A Heuristic Approach to Minimal Risk Portfolio Selection

C.A. Coutiño, B.Ma. Villarreal, and J.A. Torres (Mexico)


Portfolio Selection, Heuristic Methods, Optimization


This paper presents a group of techniques (hillclimbing, genetic algorithm and simulated annealing) for portfolio selection based on evolutionary programming as a tool for optimization. The goal is to find the mix of stocks that minimize risk expressed as standard deviation. The results showed that simulated annealing found the best portfolios which can be considered as a good alternative for the investor to make decisions. This research presents a real world application at the Mexican Stock Exchange. The heuristic algorithms were implemented based on the Markowitz Model where the investor can select the size of the portfolio as well as the expected return.

Important Links:

Go Back