Behavioral finance, emerging in the 1990s as a counterpoint to Modern Portfolio Theory (MPT), pictures a world in which investing decisions are far more complex than cold trade-offs weighing only numerical measures of risk and return. “Theoretical models of efficient financial markets that represent everyone as rational optimizers can be no more than metaphors for the world around us. Nothing could be more absurd than to claim that everyone knows how to solve complex stochastic optimization models,” wrote Robert Shiller (emphasis in the original).¹

“. . . [A]lthough to err is indeed human, financial practitioners of all types, from portfolio managers to corporate executives, make the same mistakes repeatedly,” wrote Professor Hersh Shefrin of Santa Clara University in 2002. Behavioral finance recognizes the influence that human emotions and reactions—hopes of earning great profits, fear of difficult choices, and inconsistent reasoning about money—exert in economic and investing decisions.² Accordingly, behavioral finance has grown into a rich and diverse body of knowledge, earning Nobel prizes for its researchers. We limit our discussion to a few representative aspects.

Loss Aversion

MPT dictates that all investors make identical, dispassionate evaluations of investments based solely on expected return and volatility, but this assumption ignores the impact of a phenomenon known as loss aversion. People’s attitudes toward gains and losses are not symmetrical: loss aversion holds that the psychic pain an individual feels from losing, say, 10 percent on an investment is greater than the pleasure would be from gaining 10 percent. While such a reaction is asymmetrical, it’s not irrational: losses are limited to 100 percent of capital, while gains can be infinite, so the opposite of the benefit of a 110 percent gain can’t be a 110 percent loss.³

Loss aversion presents a challenge to the one-size-fits-all framework of MPT, because in the behavioral finance framework, each person’s internal calculation of loss aversion is thought to be unique. One investor might see the trade-off of an 8 percent loss as equal to a 10 percent gain, while another may be more sensitive to losses, and view a 5 percent setback as sufficient to offset the benefit of a 10 percent gain.⁴

Consider the stylized example of two prospective investments in Table 5.1. Investment A has a 10 percent chance of earning 100 percent, and a 90 percent chance of losing 5 percent, and thus an expected return of 5.5 percent.⁵ Investment B has a 90 percent chance of earning 15 percent, and a 10 percent chance of losing 75 percent, for an expected return of 6.0 percent.⁶ And by the numbers, the volatility of Investment A is slightly higher than that of B. Thus, according to the rational return-risk criteria of MPT, all investors would always prefer B (for its higher return yet lower risk). However, many investors would not give as much weight to the overall probabilities, and their attention instead would be drawn to the extremes of gain and loss—B’s 75 percent loss potential, versus a 5 percent possible loss for A—and choose A over B. (We conducted a survey among Epoch’s employees, and a sizable minority did in fact prefer A, in spite of B’s higher expected return based on the probabilities.)

Loss aversion illustrates that MPT’s two-dimensional framework of expected returns versus volatility is an inadequate measure of how real-world investors perceive risk.

Mental Accounting

Taking the notion of multiple views of risk one step further, behavioral economists have also identified ways in which one individual investor can simultaneously hold different opinions of the same investment. MPT asserts that investors view their assets as a single portfolio, but in reality many people think of their investments as being divided into different notional “accounts” (hence the name of this phenomenon). Perhaps the easiest way to see mental accounting at work is to observe people in casinos. A gambler who started the night with $1,000, and has managed to turn it into $1,500 after an hour of blackjack, will often think of the additional $500 as “the house’s money”—distinct from the original $1,000 stake, even though the entire $1,500 belongs to the gambler. Faced with a decision on whether to take a particular risk (for example, whether to hit or stand on a 16), many people will make a different decision depending on whether they are gambling with what they think of as their own money versus gambling with the house’s money.

In the context of investing, consider the case of an investor who purchased shares in a mutual fund two years earlier for $10,000, and since then has earned a $5,000 gain. The notion of mental accounting suggests that many investors might be willing to take greater risks with the $5,000 gain portion—the equivalent of the house’s money in the blackjack example—than with their own original $10,000.

