta greater than 1.0 in hopes of frequently outperforming their benchmarks/peer groups, even if this strategy ends up hurting their long-term beta-adjusted performance (alpha). Similarly, we argue that active equity managers, in their quest for outperformance (primarily in absolute terms—that is, not risk adjusted), purposely seek out holdings with high volatility. Hence, both high-beta stocks and high-volatility stocks are the most popular stocks; thus, both low-beta stocks and low-volatility stocks are the least popular stocks, leading to their higher relative returns.
In most cases in which an investor has decided to invest with a particular equity investment manager (mutual fund), the investor has knowingly signed up for a risky investment—presumably, in hopes of realizing an expected return premium. And if the investors are paying for active management, they are also hoping for better-than-market performance. For reasons that are beyond the scope of this book, in this context, investors seem to care more about absolute return and return relative to peer groups than they care about risk-adjusted return.
The typical active investment manager knows that the way to gather more assets is to outperform, in absolute te
rms, similar competing investment managers. In a quest to outperform on an absolute basis in a world in which leverage is typically off the table, this quest creates a supersized demand for high-beta stocks. In an odd equilibrium sense, this reduces the realized returns of high-beta stocks to a level lower than that which would be predicted by the CAPM so that the equities in question will unequivocally underperform on a risk-adjusted basis but may be able to slightly outperform on an absolute basis; hence, the higher demand.
The Momentum Anomaly. Consistent with the idea of mispricing, past winners seem to continue to win for a period of time. Such momentum should not be thought of as a premium but, rather, as characterizing a transition period during which a change in valuation is slowly (and somewhat predictably) taking place. The change can be caused by either changing popularity or a delayed reaction to changing fundamentals, but in either case, a behavioral model is needed to account for the momentum anomaly.
For example, momentum may be the result of an attention-causing event that creates more interest in a stock and/or an increase in trading activity. An increase in trading activity, and thus liquidity, coincides with a price increase, an increase in market capitalization, and probably more attention to the stock, increased liquidity, and further price increases—all part of a temporary, but unsustainable, virtuous circle. Should the price exceed that which is justified by the company’s fundamentals, the stock in question peaks in popularity. The degree to which the security is mispriced is noticed by more and more market participants, including short sellers. Then, the momentum ends. The price stagnates and eventually begins to fall. A vicious circle starts that eventually, although it may temporarily overshoot, returns the price to a more “reasonable” level.
ESG Premiums and Discounts. We use the term “environmental, social, and governance” (ESG) as an overall bucket that includes the stocks of companies that score well on ESG criteria, includes the stocks of companies that pass socially responsible investing (SRI) screens, and excludes so-called sin stocks. This bucketing limits the scope of our analysis, because clearly ESG is a multifaceted area worthy of further analysis elsewhere. Sin stocks are offered by companies/industries that people tend to dislike—for example, companies associated with tobacco, alcohol, firearms, and weapons. SRI seeks to explicitly avoid such sin stocks, and ESG-oriented investing usually strikes a balance by rating investments on the basis of various SRI attributes and showing preference for the higher rated stocks. In contrast with SRI, ESG investing does not avoid entire industries but tilts toward companies rated high on the ESG scales.
Relating specifically to sin stocks, Hong and Kacperczyk (2009) studied returns from 1965 to 2006 and found that sin stocks produced a substantial annual alpha slightly greater than 300 basis points. They argued that sin stocks attract fewer institutional investors and less analyst coverage. Return patterns among ESG investing is mixed (see Statman and Glushkov 2011 ). As with our explanation based on popularity, behavioral economists attribute the sin stock premium to affect, or judgments about good or bad feelings, experienced in relation to the investment in question.
Competitive Advantage, Brand, and Company Reputation. Given popularity’s ability to explain most of the well-known premiums and anomalies, we have been searching for characteristics that align with an intuitive definition of popularity to see if we can find further evidence to support or contradict the popularity framework. In Chapter 6 , we present our findings for three dimensions of popularity: sustainable competitive advantage, brand, and reputation. (Of these, company reputation is arguably closely linked to ESG.) Overall, we found that quartiles containing the least popular stocks—as represented by low/no sustainable competitive advantage, relatively low brand power, and relatively low company reputation—nearly monotonically produced returns superior to ensuing quartiles of more popular stocks.
Lack of attention is a behavior that probably contributes to the popularity premiums associated with competitive advantage, brand, and company reputation. Barber and Odean (2008) found that investors often limit their opportunity set to stocks that have caught their attention. They concluded that the “utility of an alternative is affected by how many agents choose that alternative. Thus, the attention attracting qualities of an alternative may indirectly detract from its utility” (pp. 812–813). Growth stocks appear to generate more attention than value stocks.
