%0 Thesis
%9 Skripsi
%A Khoirun Nikmah, NIM.: 21106010049
%B FAKULTAS SAINS DAN TEKNOLOGI
%D 2025
%F digilib:71771
%I UIN SUNAN KALIJAGA YOGYAKARTA
%K Investasi, Risiko, VaR, Machine Learning, Random Forest, Support Vector Regressions, K-Nearst Neighbor
%P 151
%T ANALISIS RISIKO SAHAM SYARIAH MENGGUNAKAN MODEL MACHINE LEARNING DENGAN PENDEKATAN VALUE AT RISK (VAR)  (STUDI KASUS: SAHAM TLKM, BRIS, DAN ASII DALAM INDEKS SAHAM JAKARTA ISLAMIC INDEX 70 (JII70) PERIODE 1 SEPTEMBER 2020 - 31 DESEMBER 2023)
%U https://digilib.uin-suka.ac.id/id/eprint/71771/
%X Stocks are one of the most popular capital market instruments among investors, particularly in the context of sharia-compliant investments. In practice, understanding potential risks is essential to ensure that investment decisions remain aligned with sharia principles. One commonly used method to estimate potential losses is Value at Risk (VaR), which provides an estimate of the maximum loss within a certain time period and confidence level. The selection of the VaR method in this study is based on its ability to simplify risk information into a single numerical value that is easy to understand. With the advancement of technology, machine learning has increasingly been utilized to predict stock prices accurately and efficiently. This study aims to evaluate stock risk using machine learning models with a VaR approach. The study uses daily stock price data from the Jakarta Islamic Index 70 (JII70), focusing on three selected stocks: TLKM (Telkom Indonesia), BRIS (Bank Syariah Indonesia), and ASII (Astra International). The data period analyzed spans from September 1, 2020, to December 31, 2023. The dataset is divided using an 80% ratio for training data and 20% for testing data. The next stage is the application of machine learning models to predict stock prices. Several algorithms used in this study include Random Forest, Support Vector Regression (SVR), and K-Nearest Neighbor (KNN). Based on the predicted stock prices, returns are calculated, which then serve as the basis for risk analysis using the VaR approach. The analysis results show that SVR is the most accurate model based on the lowest RMSE and MAE values, making it the best model for risk estimation. The VaR calculation using SVR shows that BRIS has the highest risk and return potential compared to TLKM and ASII across various time horizons, making it more suitable for investors with a aggressive risk profile.
%Z Mohammad Farhan Qudratullah, S.Si., M.Si