This study aims to compare the traditional mean–variance model with Firefly and Simulated Annealing algorithm in terms of portfolio performance. For this purpose, price data of BIST Dividend 25 Index stocks between 2018–2023 were used. Expected return, portfolio risk, Sharpe Ratio, Coefficient of variation, and Downside Risk were used for performance. The Firefly and Simulated Annealing algorithm portfolios generally have higher return rates than average and therefore carry more risk. The Firefly portfolio generally performs well in Sharp Ratio and Downside Risk. Investors can diversify with Firefly and SAs and achieve excess-market returns in Borsa İstanbul.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Machine Learning in Portfolio Optimization

  • Diler Türkoğlu,
  • Melih Kutlu

摘要

This study aims to compare the traditional mean–variance model with Firefly and Simulated Annealing algorithm in terms of portfolio performance. For this purpose, price data of BIST Dividend 25 Index stocks between 2018–2023 were used. Expected return, portfolio risk, Sharpe Ratio, Coefficient of variation, and Downside Risk were used for performance. The Firefly and Simulated Annealing algorithm portfolios generally have higher return rates than average and therefore carry more risk. The Firefly portfolio generally performs well in Sharp Ratio and Downside Risk. Investors can diversify with Firefly and SAs and achieve excess-market returns in Borsa İstanbul.