Equilibrium —Asset prices are determined in markets so that all assets are willingly held.
Understanding Historical Returns
The authors of this book, along with various collaborators of theirs, have gathered quite a few of the pieces of modern finance—pieces that, when assembled, begin to explain a lot about the way assets are priced and portfolios are constructed. Let’s start at the beginning: “Stocks, Bonds, Bills, and Inflation: Year-by-Year Historical Returns (1926–1974)” by Roger Ibbotson and Rex Sinquefield (1976a) .
This familiar work, originally published in the University of Chicago’s Journal of Business in 1976, was released as a book by the CFA Institute Research Foundation in 1977 (when the organization was called the Financial Analysts Federation). The book achieved wide distribution and influence. It addresses one of the components of NET theory: risk. How much risk is in each asset class, and what is the market price of each risk? That is, how much compensation in the form of higher return do investors, as a group, require for taking a given amount of risk?
Ibbotson and Sinquefield answered the questions of risk by measuring how much investors got as compensation for the various risks in the market. Asserting that investors conform their expectation of reward-for-risk to that which proves achievable in the market, the authors concluded that the realized reward—which, the authors revealed, had been quite large for equities as compared with bonds and bills—was a satisfactory indicator of the expected or required reward.
A New Kind of Forecasting
This insight opened up a new avenue for forecasting. Ibbotson and Sinquefield (1976b) not only measured the average return on each asset class, and thus on the difference between asset classes; they also documented all of the monthly and annual returns. Doing so made it possible to measure the variability of returns—that is, the amount of risk for which investors were being rewarded—not just the size of the reward.
By “pricing” risk in this way, Ibbotson and Sinquefield were able to extrapolate past returns into the future (making an adjustment for interest rates). They not only estimated the mean or expected return on each asset class; they also forecast the whole distribution of potential future returns. They called these extrapolations “probabilistic forecasts.”
People were already familiar with probabilistic forecasts of the weather, but in investment finance they were something new and different. Under Ibbotson and Sinquefield’s influence, probabilistic forecasts have become standard practice in financial planning. 1 “You have an X% chance of earning at least Y%”—a phrase that would have baffled most planners before Ibbotson and Sinquefield did their pioneering work—is now heard everywhere. The emphasis on risk, on deviation from the expectation, is the most important benefit of this approach.
The Supply of Capital Market Returns
But where did the money come from to provide these rich rewards?
In a companion paper to “The Demand for Capital Market Returns,” Diermeier, Ibbotson, and I noted that the aggregate return to investors in the capital markets must be set by the amount of profit that corporations can earn in the real economy. 2 We called our paper “The Supply of Capital Market Returns” (Diermeier, Ibbotson, and Siegel 1984 ), where we observed that corporate profits cannot grow indefinitely (and are unlikely to shrink indefinitely) as a percentage of GDP; thus, real corporate profit growth should proceed at about the rate of real GDP growth over the very long run.
Price-to-earnings ratios also cannot rise or fall indefinitely, so the real GDP growth rate, we argued, is a good proxy for the expected real capital gain of an equity portfolio. Moreover, investors receive dividends and other cash payouts, such as buybacks, and these rewards are in addition to profit growth because they are paid out of profits that are not reinvested in the company. Inflation also must be accounted for. The sum of all of these inputs gives a supply-side estimate of the return available to investors from capital markets, which is the aggregation of all the individual security returns addressed by the demand-side approach in the current book.
The authors working with Roger Ibbotson have produced several more articles on the supply model, including a Graham and Dodd Award–winning article by Ibbotson and Chen (2003) and a fine integrative piece by Straehl and Ibbotson (2017) .
The Liquidity Factor
With a market price for risk established in Stocks, Bonds, Bills, and Inflation , the natural next step was to price the other attributes, such as liquidity (which we called “marketability”), that we identified as affecting asset prices in Ibbotson et al. (1984) . In an important precursor to the current book, Ibbotson, Chen, Kim, and Hu (2013) considered one of the factors in isolation: liquidity. Their article asked whether the fact that many investors are averse to illiquidity means that illiquid assets offer superior returns to investors who are not so averse to it. In the abstract of the article, the authors wrote,
Liquidity should be given equal standing with size, value/growth, and momentum as an investment style … [and] is an economically significant indicator of long-run returns. The returns of [the] liquidity [factor] are sufficiently different from those of the other styles that it is not merely a substitute. (p. 30)
The authors back-tested a strategy based on this idea and found large excess returns earned by portfolios of illiquid stocks. By subsuming liquidity into the larger category of popularity—a stock may be popular for many reasons, liquidity being one. In a Graham and Dodd Scroll–winning article, Idzorek, Xiong, and Ibbotson (2012) applied similar concepts to mutual funds. Finally, Ibbotson and Idzorek (2014) and Idzorek and Ibbotson (2017) were the first to specifically name popularity as the embracing concept that includes liquidity and other preference-related factors, arriving at the conceptual framework that is presented in this book.
