Classical finance theory suggests that in efficient markets, rational investors
hold diversified portfolios and security prices reflect the discounted value of
expected cash flows. Future cash flows, in turn, depend on systematic risk
alone, and investor sentiment has no effect on security prices. Contrary to the
theory, the authors argue that investor sentiment may, in fact, affect security
valuation. They find that when investor sentiment is low, subsequent returns are
relatively high for stocks of small, young, highly volatile, unprofitable,
nondividend-paying, extreme growth, and distressed companies. On the other hand,
when sentiment is high, the returns for these stocks are relatively low.
According to classical finance theory, the present value of expected cash flows
discounted at a risk-adjusted discounted rate determines the value of securities and
investor sentiment plays no role in the valuation. The authors hypothesize that investor
sentiment may, actually, have a significant effect on stock prices. They suggest that
because it is difficult to value the stocks of certain companies—new, small,
young, highly volatile, unprofitable, nondividend-paying, extreme growth, and distressed
companies—stock prices are likely to be more affected by investor sentiment. The
authors suggest that the valuation of new companies, nondividend-paying companies, and
companies on the extreme ends of the growth scale is too subjective. The subjective
valuation makes these stocks speculative such that changes in investor sentiment drive
the stocks’ prices. The authors contend that, in practice, stocks that are harder
to value also tend to be harder to arbitrage. On the other hand, the valuation of
companies with a long earnings history, tangible assets, and stable dividends is much
less subjective and much less likely to be affected by sentiment.
The authors create a composite sentiment index by constructing six proxies. These proxies
include the closed-end fund discount, NYSE share turnover, the number of and average
first-day returns on initial public offerings, the equity share in new issues, and the
dividend premium. The data include all the stocks in the merged CRSP–Compustat
database. The authors use monthly stock returns between 1963 and 2001 to construct
equally weighted decile portfolios based on several company characteristics to test the
effect of sentiment at the beginning of a period on the cross-section of subsequent
stock returns.
Analysis reveals that when sentiment is low (below the sample average), the subsequent
returns are higher for newly listed stocks than for older stocks, for more volatile
stocks than for less volatile stocks, for unprofitable stocks than for profitable
stocks, and for nondividend-paying stocks than for dividend-paying stocks. When
sentiment is high, these patterns are reversed. However, although small stocks are found
to provide higher returns when sentiment is low, the size effect disappears when
sentiment is high.
Sentiment has a similar effect on extreme growth and distressed companies. These
companies fall at the two extremes when stocks are categorized into deciles by sales
growth, book-to-market ratio, or external financing activity; the more stable companies
fall in the middle deciles. The authors find that when sentiment is low, the subsequent
returns on stocks on the two extremes are particularly high whereas the stocks in the
middle do not appear to be affected by sentiment. This result is consistent with the
authors’ theoretical construct. The extreme growth and distressed stocks would be
expected to be affected by sentiment because their valuations are relatively subjective
and they are relatively hard to arbitrage.
The conclusions of this study challenge classical finance theory and suggest that stock
valuation needs to incorporate investor sentiment.