an investor does not already own as well as those she does, but prospect
theory applies only to the stocks she owns. Thus, a belief in mean-reversion
implies that investors will tend to buy stocks that had previously declined
even if they do not already own these stocks, while prospect theory makes
no prediction in this case. Odean (1999) finds that this same group of in-
vestors tends to buy stocks that have, on average, outperformed the CRSP
value-weighted index over the previous two years. This would appear in-
consistent with a pervasive belief in mean-reversion.
E. Employee Stock OptionsInvestors are more likely to sell stocks that they are holding for a gain over
purchase price than those they are holding for a loss. How do they treat se-
curities for which there is no clear purchase price? Heath, Huddart, and
Lang (1999) help to answer this question by examining when employees
exercise company stock options. Because employees do not pay for their
stock options, there is no purchase price that can serve as a reference point.
Options are generally issued with ten years to expiration and are typically
exercised for cash. Heath, Huddart, and Lang examine stock option rec-
ords for over 50,000 employees at seven publicly traded corporations over
a ten-year period. For each option grant (options issued on a single date),
they calculate the fraction of outstanding options exercised each week.
They regress this fraction on several variables including the fraction of op-
tions recently vested, the fraction of options soon to expire, the ratio of the
option’s intrinsic value to its expected value, recent returns, and a dummy
variable for whether the current price exceeds the maximum price obtained
in the previous year (excluding the previous month). As one would expect,
recently vested options, those soon to expire, and those with a greater in-
trinsic value relative to expected value are more likely to be exercised. Em-
ployees are more likely to exercise options when the exercise price is above
the stock’s previous year’s maximum, and they are more likely to exercise
options that have recently appreciated. These findings are consistent with
the disposition effect and suggest that when purchase price is not an avail-
able reference point, many investors focus on recent maximums as a refer-
ence point.
F. Finnish InvestorsGrinnblatt and Keloharju (2001) examine trading records for 1995 and
1996 for all Finnish stock investors—households, institutions, and foreign-
ers. Estimating logit regressions for the decision to sell rather than continue
to hold a stock, Grinblatt and Keloharju find that all investor groups are
more likely to sell stocks that have outperformed the market in recent
days and weeks. Investors are less likely to sell when faced with a larger
552 BARBER AND ODEAN