ANALISIS RISIKO SAHAM SYARI’AH MENGGUNAKAN METODE LONG SHORT-TERM MEMORY (LSTM) DENGAN PENDEKATAN VALUE AT-RISK (VaR)

Nur Aisyiah Jamil, NIM.: 21106010079 (2026) ANALISIS RISIKO SAHAM SYARI’AH MENGGUNAKAN METODE LONG SHORT-TERM MEMORY (LSTM) DENGAN PENDEKATAN VALUE AT-RISK (VaR). Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

	Text (ANALISIS RISIKO SAHAM SYARI’AH MENGGUNAKAN METODE LONG SHORT-TERM MEMORY (LSTM) DENGAN PENDEKATAN VALUE AT-RISK (VaR)) 21106010079_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf Download (3MB)
	Text (ANALISIS RISIKO SAHAM SYARI’AH MENGGUNAKAN METODE LONG SHORT-TERM MEMORY (LSTM) DENGAN PENDEKATAN VALUE AT-RISK (VaR)) 21106010079_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf Restricted to Registered users only Download (4MB) | Request a copy

Abstract

This study aims to analyze investment risk in Islamic stocks by applying the Long Short-Term Memory (LSTM) method and the Value at-Risk (VaR) approach. The research objects include the stocks of PT Aneka Tambang Tbk (ANTM) and PT Bumi Resources Minerals Tbk (BRMS), which are listed in the Indonesian Sharia Stock Index (ISSI). These stocks were selected because gold prices in Indonesia have tended to increase over several periods, thereby generating positive sentiment toward gold mining stocks. In addition, the price data exhibit more upward movements than declines, making them suitable for further analysis. The data used consist of daily closing prices from June 5, 2023, to May 28, 2025, obtained from the Indonesia Stock Exchange (IDX). The LSTM model is employed to predict stock price movements based on historical patterns using variations in the number of neurons (5, 10, and 20), epochs (50 and 100), and optimizers (RMSprop, Adam, and AdamW). Model performance is evaluated using Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE). The evaluation results show that the MAE values for the analyzed stocks are below 0.05, indicating good predictive performance. Furthermore, investment risk is analyzed using the Value at Risk (VaR) method through the Monte Carlo simulation approach, which is categorized as a non-parametric method. The validation test using the Likelihood Ratio (LR) indicates that the VaR estimates for 7-day and 30-day horizons are inaccurate, as the LR values exceed 3.841, while the 1-day VaR is considered accurate because the LR value is below 3.841. This study is conducted based on the consideration that investment in Islamic stocks should not only focus on returns but also take into account the potential risks involved. Overall, the results demonstrate that the combination of LSTM and VaR methods is effective in predicting stock price movements and measuring the level of investment risk in Islamic stocks.

Share this knowledge with your friends :

Actions (login required)

View Item