Ambiguity and gender differences in financial decision making : an experimental examination of competence and confidence effects

This paper reports the results of an experiment that brings together psychological measures of competence and overconfidence with laboratory economic measures of individual valuations of uncertainty. We examine the valuations of risky and ambiguous lotteries in a financial decision context. The experiment can be viewed in two parts. The first part replicates an experimental design reported by Heath and Tversky (1991) but within a financial market context. This part produces two measures: 1) competence, the perception of feeling knowledgeable or competent in an area and 2) overconfidence, the well documented result that many individuals overestimate their ability. These measures, together with an indicator of objective knowledge, were used to explain elicited certainty equivalents in the second part of the experiment. Certainty equivalents were elicited for lotteries that were contingent on the price movements of real stock and bond funds, the price changes of simulated virtual funds, and pure risk lotteries. These represent three different levels of uncertainty: two-sided ambiguity, one-sided ambiguity and pure risk. Our results show a significant relationship between individual overconfidence and competence measures and elicited values of lotteries in a financial decision context. Further, the interaction of overconfidence, competence and knowledge measures with gender produce nearly opposite effects.