Is Self-Reported Risk Aversion Time Varying?

Seeun Jung (1, 2) , Carole Treibich (1, 2, 3)

1 PSE - Paris-Jourdan Sciences Economiques
2 PSE - Paris School of Economics
3 GREQAM - Groupement de Recherche en Économie Quantitative d'Aix-Marseille

We examine a Japanese Panel Survey in order to check whether self-reported risk aversion varies over time. In most panels, risk attitude variables are collected only once (found in only one survey wave), and it is assumed that self-reported risk aversion reects the individual's time-invariant component of preferences toward risk. Nonetheless, the question could be asked as to whether the financial and macro shocks a person faces over his lifetime modify his risk aversion. Our empirical analysis provides evidence that risk aversion is composed of a time-variant part and shows that the variation cannot be ascribed to measurement error or noise given that it is related to income shocks. Taking into account the fact that there are time-variant factors in risk aversion, we investigate how often it is preferable to collect the risk aversion measure in long panel surveys. Our result suggests that the best predictor of current behavior is the average of risk aversion, where risk aversion is collected every two years. It is therefore advisable for risk aversion measures to be collected every two years in long panel surveys.

Mots clés: Risk Aversion; Panel Data

Domaines: Economies et finances

Seeun Jung, Carole Treibich. Is Self-Reported Risk Aversion Time Varying?. 2014. ⟨halshs-00965549v2⟩