Exhibit 2.1 summarizes the various premiums/anomalies and the respective summary popularity-based explanations that we discussed in this chapter.
Other Premiums and Anomalies. Many more premiums exist than just the few we listed in this chapter. In fact, so many have been identified that Cochrane (2011) refers to a “zoo of new factors” (p. 1047). Green, Hand, and Zhang (2017) listed almost 100 potential factors, and Harvey, Liu, and Zhu (2016) listed more than 300 factors identified in the financial literature as providing high returns. Of course, the authors point out that the factors are cross-correlated and can be categorized into subgroups. Also, many appear to arise from after-the-fact data mining, so the hurdle rate for accepting them as premiums should be high.
In this chapter, we focused on the premiums that are generally recognized by the academic community. One could reasonably question our selected set, because they are primarily the premiums that were discovered the earliest rather than the set of premiums that have the most statistical significance. Our most important criterion for inclusion, however, was that each premium have some economic explanation. For us, this explanation is related to popularity.
Exhibit 2.1. Popularity-Based Explanations of Premiums and Anomalies
Premium/Anomaly Characteristic or Dimension of Popularity
Popularity-Based Explanation
Equity premium
Stocks are riskier than safe assets. Risk is unpopular.
Size
Small-cap stocks are riskier than large-cap stocks. They are also less liquid, less well covered, and have lower investment capacity.
Value
Value stocks are often out of favor (unpopular), are less well-known, and/or operate in the less glamorous industries.
Liquidity
Investors prefer more liquidity to less.
Severe downside risk
Investors dislike large losses.
Low volatility/beta
Active managers prefer high-beta stocks in hopes of outperforming benchmarks.
Momentum
Attention-causing events create interest, which increases trading activity and liquidity and results in an unsustainable virtuous mispricing circle.
ESG
Investors tend to avoid sin stocks and seek out responsible investments.
Competitive advantage, brand, and reputation
Stocks of companies with desirable attributes—competitive advantage, brand power, or company reputation—are sought out beyond their economic benefits.
Source: Based on Exhibit 1 in Idzorek and Ibbotson (2017) .
______________________________
Conclusion
In this chapter, we examined some of the well-known premiums and anomalies and found that all are consistent with the direction predicted by popularity: Those that embody a more-or-less permanent unpopular characteristic have been rewarded with a popularity premium. Characteristics that rapidly change in popularity can result in mispricing, especially to the extent that the popularity changes are predictable.
Except for the momentum and low-beta/low-volatility anomalies, most of the premiums we examined are thought by some researchers to be consistent with efficient markets, probably because they have been considered risk premiums. What is becoming increasingly clear from empirical results, however, is that many of the premiums are not associated with extra risk and, in some cases, are associated with a risk reduction. Thus, we need popularity to explain not only the premiums but also many of the anomalies that we observe in capital markets. In the following chapters, we show how popularity does so.
Appendix A. Psyc
hic Returns in Art Markets
Popularity is a social phenomenon associated with being admired, sought after, well-known, and so forth. Art collecting and investing can shed some light on the popularity hypothesis. What drives passionate collectors or art investors is the individual interpretation of artwork. Unlike stocks, art has no balance sheet, cash flow, or earnings to help determine its value. The valuation of art is almost entirely subjective.
Baumol (1986) calculated the returns in the art market over three centuries (1650–1960) and found that the average annual real (inflation-adjusted) return was about 0.55%. Thus, he called investing in art a “floating crap game” (p. 10). He reported that the real return on bonds over the study period was about 2.5%. Given that art as an investment provides much lower returns than stocks and bonds, in line with the equilibrium approach and the no arbitrage conditions in financial economics, Baumol’s findings suggest that the difference (1.95 percentage points) must be attributed to nonpecuniary returns or the utility derived from the aesthetic pleasure in art investments. In other words, art owners are willing losers from a financial return perspective because they derive utility from art. From this perspective, even though the investors own the art, it is almost as if they are paying a rental fee.
Spaenjers, Goetzmann, and Mamonova (2015) argued that to understand the economics of the market for art, one needs to examine the formation of art prices on a disaggregated level. Each individual piece of artwork gives rise to a market for trading in its private-value benefits (or nonfinancial utility). Within this framework, they discussed recent theoretical and empirical studies of the various forces driving the willingness to pay of bidders at art auctions. Emerging conclusions of this body of work are that the enjoyment associated with art ownership is multifaceted and that preferences interact with wealth in determining the magnitude of private values.
Popularity Page 5