Mean-Variance and Scenario-Based Approaches to Portfolio Selection

Abstract

In this cautionary tale, the sorcerer's apprentice uses his mentor's magic before he knows how to control it. Matters quickly get out of hand. The author argues that the scenario-based expected utility maximization approach to portfolio optimization presents similar opportunities for misadventure. He shows how to avoid the danger, Alas, as with all sorcery, when the illusion is stripped away one sees that there is less there than initially supposed. The conventional mean-variance approach gives comparable answers with less bother and peril.