A simple nonparametric approach to low-dimension, shortfall-based portfolio selection

This paper develops a simple, low-dimension portfolio selection rule based on minimizing the probability of realizing a return below some pre-determined benchmark or target rate. Unlike most shortfall-based methods, which employ approximations to the shortfall probability, this method operates directly on the complementary Heaviside function representation of the in-sample shortfall probability. Thus, no behavioral assumptions, other than the notion of shortfall minimization, enter the portfolio selection process.