Popularity

by Roger Ibbotson

  Our fourfold classification of security characteristics partially overlaps with the threefold classification in Statman (2017) , in which investors are described as holding securities for utilitarian, expressive, and emotional reasons. Utilitarian reasons correspond to risk and frictional characteristics, and expressive and emotional reasons correspond to psychological characteristics.

  In this book, we focus primarily on the stock market, although we believe the concepts can be applied to fixed-income securities, real estate, and numerous other real assets. 6 Periodically, as necessary, we attempt to distinguish between characteristics of a company and characteristics of the security in question—both of which can have attributes that are more or less popular among investors. Assets are priced not only by their expected cash flows but also by the popularity of the other characteristics associated with the company or security. The less popular stocks have lower prices (relative to the expected discounted value of their cash flows), thus higher expected returns. The more popular stocks have higher prices and, therefore, lower expected returns. Popularity can be related to risk (an unpopular characteristic), and it can also be related to other rational preferences. But popularity can also be related to behavioral concepts. For instance, investors may want to brag about their past winners (or purchase recent winners—for example, in the practice called “window dressing”) or hold recognizable securities that are consistent with their social values. Any aspect that can affect the popularity of a stock will affect its demand and thus its price. 7

  Popularity is a bridge between classical finance and behavioral finance because both types of finance rely on preferences. Popularity is an expression of these preferences, whether they are rational, irrational, or somewhere in between. 8 Popularity does not make a value judgment but, instead, takes preferences as a given and recognizes that preferences can change over time. This book is presented in an equilibrium framework, so asset prices and expected returns reflect the aggregate impact of investor preferences.

  Principles and Models of Classical Finance

  Classical finance rests on several principles that largely come from economics:

  Rationality. This principle states that investors are rational utility maximizers who care about cash flows, expected return, and risk. Although considering liquidity and tax efficiency is also rational, classical finance often assumes away the nonrisk aspects of investments.

  Risk-free arbitrage. In economics, this concept is known as “the law of one price.” 9 In finance, it specifically means that in the absence of any frictions, such as transaction costs, two securities with the same payouts must have the same price.

  Equilibrium. This condition is “supply equals demand.” The prices of securities are such that every available share is held by some investor in the quantity that the investor wants to hold at the prevailing prices. A frictionless equilibrium has no transaction costs and no risk-free arbitrage opportunities.

  Efficient markets. In an efficient market, security prices reflect all relevant information regarding the securities’ value. In an efficient market, all prices are “fair”—meaning that they equate to their intrinsic values (which are usually not directly observable) and investors can “beat the market” only through luck or random variation around market benchmarks. As Fama (1970) pointed out, any test of market efficiency must be a test of a joint hypothesis with some other model that explains rational differences in expected returns.

  Models in classical finance generally assume rationality and market efficiency. Where they differ is in whether they assume that risk-free arbitrage eliminates all price discrepancies or whether they make the assumption of equilibrium. The best-known examples of arbitrage models are the arbitrage pricing theory (APT) (Ross 1976 ) and the Black–Scholes option pricing model (Black and Scholes 1973 ). Because these models make no direct reference to investors, they make no reference to investor preferences.

  The best-known equilibrium model is the CAPM (Sharpe 1964 ; Lintner 1965 ; among others). In the CAPM, each investor maximizes a utility function that is higher as expected return increases and lower as variance of return increases. The model takes into account a single preference parameter for risk aversion that identifies how various investors make the trade-off between risk and expected return. In equilibrium, the model assumes a single risk premium, which is a function of the asset-weighted aggregation of the risk aversion parameters. 10

  The practical distinction between arbitrage and equilibrium is in what the two approaches say about prices. Arbitrage models describe the relationships between security prices but do not explain where prices ultimately come from. For example, the Black–Scholes model gives formulas for options on stocks but does not explain how the prices of stocks are determined. The APT, another arbitrage model, says the expected return on a security is a linear function of exposures to a set of factors, with risk premiums on the factors as the coefficients. It does not, however, explain where the risk premiums come from. In contrast, equilibrium models explain security prices in terms of both the characteristics of the securities and the preferences of investors. The CAPM and the popularity asset pricing model (PAPM) that we present in Chapter 5 are equilibrium models.

  The New Equilibrium Theory (NET) of Ibbotson, Diermeier, and Siegel (1984 ; hereafter, IDS) is a precursor to the popularity framework that includes frictional characteristics—such as asset liquidity, taxation, and divisibility—in a general asset pricing framework. These characteristics do not necessarily have to be linear in asset pricing, although for simplicity we specify a model later in this book in which they are. The key is that rational investors have different preferences for these characteristics, making some securities more popular than others. In a rational world, equilibrium prices and expected returns are affected by an asset’s frictional characteristics. Although NET was imagined as a potential equilibrium theory, IDS (1984) only outlined what such a theory might look like. In Chapter 5 of this book, we present a fully developed equilibrium asset pricing theory in which investors have preferences for all types of characteristics.

