Analisis Portofolio Saham Optimal Model Mean Absolute Deviation Berdasarkan Hasil Peramalan Menggunakan Metode Support Vector Regression - Genetic Algorithm

Safira, Regita Adelia (2025) Analisis Portofolio Saham Optimal Model Mean Absolute Deviation Berdasarkan Hasil Peramalan Menggunakan Metode Support Vector Regression - Genetic Algorithm. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5006211050-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (7MB) | Request a copy

Abstract

Pasar modal Indonesia telah mengalami perkembangan pesat, dengan jumlah investor pada akhir Januari 2025 telah menembus 15 juta single investor identification (SID). Salah satu instrumen investasi paling popular adalah saham. Bursa Efek Indonesia (BEI) menyediakan indeks saham seperti IDXHIDIV20 yang berisi saham-saham dengan dividend yield tinggi. Peramalan pergerakan harga saham sangat penting untuk membantu investor dalam pengambilan keputusan, mengingat tingginya volatilitas di pasar modal. Selain itu, investor juga dihadapkan pada tantangan untuk mengelola portofolio yang optimal guna memaksimalkan keuntungan dan meminimalkan risiko. Maka dari itu, dilakukan penelitian untuk mengestimasi Value at Risk (VaR) dan Conditional Value at Risk (CVaR) dengan simulasi Monte Carlo dari portofolio saham optimal model Mean Absolute Deviation (MAD). Estimasi risiko berdasarkan data aktual dan hasil peramalan harga saham selama satu bulan ke depan menggunakan metode Support Vector Regression (SVR) dengan optimasi Genetic Algorithm (GA) yang efektif untuk menangkap pola nonlinier yang sering ada pada data harga saham. Data yang digunakan dalam penelitian ini adalah data closing price harian lima saham yang konsisten tergabung dalam indeks IDXHIDIV20 selama periode 3 Januari 2022 hingga 31 Januari 2025. Hasil penelitian menunjukkan bahwa model SVR-GA memberikan performa prediksi lebih baik dibandingkan SVR tanpa optimasi dengan selisih nilai MAPE pada data testing berada di rentang 0,00013% hingga 0,02276% sehingga model SVR-GA dipilih sebagai model terbaik untuk meramalkan harga saham. Portofolio optimal model MAD dibentuk dari tiga saham dengan expected return positif, yaitu BBCA (41,08%), INDF (44,24%), dan PTBA (14,68%). Estimasi risiko menggunakan Monte Carlo menunjukkan bahwa nilai VaR dan CVaR portofolio pada tingkat kepercayaan 95% masing-masing sebesar 1,487853% (VaR) dan 2,021914% (CVaR), sedangkan pada tingkat kepercayaan 99% sebesar 2,322626% (VaR) dan 2,817444% (CVaR). Hasil backtesting membuktikan bahwa pengukuran risiko yang dilakukan telah valid. Selain itu, portofolio MAD terbukti memberikan risiko yang lebih rendah dibandingkan masing-masing saham penyusunnya. Hasil dari penelitian ini dapat digunakan bagi investor sebagai referensi dalam menyusun portofolio dan mengelola risiko investasi.
======================================================================================================================================
The Indonesian capital market has experienced rapid growth, with the number of investors at the end of January 2025 has reached 15 million single investor identification (SID). One of the most popular investment instruments is stocks. The Indonesia Stock Exchange (IDX) provides stock indices such as IDXHIDIV20, which contains stocks with high dividend yields. Forecasting stock price movements is essential to assist investors in decision-making, given the high volatility in the capital market. In addition, investors are also faced with the challenge of managing an optimal portfolio to maximize returns and minimize risks. Therefore, this research will be conducted to estimate Value at Risk (VaR) and Conditional Value at Risk (CVaR) with Monte Carlo simulation of an optimal stock portfolio using Mean Absolute Deviation (MAD) model. The risk estimation is based on actual data and results of stock price forecasting for one month ahead using the Support Vector Regression (SVR) method with Genetic Algorithm (GA) optimization, which is effective in capturing nonlinear patterns that often found in stock price data. The data used is the daily closing price data of five stocks that are consistently included in the IDXHIDIV20 index during the period from January 3, 2022 to January 31, 2025. The results of the study indicate that the SVR-GA model provides better prediction performance than SVR without optimization, with a MAPE testing difference is in the range from 0,00013% to 0,02276%. Thus, SVR-GA is selected as the best model for stock price forecasting. The optimal portfolio of the MAD model is formed from three stocks with positive expected returns, BBCA (41,08%), INDF (44,24%), and PTBA (14,68%). The risk estimation using Monte Carlo shows that the VaR and CVaR of the portfolio at a 95% confidence level is 1,487853% (VaR) and 2,021914% (CVaR), respectively, while at a 99% confidence level is 2,322626% (VaR) and 2,817444% (CVaR). The backtesting results prove that the risk measurement is valid. Additionally, the MAD portfolio has been proven to provide lower risk compared to each of its constituent stocks. The results of this research can be used by investors as a reference in constructing portfolios and managing investment risks.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Genetic Algorithm, Mean Absolute Deviation, Monte Carlo, Support Vector Regression, Value at Risk
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. Q Science > QA Mathematics > QA276 Mathematical statistics. Time-series analysis. Failure time data analysis. Survival analysis (Biometry) Q Science > QA Mathematics > QA353.K47 Kernel functions (analysis) Q Science > QA Mathematics > QA401 Mathematical models. Q Science > QA Mathematics > QA76.9.D343 Data mining. Querying (Computer science)
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Actuaria > 94203-(S1) Undergraduate Thesis
Depositing User:	Regita Adelia Safira
Date Deposited:	31 Jul 2025 03:37
Last Modified:	31 Jul 2025 03:37
URI:	http://repository.its.ac.id/id/eprint/123148

Actions (login required)

View Item