A Stream of Prospects or a Prospect of Streams: On the Evaluation of Intertemporal Risks

Abstract

Agranov, and conference participants at the American Social Science Association Annual Meeting and the Southwest Experimental and Behavioral Economics Conference. Andreoni also thanks the National Science Foundation for financial support. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research. NBER working papers are circulated for discussion and comment purposes. They have not been peer-reviewed or been subject to the review by the NBER Board of Directors that accompanies official NBER publications.

Intertemporal Consumption With Risk: A Revealed Preference Analysis

“…Table 7 shows that 67.96 percent of participants have a higher efficiency level for correlation aversion The support we find in our data for correlation aversion as opposed to correlation neutrality suggests that the behavior of our experimental subjects is not well represented by the DEU model, and more generally, by any model that is separable across time (see ( 12)). On the other hand, there is experimental evidence indicating that a subject's preference for lotteries that pay out at time t are unaffected by the presence of lotteries, with independent realizations, in a future period t ′ , which supports a time-separable utility formulation (see Andreoni et al (2017)). In fact, it is possible to reconcile these results because there are utility functions that satisfy correlation aversion and are also time-separable when lotteries are independent.…”

“…11 This paper has given rise to interesting follow up discussion. See Cheung (2015), Miao and Zhong (2015), Epper and Fehr-Duda (2015), Andreoni, Feldman, and Sprenger (2017).…”

“…The evidence put forth provides useful guidance to researchers working on models of choice over time and risk. The ideal model should allow for (not necessarily require) correlation aversion and stochastic impatience; it should also be consistent with violations of RSTL (see DeJarnette et al (2019)) and perhaps permit some limited degree of time separability (see Andreoni et al (2017)).…”

“…A discussion found in Andreoni and Sprenger (2012a), Epper and Fehr-Duda (2015), and Andreoni et al (2017) concerns whether utility should be time-separable or stateseparable. Andreoni et al (2017) observe that with the time-separable formulation, an agent's preference over lotteries that pay out in period t are unaffected by the presence of lotteries that pay out independently in some other time period t ′ > t; they refer to this property as common future risk invariance (CFRI). Andreoni et al (2017) find some support for this behavior in their experiment.…”

Intertemporal Prospect Theory

“…The overall PT value is then calculated as the atemporal PT value of the present values. This method (or a version of it with rank-dependent utility, not distinguishing between gains and losses but including probability weighting) is used in the studies by Chew and Epstein (1990), Bleichrodt and Eeckhoudt (2006), Halevy (2008), Drouhin (2015), Epper and Fehr-Duda (2015), Blavatskyy (2016), Andreoni et al (2017), and Andersen et al (2018).…”