Ambiguity, Prudence and Optimal Portfolio

Abstract

This paper extends the robust mean-variance analysis of Maccheroni et al. (2013) by investigating the contribution of ambiguity prudence to the optimal stock allocation when the investor evaluates portfolio compositions as described by the smooth model under ambiguity criterion. Ambiguity prudence captures an aversion towards model uncertainty that increases the more the investor believes that unfavourable events are likely to realise. I derive a higher-order approximation of the certainty equivalent to disentangle the contribution of preferences and beliefs. I analyse two relevant portfolio problems to show that ambiguity prudence puts non-linearities into the investor's valuation that induce sizeable variations from the robust mean-variance solution.