The mean-variance framework: is there a superior portfolio selection strategy?

Abstract

More than 60 years has passed since Markowitz introduced the mean-variance framework for selecting optimal investment portfolios, but it has remained actual until nowadays. Despite its theoretical attractiveness, however, the mean-variance framework shows fallacies when applied to real world data due to the estimation error of its components. In this thesis I will compare different portfolio selection strategies, addressing the issue of improving the estimation of the variance-covariance matrix.