A Clonal Selection Algorithm for the Design of an Optimal Investment Portfolio

One of the applications of artificial intelligence is combinatorial or numerical optimization by means of computational algorithms. This proposal documents the obtainment process of an optimal investment portfolio of 10 assets of the Mexican Stock Exchange from the diversification of the amounts to be invested, which is calculated by a population based metaheuristic called immunological algorithm of clonal selection, a technique supported in the evolutionary principle of the immune system of upper mammals, whose objective function was formulated as maximizing the gain-risk ratio described by the Markowitz Theory of Investment Portfolios, the possible solution vectors, it is to say, the amounts to be invested, are encoded with real numbers. In this proposal, the results obtained were analyzed with statistical tests which allowed to determine the optimal parameters and stability of the proposed algorithm that made possible to obtain the maximum profit-risk ratio.