The phenomenon manifests itself as well in the way that people divide their assets into multiple portfolios, not just mentally but in reality, and adopt different levels of risk appetite in their various portfolios. An investor may perceive a particular stock as being too risky for the college savings fund, but perfectly fine for the vacation home fund. Thus, many people do not make investment decisions by seeing their assets as an integrated whole, which runs against the assumptions that MPT makes about investor behavior.

Minimizing Regret

Another area where MPT is incomplete, by not taking into account the complexity of real-world investor behavior, is the assumption that all investors seek only to maximize their return for a given level of risk. Behavioral finance theorists argue that this is just one of multiple objectives, and that investors simultaneously seek both to maximize return and to minimize regret, objectives that often clash.⁷

Minimizing regret is another manifestation of loss aversion. In the loss aversion examples, one investment had a potential, although unlikely, loss of 75 percent. Many investors might avoid that choice, realizing that they will regret the loss if things turn out badly—even when, in MPT terms, it has a higher expected return and lower expected risk.

Regret minimization can also cause investors to hold on to investments which looked promising at the outset, but which have turned out to show poor prospects. Investors will use their purchase price as a reference point, and become reluctant to sell for less and “lock in” a loss (bringing the investment to an end, and thus feeling definite regret). Some other investment, or probably many, might offer better current return opportunities, but a regret-minimizing investor could hope that the first investment will someday recover to at least its purchase price, where it could be sold regret free.

Overconfidence

Holding too much confidence in ideas and aptitude extends to many fields of endeavor. In investing, it can lead investors, both individuals and professionals, to believe they know more than they really do about the market or individual stocks. Overconfidence can take the form of erroneous forecasts—such as false notions of where a company’s revenue growth or profit margins are heading—or an investor’s belief that he has learned a bit of news to which the market as a whole has not yet caught on. Professor Hersh Shefrin asserts that overconfidence hurts investors in two ways: first, having the wrong information leads them to buy the wrong securities; second, an excess of confidence causes them to trade more often than they should.⁸

Extrapolation and Reversal

Investors also are likely to see patterns where none exists. One common misconception is the extrapolation of a trend in returns—that a winning stock will keep winning, and losers will keep falling. The opposite, known as the gambler’s fallacy, holds that what has gone up must come down, and that trends will reverse. These sorts of forecasts are constructive when they are founded in changes in the facts of fundamentals of a company or a market climate, but when they simply express a view that it’s time for a change, they’re likely to be biased (and wrong).

These are just a sample of the many anomalies and exceptions to the systematic, dispassionate thinking that MPT has ascribed over time to the financial markets. Through the lens of behavioral finance, it is clear that investors do not all think and act alike and that their actions are not entirely rational and optimal.

Investor Behavior in Action

To illustrate how misperceptions make their way into stock prices, we consider the impact of the markets’ reactions to company earnings announcements (or post-earnings-announcement drift, a phenomenon widely studied by financial economists). If markets were efficient on earnings information, stock prices would adjust the moment that companies announce their earnings, or perhaps within a few days. Instead, researchers over 50 years or so have found that full adjustments to earnings news tend to be diffused over time. For instance, Hersh Shefrin cited studies suggesting that industry analysts—professional investors, that is—underreact to new earnings information when revising their forecasts, with the result that “one positive earnings surprise is followed by another, and then by yet another.”⁹

“Think about what this pattern implies,” Shefrin added: “It pays to hold stocks that have experienced recent large positive earnings surprises, because the market does not adjust fully to the good news. Instead, the market adjusts over the three quarters that follow an announcement.”

This is a sign of overconfidence at work, Shefrin contended—that analysts are overconfident in their prior forecasts, and underweight evidence that disconfirms their views. He also noted that other researchers had concluded that analysts underreact to news they gathered from public sources, such as earnings reports, while overreacting to information they turned up on their own. “The result is conservatism,” he concluded. “Permanent changes in circumstances get mistaken for temporary ones, at least up to a point.”

