Investment Decisions and Negative Interest Rates

Abstract

Prospect Theory’s value function suggests that investors would be risk averse in the gain domain and risk seeking in the loss domain—that is, the reflection effect. However, most of the experimental evidence relies on choice tasks in the gain domain between prospects marked in dollar amounts and considering non-mixed lotteries. There is not much work that examines environments with properties typical in investment decisions where the task is fund allocation involving mixed lotteries with outcomes being rate of return. The recent negative deposit rates in Europe demonstrate the importance of this question and, in particular, understanding investment decisions in the loss domain. This paper fills this gap by using a series of laboratory experiments mimicking these properties of investment decisions with a range of investment amounts and with the money to invest either being earned and literally on the table or not. Yet, no matter the settings, we find no evidence for the reflection effect in investment, and behavior is most consistent with maximizing expected return. This holds regardless of the language used (abstract or not), whether we use a two- or a three-state lottery, whether the task is continuous rather than discrete, or the risky portfolio is a mixed lottery. This paper was accepted by Chen Yan, decision analysis.