Abstract
The risky investment game of Gneezy and Potters :631–645, 1997) has been proposed as a simple tool to measure risk aversion in applied settings, especially attractive in settings where participants may have limited education. However, this game can produce a significant endowment effect, so that analysis of the behavior in this game should not be done in the Expected Utility Theory framework. The paper illustrates this point, by showing that risk tolerance can be much higher when the initial endowment concerns a risky lottery.