Portfolio Optimization: An Introduction to Higher Order Moments

Abstract

This paper explores the possible improvements to portfolio construction techniques that the inclusion of skewness could bring. The main analysis has been done using the daily and monthly returns of 205 S&P500 components, ranging from 1990 to 2020, with a three-year rolling window approach for the selection of the stocks included in each portfolio. What this study will demonstrate is that the higher moments are necessary for portfolio optimization, and they can enhance the performances of the Markovitz approach that does not account for skewness as an input by itself