Portfolio selection based on technical trading rules optimized with a genetic algorithm

In this paper, we propose a portfolio selection method based on a set of technical trading rules, which are optimized by a genetic algorithm. The aim of the research was to check if it is possible to obtain a set of trading rules deriving from technical indicators, which could be used to create a portfolio able to outperform standard portfolio models based upon Modern Portfolio Theory. On the contrary to the typical portfolio approach incorporating expected return and variance, presented method relies on market momentum analysis and stock timing using selected technical indicators.