The assumption that investors hold rational expectations of market returns is central to many asset pricing models. However, in recent years, surveys of investors have revealed that market participants’ reported expectations often deviate from the objective predictions of financial models working with large pools of data. One theory is that these deviations are the result of persistent pessimism on the part of investors: survey respondents, according to this hypothesis, are discounting the rationally expected rate of return to reflect the risk of investing in stocks.
To examine whether investors have a pessimistic bias, Oxford’s Klaus Adam, the Bank of Canada’s Dmitry Matveev, and Chicago Booth’s Stefan Nagel examined existing evidence—including surveys of individual investors, professional investors, and CFOs going back to the 1980s—to compare expected returns with realized returns. The research suggests that, contrary to the pessimism hypothesis, investors are just as likely to be optimistic.
Investor expectations closely matched realized market returns over the full length of time the researchers examined. But at any given time, expectations tended to be procyclical: investors expected higher returns during boom times in the stock market and lower returns during market contractions, even though many asset pricing models work in precisely the opposite direction. Thus, the apparent conformity of investor expectations to market returns on average over time actually reflected investors’ biases—alternately optimistic and pessimistic, with the two balancing each other out.
“In boom times, like the late 1990s, investors are too optimistic,” Adam, Matveev, and Nagel write.“Following crashes, like in early 2009, investors are too pessimistic.”
The researchers suggest several explanations that could contribute to investors’ tendency to deviate from rational expectations in this predictable way. Prior research by Nagel and University of California at Berkeley’s Ulrike Malmendier has found that investors tend to let their personal experiences take an oversized role in establishing risk tolerance and inflation expectations, and perhaps this tendency plays some part in shaping expectations of market returns as well. Separately, Yale’s Nicholas Barberis, Harvard’s Robin Greenwood and Andrei Shleifer, and California Institute of Technology’s Lawrence Jin have found evidence that some investors’ expectations are formed by extrapolating from recent returns.
Whatever the explanation, Adam, Matveev, and Nagel suggest their findings are difficult to reconcile with the rational expectations hypothesis. “It does not appear possible to simultaneously match asset price dynamics and the survey evidence without entertaining some departure from rational expectations, such as extrapolative expectations,” they write.