PEMODELAN VOLATILITAS HARGA EMAS ANTAM BERBASIS DATA HARIAN MENGGUNAKAN MODEL GARCH

Abstrak

This study aims to model and forecast the volatility of Antam gold prices using the Generalized Autoregressive Conditional Heteroscedasticity (GARCH) model. The data consisted of daily Antam gold prices from January 2020 to May 2025, obtained from Investing.com, with a total of 1,446 observations. The analysis was conducted by transforming the data into logarithmic returns, performing descriptive statistical analysis, testing stationarity using the Augmented Dickey-Fuller (ADF) test, identifying the appropriate model through the Autocorrelation Function (ACF) and Partial Autocorrelation Function (PACF) plots, conducting the ARCH-LM test, estimating GARCH models, selecting the best model based on the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC), performing diagnostic tests, and evaluating forecasting performance using the Root Mean Square Error (RMSE) and Mean Absolute Error (MAE). The results indicate that the return series is stationary and exhibits ARCH effects, making it suitable for GARCH modeling. Based on the AIC value, the GARCH(1,2) model was identified as the best model with an AIC value of -4.350011. However, the GARCH(1,1) model was employed for forecasting because it is more parsimonious and satisfies the diagnostic tests. The forecasting evaluation yielded an RMSE of 0.03258798 and an MAE of 0.0241362, indicating good forecasting performance. Therefore, the GARCH model is effective for modeling the volatility of Antam gold prices and supporting investment decision-making.