Genetic algorithms for fuzzy multi-objective approach to portfolio selection

This research deals with a model with better efficiency for selection of portfolio making use of cardinal constraints, which are explained in previous sections. Such a method, which is a combination of fuzzy models and MCDM considering the constraints intended by investors, has not been used in previous models. We have considered transactions cost, because they are among factors important for an investor, and their being ignored in a portfolio selection method will result in inefficient portfolio. Sector value constraint is among other constraints considered here. Such a constraint aims to raise investment rate in sectors with higher values. Cardinal constraints (number of shares existing in a portfolio and shares weight constraints), minimum purchase rate (for prevention of very small investments) and maximum purchase rate (for absorption of diversified and sufficient investment rates) are also added to the proposed method, which in turn results in increased model efficiency and its proximity to reality. A genetic algorithm has been suggested for solving the model, the results of which imply increased efficiency of the problem considering transaction cost as well as increased shares in portfolio.