Comparison of Markowitz and Genetic Algorithm Models for Saudi Arabian Stocks

Mega Tri Candeni, D. Waluyo, Ana Kadarningsih, Yenny Ernitawati, Dwi Eko, Waluyo

Abstract

This study aims to compare the Markowitz method and the Genetic Algorithm in forming an optimal portfolio in the Saudi Arabian stock market, which is known to have dynamic characteristics and high levels of volatility due to the influence of economic reforms, oil price fluctuations, and integration with global markets. The study uses daily closing price data for companies indexed in the Tadawul All Share Index for the period January 2, 2022, to October 30, 2025, obtained through Yahoo Finance. Sample selection uses a purposive sampling method based on data completeness and Coefficient of Variation selection. The optimization process is carried out using the Python programming language through the Efficient Frontier Markowitz approach and the Genetic Algorithm. While portfolio performance evaluation is carried out using the Sharpe Ratio, Sortino Ratio, and Omega Ratio to obtain a more comprehensive assessment of risk and return efficiency. The results show that the Markowitz method tends to produce higher expected returns, but with a greater level of risk, while the Genetic Algorithm produces a portfolio with a relatively lower level of risk and more stability. Furthermore, conventional stock portfolios performed better than Islamic stock portfolios based on the performance ratio evalu-ation used. This finding suggests that portfolio optimization methods should be tailored to investor risk preferences and the characteristics of the investment market.

Source: semanticscholar · PDF

Read the AI summary, key takeaways and discussion on WOBR Quant Research.