(i.e., greater than 30 percent) capital loss than a smaller loss. They pay at-
tention to taxes in December, especially in late December. Controlling for
past market returns, investors who are holding a stock for a loss are less
likely to sell than those holding for a gain. And investors are more likely to
sell a stock that has hit its high price within the last month. Grinnblatt and
Keloharju also estimate logit regressions comparing selling decisions to
purchase decisions. They find that, generally, high past returns make it
more likely that a Finnish investor will sell rather than buy a stock. Finnish
investors are also more likely to buy a stock that is at a monthly low and
sell a stock that is at a monthly high. High volatility increases the propen-
sity of households to buy, rather than sell, a stock. And, consistent with the
life cycle hypothesis, the oldest investors are more likely to sell stocks and
the youngest more likely to buy.
G. Real EstateThe reluctance of investors to realize losses is not confined to stock and op-
tion losses. Genesove and Mayer (2001) find that homeowners, too, exhibit
a disposition effect. Genesove and Mayer examine individual listings for
Boston condominiums between 1990 and 1997. Their data track listing
dates and prices, exit dates, type of exit (e.g., sale or withdrawal), and
property characteristics such as square footage and number of bedrooms,
previous appraisal prices, owner occupancy, and mortgage information.
They determine which sellers face an expected loss by estimating current
condominium values and comparing these to original purchase prices. They
find that sellers facing losses set higher asking prices of 25 to 35 percent of
the difference between the expected selling price of a property and their
original purchase price. On average, these sellers attain higher selling prices
of 3 to 18 percent of that difference though their sales come, on average,
after more time on the market than those of sellers not facing losses.
2.Overconfidence and Excessive TradingPsychologists find that people tend to be overconfident. That is, they tend
to overestimate their abilities and the precision of their information. Odean
(1998b) demonstrates theoretically that overconfident investors trade more
than those who are not overconfident and, as a result of excessive trading,
lower their expected utilities.^8
Overconfidence increases trading activity because it causes investors to
be too certain about their own opinions and to not consider sufficiently the
INDIVIDUAL INVESTORS 553(^8) Other theoretical treatments of overconfident investors include De Long, Shleifer, Sum-
mers, and Waldmann (1991), Benos (1998), Kyle and Wang (1997), Daniel, Hirshleifer, and
Subramanyam (1998), and Gervais and Odean (2001).