  To the extent that an asset’s characteristics are permanent and cannot easily be reconstructed or securitized in some form, the disliked characteristics result in expected return premiums that can be predicted over time. An example might be the small-capitalization premium; small-cap stocks are disliked in general because they are riskier and less liquid than large-cap stocks and require more analysis per dollar of investment. If value stocks were permanently unpopular, perhaps because poorly managed or unlucky companies are perceived as poor investments (although we would argue that they can be great investments), they might also receive a long-term premium without being any riskier than growth stocks.

  In an equilibrium in which investors have preferences regarding company and security characteristics that are relatively permanent, companies and securities with disliked characteristics will have lower prices and higher expected returns than popular ones, giving rise to what we consider long-term “premiums.” In classical finance, the primary disliked characteristic is risk and only the systematic or nondiversifiable part of a security’s total risk bears a premium. (Even with rational investors, assets that are difficult to diversify, such as owner-occupied houses or human capital, may have prices that include premiums for idiosyncratic risk.) Furthermore, other characteristics—such as liquidity, taxability, or divisibility—would and should be priced by rational investors.

  In this book, we take an equilibrium approach: Investor preferences influence prices and expected returns. In such an equilibrium, natural clienteles may come about in which various investor desires and holdings differ; prices, however, reflect the aggregation of those preferences in the context of the relatively fixed supply of characteristics in the marketplace. The popularity preferences are expressed as demand for those characteristics and an overall demand for expected returns.

  Principles of Behavioral Finance

  In behavioral finance,
individuals suffer from various biases that limit their rationality. Included are many of the biases and heuristics discussed by Tversky and Kahneman (1974) , such as the affect heuristic, framing, loss aversion, anchoring, the endowment effect, and overconfidence.

  Behavioral economics or behavioral finance is applied in numerous ways. One way is to nudge investors toward better behavior, as suggested by Thaler and Sunstein (2008) . Practitioners or academics might develop schemes that help investors save more, plan better for contingencies, and diversify more effectively. In our context, however, we are interested in understanding how investor behavior might affect security prices and returns, not in modifying behavior.

  We illustrate in Chapter 5 how, in equilibrium, investor preferences affect security prices and expected returns. In a PAPM illustration, one of the investors is purely rational and cares only about classical characteristics. This investor sees popular securities as being overpriced and takes advantage of this insight by taking levered short positions in those securities and levering up long positions in unpopular securities. The purely rational investor thus influences prices but does not determine them. Therefore, no complete arbitrage takes place in the equilibrium environment because prices reflect the aggregate demand for each security as influenced by all the participants in the market.

  We classify the behavior biases or distortions into two distinct types: psychological and cognitive. The psychological biases express desires or needs of the investors; some investors knowingly and willing pay extra to achieve some other good, such as influencing a social goal. The cognitive aspects of behavioral finance usually involve some mistakes investors make, such as overestimating the earnings of a growth company or imagining (against evidence) that momentum will continue.

  Demand and Supply

  In the primary capital market, companies demand capital and, in exchange for the provision of capital, supply earnings and, subsequently, cash flows. The cash flows come in the form of dividends, buybacks of shares, cash paid for shares of acquisition targets, and so on. Looking at this process from the other side, investors supply capital and demand cash flows and returns.

  The concept of popularity looks at the market from a demand perspective and takes the expected cash flow streams as given. Many investors value a stream by applying a discount rate to it. This discount rate, equivalent to investors’ expected return, will be higher or lower depending on the risk, recognizability, liquidity, or general desirability of the stock above and beyond its expected cash flow characteristics. Thus, popularity is a measure of demand: The more demand for a stock, the higher the price and, consequently, the lower the expected return. In the primary capital market, then, capital flows from investors to corporations (through the corporations’ costs of capital), which attempt to invest in projects that will generate returns, which enriches corporations and ultimately provides returns back to investors. Figure 1.1 illustrates these relationships.

  In addition to expected returns, investors demand less risk, more liquidity, tax efficiency, dividends, and other desired characteristics. Figure 1.2 takes these preferences into account.

  Figure 1.1. Supply and Demand for Capital and Returns: Flow Chart

  Figure 1.2 depicts the following exchanges between investors and corporations:

  Investors supply capital to the company (the corporate treasury) by purchasing stock and bond issuances (in the primary [direct] capital markets).

  The company (corporate treasury) supplies assets to corporate operations/projects in the form of working capital and assets.

  Corporate operations/projects spend the money in hopes of providing earnings, cash flows, brand power/recognition, and so on, back to the company (corporate treasury).

