II. Stock Returns and the Equity PremiumThe second section of the book discusses the overall stock market and two
distinct puzzles. The first puzzle is the long-term predictability of the stock
market. Although the stock market was long thought to be a random walk,
and thus unpredictable, numerous researchers have found that over long
horizons (several years) the returns on the stock market are at least some-
what predictable. Specifically, when stock prices are very high, as judged by
price/earnings or price/dividends ratios, then subsequent returns tend to be
low. John Campbell and Robert Shiller review this evidence in Chapter 5.
Their analysis suggests that stock prices were too high by the mid-1990s,
and they were predicting (albeit too early) the bear market that eventually
arrived. In fact they are indirectly responsible for one of the most famous
phrases to emerge from the 1990s. The story is that Campbell and Shiller
presented an early version of this chapter to the Federal Reserve Board a
few days before Chairman Alan Greenspan made a speech containing the
words “irrational exuberance.” Neither John nor Bob recall using that
phrase in their talk, but they did seem to give Greenspan the idea. In a fair
exchange, Shiller then borrowed the phrase back for the title of his best-
selling book.
The other puzzle discussed in this section is the “equity premium” puz-
zle. Briefly put, the historical difference between the return on equities and
the risk free rate has been judged too big to be explained within traditional
asset pricing models of expected utility maximization. In Chapter 6 Shlomo
Benartzi and I offer a behavioral explanation for part of this puzzle based
on two concepts from the psychology of decision making: loss aversion(the
tendency to weigh losses much more heavily than gains) and narrow fram-
ing(the tendency to consider returns over brief periods of time rather than
the long run. We show that the equity premium can be understood if people
weigh losses twice as much as gains and evaluate their portfolios roughly
once a year. In the appendix to the chapter, added for this book, we review
a few experimental studies offering empirical tests of myopic loss aversion.
In Chapter 7 loss aversion is incorporated into a much more sophisticated
equilibrium model by Nick Barberis, Ming Huang, and Tano Santos. They
also incorporate a third behavioral concept: the house money effect,^3 the
idea that when investors consider themselves to be “ahead” in the game
they are playing, they are more willing to take risks (since they are playing
with the “house money,” in gambling parlance).
PREFACE xiii(^3) See Thaler and Johnson, 1990