one-third to win two and half times the amount invested. The subjects were
provided an endowment, and they had to decide how much to invest in the
above lottery. Half of the subjects played twelve individual rounds, whereas
the other half played them in blocks of three. Those playing in blocks of three
were provided feedback at the end of the third, sixth, ninth, and twelve
rounds. The feedback included a side-by-side display of the three individual
outcomes, though the sum of the three amounts was not displayed. Again,
the results are consistent with myopic loss aversion. Those playing the bets
in blocks of three took more risk, and as a result, earned higher returns. The
difference in allocations to the lottery varied between 10 and 20 percent,
depending on the specific rounds.
We have also conducted several experiments on myopic loss aversion
(Benartzi and Thaler 1999). In one of our tests, we gave thirty-nine under-
graduate students a choice between a certain amount and multiple plays of
a gamble. The subjects were presented with three gambles that had differ-
ent payoffs if played once, but virtually the same distribution of final pay-
offs if played repeatedly. The specific gambles are displayed below and the
certain amount was $3. Whereas a single play of the first gamble entails a
loss fifty times greater than that of the third gamble, the final distribution
of payoffs at the end of the multiple plays are virtually the same for the
three gambles.
- 90% chance to win $0.10
10% chance to lose $0.50
Played 150 times. - 50% chance to win $0.25
50% chance to lose $0.15
Played 120 times. - 10% chance to win $0.75
90% chance to lose $0.01
Played 90 times.
Traditional economic analysis predicts that the attractiveness of the three
gambles is approximately the same, because they all have the same distribu-
tion of final payoffs. Myopic loss aversion, however, predicts that the char-
acteristics of the single play of the gambles make a difference. In particular,
myopic loss aversion predicts that the potential loss of the single play has
an important role in decision making. The results are consistent with my-
opic loss aversion. Only 49 percent of the subjects preferred the first gam-
ble to the certain amount, in which they faced a potential loss of $0.50 on
each trial. In comparison, 64 percent and 75 percent preferred the second
and third gambles to the certain amount, where the potential loss per trial
was $0.15 and a penny, respectively. We should highlight that loss aversion
MYOPIC LOSS AVERSION 219