Portfolio optimization using batch enabled particle swarm optimization
Deepika Kaushik, Mohammad Nadeem · 2025
Portfolio Optimization (PO) is a prominent optimization problem in computational finance. A portfolio is the collection of assets and the weight assigned to each asset needs to be optimized to maximize the returns and minimize the risks. In the proposed work, the Markowitz model also referred to as the Mean-Variance (MV) model is adopted for solving this multi-objective problem. The study used an improved version of Particle Swarm Optimization (PSO) in which, the population is divided into sub-populations(batches) each of which executes independently without any migration among themselves. The portfolio has been optimized for Dow Jones and NASDAQ100 datasets. The results obtained are superior to other state-of-the-art algorithms performing well for this problem. Thus, proving the virtue of the used approach for PO.