ANALISIS RISIKO SAHAM SYARIAH MENGGUNAKAN VALUE AT RISK (VAR) SIMULASI MONTE CARLO BERBASIS MODEL BI-DIRECTIONAL LONG SHORT TERM MEMORY (BI-LSTM)

Katya Royhana Dewi, NIM.: 22106010076 (2026) ANALISIS RISIKO SAHAM SYARIAH MENGGUNAKAN VALUE AT RISK (VAR) SIMULASI MONTE CARLO BERBASIS MODEL BI-DIRECTIONAL LONG SHORT TERM MEMORY (BI-LSTM). Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (ANALISIS RISIKO SAHAM SYARIAH MENGGUNAKAN VALUE AT RISK (VAR) SIMULASI MONTE CARLO BERBASIS MODEL BI-DIRECTIONAL LONG SHORT TERM MEMORY (BI-LSTM)) 22106010076_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version Download (3MB) | Preview
	Text (ANALISIS RISIKO SAHAM SYARIAH MENGGUNAKAN VALUE AT RISK (VAR) SIMULASI MONTE CARLO BERBASIS MODEL BI-DIRECTIONAL LONG SHORT TERM MEMORY (BI-LSTM)) 22106010076_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf Restricted to Registered users only Download (7MB) | Request a copy

Abstract

The fluctuating movement of the Jakarta Islamic Index (JII) indicates that Sharia-compliant stocks continue to exhibit market risk despite adhering to Islamic principles. This study aims to analyze the risk of the JII during the 2020–2025 period by integrating Bidirectional Long Short-Term Memory (BiLSTM), Monte Carlo simulation, and Value at risk (VaR). The data used in this study consist of the daily closing prices of the JII obtained from Yahoo Finance and transformed into log returns. A BiLSTM model was developed using a 60-day sliding window, a batch size of 32, and a maximum of 50 training epochs with an early stopping mechanism. The model predictions were then denormalized and used as the basis for a Monte Carlo simulation based on the Student-t distribution to estimate VaR at the 95% and 99% confidence levels. The results show that the BiLSTM model achieved a Mean Squared Error (MSE) of 0.000215, a Root Mean Squared Error (RMSE) of 0.014672, a Mean Absolute Error (MAE) of 0.010739, a Mean Absolute Percentage Error (MAPE) of 133.769363%, and a Directional Accuracy of 30.449827%. Fitting the residuals to the Student-t distribution produced a degree of freedom (df) of 4.903950, indicating a fat-tailed return distribution. The Monte Carlo simulation yielded a 1-day VaR of -0.029262 at the 95% confidence level and -0.048906 at the 99% confidence level. Backtesting using the Kupiec Proportion of Failures Test after calibration confirmed that both VaR estimates were statistically valid. Accordingly, the BiLSTM–Monte Carlo approach may serve as an alternative method for measuring the daily downside risk of the Jakarta Islamic Index.

Share this knowledge with your friends :

Actions (login required)

View Item