Portfolio optimization with mean-variance approach using hunting search meta-heuristic algorithm

Abstract

This paper presents a new meta-heuristic solution to find the efficient frontier using the mean-variance approach. Portfolio optimization problem is a quadratic programming model and, changes to NP-hard if the number of assets and constraints has increased, and it cannot be solved using common mathematical methods in a reasonable time. Therefore, a heuristic or meta-heuristic algorithm should be used that is the appropriate solution. This paper optimizes portfolio using a new meta-heuristic algorithm called hunting search algorithm. To determine the strengths and precision of proposed algorithm, a case study is designed using Iran stock market data from 1/3/1389 to 1/3/1390 for big thirty companies. The proposed algorithm finds the efficient frontier precisely and in timely manner. To determine abilities of the algorithm, two verified examples, Hang Sang 31 and Dax100 are also solved with it. Results show that hunting search algorithm has a high speed and high accuracy in order to solve portfolio optimization problems, and it can be used to find the efficient frontiers in various portfolio optimization problems.

Keywords

• portfolio optimization
• markowitz mean–variance model
• hunting search algorithm