OPTIMASI PORTOFOLIO SAHAM IDX30 MENGGUNAKAN HIERARCHICAL CLUSTERING METODE WARD DENGAN PEMBOBOTAN ALGORITMA GENETIKA

Maria Artameivia Putri, Hendra Perdana, E. Sulistianingsih

Abstract

Stock portfolio is an investment consisting of stocks from different companies with the expectation that if the price of one stock decreases, the price of another stock may increase. To optimize returns and minimize risks, an investment strategy through portfolio optimization is required. This study aims to analyze the implementation of hierarchical clustering using the Ward method with genetic algorithm weighting in optimizing the IDX30 stock portfolio. The Ward method is used to group stocks based on similarities in financial ratios, while the genetic algorithm is used to determine optimal investment weights. The data used consist of 28 stocks included in the IDX30 index during the period from February 2024 to January 2025, with 6 stocks meeting the criteria of having positive expected returns and positive financial ratios. The financial ratio variables used in this analysis are Earnings per Share (EPS), Return on Equity (ROE), Debt to Equity Ratio (DER), and Price Earnings Ratio (PER). Based on the analysis results, the optimal IDX30 stock portfolio consists of PT Charoen Pokphand Indonesia Tbk (CPIN), PT Indofood Sukses Makmur Tbk (INDF), PT Perusahaan Gas Negara Tbk (PGAS), PT Bukit Asam Tbk (PTBA), and PT United Tractors Tbk (UNTR), with investment allocations of 0.23% for CPIN, 50.37% for INDF, 49.16% for PGAS, 0.12% for PTBA, and 0.13% for UNTR. The portfolio produces an expected return of 0.00112 and a portfolio risk of 0.01340. However, this study does not perform sensitivity analysis on the genetic algorithm parameters; therefore, future research may evaluate solution stability through parameter sensitivity testing.

Source: semanticscholar · PDF

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