As IIKX show through empirical tests, popularity is much more than liquidity. It includes such components as brand value, competitive advantage, and reputation as well as more conventional factors, such as high growth rates, profitability, and high beta. All of these attributes, say IIKX, should be avoided by investors seeking above-market returns because assets that have these characteristics are oversubscribed by other investors. By selecting assets with the opposites of these characteristics, investors can expect to earn excess returns.
Conclusion: It’s Hard but Not Impossible to Beat the Market
Investing in stocks or other assets that most people don’t want has a long and rich history, proceeding from Graham and Dodd (1934) through Warren Buffett and many scholars, active managers, hedge fund entrepreneurs, and private equity managers. They all take advantage of some aspect of the popularity hypothesis set forth in this book.
Yet, investing in unpopular assets is hard. First, they are typically unpopular for a reason. Mounting losses instead of bountiful profits, declining market share or a shrinking market for one’s product, an unusual loading of debt, and other characteristics that drive investors away are often indicators of continued poor performance rather than of what one value manager optimistically calls “troubles that are temporary.” This value trap is the pitfall that awaits investors who too blindly follow an unpopularity formula.
Investing in unpopular assets is hard for another reason: Active managers, including those believing themselves to be contrarian, engage in herd behavior. Their quantitative screens all tend to identify the same stocks. If managers focusing on unpopular assets have already formed a cluster of demand for an asset—even if that cluster represents a minority opinion—that asset may no longer be attractively priced.
Following these ideas to their logical conclusion brings the well-known fallacy of aggregation into consideration: Any strategy or factor that is widely enough used will fail. It is easy to imagine so much money flooding into an unpopularity strategy that no unpopular assets exist any longer. If that were to happen, the whole world would become a gigantic closet index fund. We financial economists really do lose sleep over thoughts like that.
Despite the
se concerns, the market has rewarded value investing and other strategies, such as those advocated in this book, that rely on buying what other investors are avoiding. Value, for example, has won over very long periods of time (back to 1927, according to Eugene Fama and Kenneth French 3 ) and by an economically significant margin. But value has not been on top recently; a small number of large and fast-growing companies have increased in relative popularity and beaten almost everything else. Like all other trends in investing, this one will surely turn sooner or later.
Meanwhile, read this book. It returns the CFA Institute Research Foundation, which is proud to present it, to its roots in quantitative financial research while helping to bridge a philosophical divide. And it might contain a key to that most elusive of Greek letters, alpha. Past performance is obviously no guarantee of future results, but it sure is a hint.
Laurence B. Siegel
Gary P. Brinson Director of Research
CFA Institute Research Foundation
July 2018
Preface
The idea that the popularity of an asset affects its pricing, and ultimately its return, is not new but is often overlooked in the mathematics of asset pricing models. Popularity is really just another word for demand, and of course, neoclassical economics—on which standard finance is based—is all about supply and demand. In the short run, the supply of an asset, such as the number of shares of a stock, is relatively fixed. Even when a company is subject to no news, however, the daily price fluctuates. This fluctuation is driven primarily by changes in the demand for the stock.
Popularity can shift daily or even hourly, but it can also be a relatively stable phenomenon. Some companies are inherently attractive or popular, while others remain uninteresting for long periods of time. Some companies have characteristics that investors seem to like, such as a great story behind them with exciting prospects ahead. Other companies plug along with good results but do not inspire the imagination of investors. These boring or even unattractive companies will have lower valuations, and thus higher costs of capital, than the popular companies. We assert it is strategies built on these overlooked stocks, however, even if such strategies appear to underdeliver day to day, that perform the best over the long run.
Recognizing that popularity can affect pricing does not necessarily lead to immediate excess returns. Rather, popularity is usually associated with valuation, an indicator of long-term future performance rather than a predictor of short-term or technical supply/demand imbalances. Benjamin Graham (2006) noted this aspect long ago in his book The Intelligent Investor :
Buying a neglected and therefore undervalued issue for profit generally proves to be a protracted and patience-trying experience. And selling short a too popular and therefore overvalued issue is apt to be a test not only of one’s courage and stamina, but also of the depth of one’s pocketbook. (pp. 31–32)
This book is based on the insight that Graham and others have always had—namely, that popularity affects security prices and thus expected returns. Whereas Graham focused on mispricing, we focus on long-run premiums. Relative popularity is driven by the collective wisdom—or perhaps not-so-wise collective opinion—of the crowd/investors, so going against the collective wisdom that drives popularity is inherently contrarian. We show that popularity is a broad umbrella under which nearly all market premiums and anomalies, including the classic value and small-cap anomalies, fall. We show this by drawing from both classical and behavioral finance to extend existing asset pricing models to include any security characteristic that investors might care about.
