Algorithm of construction of optimum portfolio of stocks using genetic algorithm

• DOI: 10.1007/s13198-014-0293-7
• Publication Date: 2014-09-19T11:45:21Z
• Authors: Pankaj Sinha; Abhishek Chandwani; Tanmay Sinha
• Keywords: Optimum Portfolio, Genetic Algorithm, Portfolio Construction, MATLAB; jel:C63; jel:C61; 0202 electrical engineering, electronic engineering, information engineering; 02 engineering and technology; jel:C6; jel:G11

The objective of this paper is to develop an algorithm to create an optimum portfolio from a large pool of stocks listed in a single market index SPX 500 Index: USA (for example) using genetic algorithm. The algorithm selects stocks on the basis of a priority index function designed on company fundamentals, and then genetically assigns optimum weights to the selected stocks by finding a genetically suitable combination of return and risk on the basis of historical data. The effect of genetic evolution on portfolio optimization has been demonstrated by developing a MATLAB code to implement the genetic application of reproduction, crossover and mutation operators. The effectiveness of the obtained portfolio has been successfully tested by running its performance over a 6 month holding period. It is found that genetic algorithm is successful in providing the optimum weights to stocks which were initially screened through a predetermined priority index function. The constructed portfolio beats the market for the considered holding period by a significant margin.