While academics and investors have known about these tendencies for decades, the market apparently has not bid them away. Academics Haigang Zhou and John Qi Zhu, writing in the Financial Analysts Journal in 2012, developed a profitable trading strategy from stocks with large positive and negative earnings surprises, based on the trading history of 1971 through 2009. “Market participants seem largely to underreact to, or are simply unaware of, the latent ‘extremely good (or bad) news’ signaled by the direction of jumps around earnings announcements,” they conclude.¹⁰

These points are important to active investment managers, as investors’ reactions, overreactions and missteps lead to the mispricing of securities, and create some inefficiencies and opportunities in the markets. However, it’s a long way from academic pronouncements that securities markets are efficient or inefficient to assembling actively managed portfolios that outperform the market.

MPT Still Lives

MPT’s proponents have not been content to let behavioral finance steal the academic spotlight. In part, this may simply be due to the worldview of most economists. As MIT Professor Andrew Lo noted, in a paper on corporate culture:

[T]he culture of economics . . . prizes the narrative of rational economic self-interest above all else. Given two competing explanations for a particular market anomaly, a behavioral theory and a rational expectations model, the vast majority of economists will choose the latter—even if rationality requires unrealistically complex inferences about everyone’s preferences, information, and expectations.

Lo concludes that the mathematical elegance of an equilibrium according to rational expectations usually trumps a “messy and imprecise narrative”—such as those founded on less formal behavioral economics.¹¹ Thus, when investment practitioners and skeptical academics documented the anomalies to MPT—showing that stocks failed to perform in line with the predictions of the Capital Asset Pricing Model (CAPM) and its single risk factor, beta—its adherents countered that the early CAPM had been too simplistic (or, in the technical parlance, misspecified), and responded with newer, more complex versions of MPT.

In 1992, Professor Eugene Fama, who formulated the Efficient Market Hypothesis, and Professor Kenneth French of Dartmouth College improved on the simple CAPM with their documentation of the “value effect”—an inverse relationship between how stocks are valued in the market and their subsequent returns.¹²

Fama and French devised a three-factor model, expanding the simple CAPM to include a variable on firm valuation, and another metric on company size (thus creating a three-factor CAPM). They found that the additional factors greatly enhanced the explanatory power of plain-vanilla beta based on volatility alone. Simply stated, “value stocks” that are given cheaper valuations—lower ratios of price to earnings, cash flow, or book value—tend to show better subsequent returns than stocks with higher valuation ratios. The value effect has since been studied and documented far and wide, over many time periods and in many markets around the world. Similarly, the size effect shows that smaller companies earn systematically higher returns than larger companies.¹³ One plausible explanation is that investors see lower-valued companies, often selling cheaply because they are down on their luck, as well as smaller companies, as more risky, and thus demand greater returns for investing in them.¹⁴

Fama and French also offered crucial inferences on the risks in stocks, and the shortcomings of the rational investor views of simple beta and the CAPM: “If assets are priced rationally, our results suggest that stock risks are multidimensional. One dimension of risk is proxied by [company] size. . . . Another dimension of risk is proxied by . . . the ratio of the book value of common equity to its market value. . . . It is also possible, however, that [the ratio of book value to market value] just captures the unraveling . . . of irrational market whims about the prospects of firms.”¹⁵

Recently, Fama and French have created yet another version of the CAPM, this time adding two more company risk factors: firm profitability and levels of investment in plant, equipment, and acquisitions.¹⁶ They find—not surprisingly—that higher profitability tends to be associated with better stock performance. More unexpected, however, is that higher levels of investment by companies—aggressive expansions of their operations—are associated with poorer stock performance. Taken together, the returns to these two risk factors suggest that companies able to generate greater profits with lower capital investment are likely to be rewarded with higher returns on their shares. (In Chapter 10 we discuss in detail the importance of companies’ return on invested capital in the investment process.)

The academic debate may go on forever, as the MPT and behavioral finance camps compete with new and improved models of investors and markets. But the dogmatic views in the 1960s and 1970s, which dictated that all investors should hold the same “market portfolio,” and effectively ruled out that active investment management could add value over a passive strategy, have been greatly undermined by the development of behavioral finance.