Towards Portfolio Selection Strategy Using Cultural Algorithm Based Solution Approach

Abstract

One of the critical issues in financial management is portfolio selection and optimization. It seeks to determine the optimal resource allocation for a group of assets. Since Harry Markowitz established the traditional Mean- Variance model in 1952 and William Sharpe subsequently refined it, this subject has been researched, and several models have been put forward. The effectiveness of nature-inspired algorithms in solving challenging computational optimization problems has prompted academics to create and use these algorithms for a range of optimization issues. This study proposes an unconstrained portfolio optimization strategy using a cultural algorithm (CA) to maximize the Sharpe ratio. The cultural algorithm is an evolutionary algorithm. It includes both the population and knowledge components (belief space). The experimental evaluation of the suggested strategy is shown by comparative analysis with the genetic algorithm (GA) performance. The proposed technique has produced very competitive results on the standard benchmark dataset, namely, DAX 100, Hang Seng 31, FTSE 100, and S&P 100 employed in our study.