CONSTRAINED PORTFOLIO SELECTION OPTIMIZATION USING CALIBRATED GENETIC ALGORITHM

Abstract

Portfolios are a proper collection of investments choosing by an organization or a person. Hence portfolio optimization is a very prominent problem by optimizing which we can attain more profit with the less risk. In this paper, we consider a portfolio optimization problem with some constraints such as boundary and cardinality constraints. Although adding these constraints makes the model more adapted to real world problems, it leads to disability of exact methods to find the optimal solutions in a reasonable time. In order to find optimal solutions, we use an improved genetic algorithm inspiring by natural evolution. We indicate that our proposed algorithm is more efficient and effective than many genetic algorithms which have been previously presented. Parameter setting of the proposed algorithm is done by means of a statistical method based on Taguchi technique. Finally, we perform some computational experiments which confirm the efficiency of our algorithm for portfolio optimization.

Keywords

• portfolio selection
• genetic algorithm
• portfolio optimization
• constrained portfolio
• taguchi technique