The capital asset pricing model (CAPM), which has dominated finance for the last 50 or more years, is simple and elegant. It is an equilibrium model built on neoclassical economics. From a practitioner perspective, it is extremely simple to apply. The CAPM ignores the insights of behavioral finance, however, and in numerous and systematic ways fails to accurately model asset prices. 4 In this book, we move from an intuitive understanding of popularity to, first, a framework for understanding how popularity predicts the direction of various premiums and anomalies relative to the CAPM and, eventually, to the development of a formal asset pricing model that incorporates the central idea of popularity, which we call the “p opularity a sset p ricing m odel” (PAPM).
Finally, for the popularity framework to be useful, it should not only be consistent with existing well-known empirical results. It should also predict premiums and anomalies that have not been considered before as priced characteristics. Examples of such characteristics are a company’s brand, reputation, and perceived competitive advantage. In this book, we show empirically that these characteristics are priced.
Hence, in both theory and empirical work, this book presents popularity as a bridge between classical and behavioral finance.
1. Introduction
The existence of various market premiums and anomalies is well established in the finance literature. To date, however, no single agreed-upon explanation for them has emerged. Investment finance is largely divided into two camps, classical and behavioral. Classical finance is based mainly on the idea that investors are risk averse, so market premiums are generally interpreted as risk premiums. In behavioral finance, premiums are considered to be the result of either cognitive errors that investors systematically make or preferences for company or security characteristics that might not be related to risks. We believe that most of the best-known market premiums and anomalies can be explained by an intuitive and naturally occurring (social or behavioral) phenomenon observed in countless settings: popularity.
What Is Popularity?
Popularity is the condition of being admired, sought after, well-known, and/or accepted. A wide range of possible categories—people, food, fashion, music, places to live, types of pet, vacation destinations, television shows, and so on—contain an implicit popularity spectrum or rank. Each of the categories has various criteria for estimating popularity.
For our purposes, the quality of the ranking criteria is not important; what is important is that any given category comprises a natural ordering in which some constituents are more popular than others. Such relative popularity evolves over time. Some aspects of popularity are systematic, or more or less permanent (for example, modern society seems to prefer thin to fat, tall to short). Other aspects of popularity may be transitory or exist only as fads (for example, necktie width, high-waisted jeans, men wearing wigs). Whether the result of systematic trends or idiosyncratic evolution, these rankings are in flux. Some popular items become relatively less popular, and some of the unpopular items become relatively more popular. While unsustainable, some popular items will temporarily become even more popular. For example, liquidity is permanently popular, but on a relative basis during times of market distress, it is especially sought after. Society places a greater relative value (monetary or otherwise) on the more popular items.
In this book, popularity refers to investor preferences—that is, how much an asset is liked or disliked. Of course, the primary preference for investors is to seek returns. Investors do not know what the returns will be, but they can distinguish one asset from another in terms of their observable characteristics, for which they may have clearly defined preferences. Thus, even with the same set of expected cash flows, investors may have more demand for one asset over another, which gives the preferred asset a higher current price and a lower expected return. An asset could be liked (or disliked) for rational or irrational reasons. 5 In this way, popularity spans ideas from both classical and behavioral finance, thus providing a bridge between the two camps.
In classical finance, the primary preference, beyond maximizing expected return, is to take less risk. This fact has given rise to various models that usually assume no other preferences. In the most well-known model, the capital asset pricing model (CAPM), the only “priced” characteristic is exposure to undiversifiable market risk. We consider a broader set of preferences that lead to other priced characteristics, which mig
ht include the rational preferences to reduce catastrophic losses, increase liquidity, be tax efficient, and so on. We also consider preferences that might be more in line with what the literature considers “behavioral,” such as desiring to hold companies with strong brands, investments with strong past price increases, or companies that have strong ESG (environmental, social, and governance) characteristics.
The popularity framework presented in this book includes a generalization of a wide range of characteristics in classical finance and behavioral finance that influence how investors value securities. We can classify these characteristics into two broad categories with two subcategories each as follows:
Classical:
Risks. In classical finance, risk usually refers to fluctuations in asset values, but risk can be interpreted more broadly as any risks to which a rational investor, who assumes away any real-world frictions in the holding and trading of securities, would be averse. Thus, risks may be multidimensional, including various types of stock or bond risks, or may arise from catastrophic events.
Frictional. These characteristics are often assumed away in classical finance, but a rational investor would consider them. Examples include taxes, trading costs, and asset divisibility.
Behavioral:
Psychological. Investors consider these characteristics because of their psychological impact. For example, buying a company with a small carbon footprint might make an investor feel good.
Cognitive. Investors consider these factors or fail to accurately interpret such factors because of systematic cognitive errors. For example, investors may overvalue the importance of a company’s brand when evaluating its stock because they do not realize that the value of the brand is already embedded in the market price of the stock.
Popularity Page 2
