Testing Ambiguity Models through the Measurement of Probabilities for Gains and Losses

Abstract

This paper reports on two experiments that test the descriptive validity of ambiguity models using a natural source of uncertainty (the evolution of stock indices) and both gains and losses. We observed violations of probabilistic sophistication, violations that imply a fourfold pattern of ambiguity attitudes: ambiguity aversion for likely gains and unlikely losses and ambiguity seeking for unlikely gains and likely losses. Our data are most consistent with prospect theory and, to a lesser extent, α-maxmin expected utility and Choquet expected utility. Models with uniform ambiguity attitudes are inconsistent with most of the observed behavioral patterns. (JEL D81, D83, G11, G12, G14)

Estimating ambiguity preferences and perceptions in multiple prior models: Evidence from the field

“…To capture such reference dependence, we adjust the α-MaxMin model by introducing separate ambiguity aversion parameters for gains and losses. Our results are consistent with laboratory evidence, such as Baillon and Bleichrodt (2015) who find that ambiguity aversion for losses and gains differ, and Kothiyal et al (2014) who provide support for reflection.…”

“…To examine reference dependence, we test the null hypothesis that mean AA 50 = meanAA −50 . A paired samples t -test strongly rejects the null hypothesis of ‘ no reference dependence ’ ( p-value <0.01); accordingly, ambiguity aversion differs for gains and for losses in the general population, consistent with the experimental results of Cohen et al (1987), Abdellaoui et al (2005), and Baillon and Bleichrodt (2015). We next test for reflection, or mean AA 50 =− meanAA −50 , which implies that ambiguity aversion for gains is reflected into ambiguity-seeking behavior for losses.…”

The Ambiguity Triangle: Uncovering Fundamental Patterns of Behavior Under Uncertainty

“…Further, in a setting where lower envelope lotteries are defined on complementary events, our setup could also shed further light on violations of probabilistic sophistication. Previous experimental work by Baillon and Bleichrodt (2015) indicates that the additivity of disjoint events required under probabilistic sophistication fails when eliciting matching probabilities for naturally occurring uncertainties.…”

The strength of sensitivity to ambiguity

“…An alternative empirical strategy for identifying our preference parameters to the one that we have used in this paper would be to apply a matching probability method, akin to one explored in recent literature (e.g., Baillon and Bleichrodt 2015;Dimmock et al 2015Dimmock et al , 2016.…”