Utility Adjusted Cumulative Area Ratio Optimisation Technique (UCAROT)

Abstract

In this paper we propose a novel portfolio optimization technique that draws its inspiration from the concept of “almost stochastic dominance”. The technique is computationally parsimonious and applicable for all return distributions and all strictly increasing utility functions. First, we develop the concept of Utility Adjusted Cumulative Area Ratios (UCAR) to make pairwise comparisons of the individual assets. This technique involves adjusting the cumulative distributions of asset returns for the investor utility function. We then input this information into an NxN Analytic Hierarchy Process (AHP) matrix to determine optimum weights. As an example we apply it to the BIST-30 Turkish Index. We form five portfolios based on various utility functions: risk averse with decreasing, increasing and constant absolute risk aversion (DARA, CARA and IARA); risk neutral; and risk seeking. We find that all these portfolios convincingly outperform the Index according to conventional mean-variance criteria as well as second order stochastic dominance.