Abstract:

We estimate a structural model using data from a novel experiment investigating how investors’ preferred frequency of portfolio evaluations balance the opposing effects of ambiguity and loss aversion. Investors in the experiment face initial ambiguity concerning return distributions for an asset. They observe draws from the true return distribution of the asset, allowing them to reduce their ambiguity through time. We exploit portfolio choices and stated beliefs over possible return distributions to estimate preferences and ambiguity updating rules. We find that 70% of investors would opt for a high frequency of portfolio evaluations, reflecting the dominating effect of ambiguity aversion over loss aversion.