Discrete fuzzy system orbits as a portfolio selection method

The purpose of this work is to approach the portfolio selection problem from a particular System Theory framework. The System will be formed by the set of public companies in the portfolio and a set of fuzzy relations on those companies. Starting with an equally split portfolio represented by a fuzzy set B, the orbit of B is computed for a particular period obtaining a portfolio to invest in the next period. We present an example finding nine portfolios to invest in 9 months and we compare them with some optimal portfolios in the efficient frontier given by the Modern Portfolio Theory and with some random generated portfolios. We find a better performance in returns for the portfolio based on the systemic method.