  The corporate treasury (through its cost of capital) provides returns and cash flow back to stock- and bondholders (which may be the original purchasers from the primary [direct] capital markets or may be acquirers of the securities in the secondary market).

  ______________________________

  Figure 1.2. Exchanges between Investors and Corporations

  Although we will focus on the upper right-hand box depicting investor demand for expected returns and security (termed “popularity”) characteristics, we recognize that it is corporations that ultimately supply the returns and characteristics through their investments in assets and projects, as expressed through their issuance of debt and equity.

  Popularity Premiums

  We define a premium as an expected excess return, relative to an appropriate benchmark, that is relatively permanent (which we emphasize to distinguish premiums from short-term mispricing). The most clear-cut example is the equity risk premium in which the benchmark is a risk-free asset or a bond index. Stocks are expected to have higher returns than bills and bonds because stocks are riskier. In most cases, risk or uncertainty is undesirable or unpopular. The extra risk or uncertainty in stocks is inherent because, legally, stockholders have a residual claim on a company’s cash flows only after all bond interest and principal are paid. Thus, stocks always have positive expected excess returns—but not necessarily higher realized excess returns, given that uncertain results almost always differ from expectations. The equity risk premium, like any other premium, will be positive in its expectation but, of course, can vary over time as the supply of and demand for capital changes. Over the long run, stocks have indeed had substantially higher returns than bonds or bills.

  Risk is only one of many potential reasons that one stock may have a higher expected return than another stock even with similar patterns of expected cash flows. For a return differential to be called a “premium,” however, it should not only be permanent and not easily changed, but it should also be relatively consistent over time. For example, some investors will always prefer a liquid investment. Usually, a company cannot easily change the liquidity of its equity. So, the demand for and supply of this liquidity is relatively permanent and will affect a stock’s price and expected excess return. The investor willing to invest in the less liquid stock should receive a premium. Arbitrageurs might try to reduce the liquidity premium, perhaps by wrapping illiquid stocks in such a liquid vehicle as an exchange-traded fund, but not all securities can be made uniformly liquid. Thus, like risk, liquidity will be priced in the market.

  This book posits a variety of characteristics that might be relatively permanent and are likely to be priced in the stock market. In the CAPM, a security’s beta is the only priced characteristic, but in the popularity framework, beta can be either popular or unpopular. The CAPM specifies that the more market or beta risk a security has, the higher its discount rate, the lower its price, and the higher its expected return. Empirical data on individual stocks, however, do not usually support this positive relationship between stock returns and betas. Rather, some analysts have suggested that because the market usually goes up and because investment managers, acting as agents for their investors (customers), want to outperform a majority of the time, these managers may prefer high-beta stocks despite their added extra risk. This phenomenon is a likely explanation of the low-beta anomaly we discuss in Chapter 2 .

  Other premiums that are generally accepted are size and value. Presumably, all small companies strive to become large companies; however, aspiring and pursuing do not necessarily create reality. Small size is usually a deterrent to institutional investors because it is usually associated with less liquidity, and small-cap stocks generally take more analysis per dollar of investment. Small-cap stocks are also riskier (an unpopular characteristic) than large-cap stocks; thus, the absolute outperformance of small versus large is consistent with the more-risk/more-return CAPM paradigm. Although directionally aligned, the realized premium of small caps has empirically exceeded that which the CAPM would predict.

  The value premium is less easily understood, especially from a CAPM perspective, because not much empirical evidence suggests that value stocks are riskier than growth stocks, at least as measured by their standard devia
tion of returns or CAPM betas. Nevertheless, investors generally think of value stocks as less attractive, perhaps even distressed, companies. Thus, if investors seek growth over value, a value premium will exist in the market. Whether this is a permanent premium is unclear.

  Many premiums are not risk based but, rather, are associated with non-risk investor preferences. For example, investors may desire and will pay up for environmentally sensitive companies or those that meet or avoid various social criteria. Companies can, of course, change their ESG policies; in fact, change is what some investors are trying to achieve. But change comes at a cost, and ultimately, investors will have to pay up for these preferences, potentially earning lower returns.

  Some preferences are purely behavioral, reflecting preferences that are not risk or return related. For example, investors prefer companies with familiar brands and companies that are successful and have good reputations. All these criteria can make for great companies but not necessarily great investments because the characteristic in question is probably already reflected in the price. Some of these criteria overlap with the criteria by which value companies are judged (e.g., value companies may have weak brands or produce items the general public does not appreciate), which may partially explain why value stocks earn a premium. Of course, even though companies usually want to improve their situations, brands and reputations and core product offerings are not easily changed. Value companies might wish to become growth companies to become more desirable to investors, but here again, a stock’s classification is not easily changed on the supply side, nor is investor excess demand for growth likely to disappear.