Title: Prediction of Daily Price of Jakarta Islamic Index Using Hybrid ARIMA-GARCH Model
Authors: Najwa Khoir Aldawiyah, M. Fariz Fadillah Mardianto, Elly Pusporani, Sediono
Volume: 9
Issue: 12
Pages: 299-306
Publication Date: 2025/12/28
Abstract:
The objective of this research is to model and forecast the daily closing prices of the Jakarta Islamic Index by comparing the ARIMA and hybrid ARIMA-GARCH approaches. The data are obtained from Investing.com and cover the period from 1 August 2023 to 31 August 2025, with the sample split into 80% training data and 20% testing data. The empirical results show that the ARIMA (1,1,0) model achieves strong in-sample accuracy with a MAPE of 0.79%, but its out-of-sample performance deteriorates substantially, as reflected by a testing MAPE of 15.40%. In contrast, the hybrid ARIMA (1,1,0)-GARCH (1,1) model consistently outperforms the standard ARIMA specification, yielding lower MAPE of 0.23% in the training period and 10.29% in the testing period. These findings imply that incorporating conditional heteroskedasticity improves forecast accuracy and provides more reliable predictions in volatile market environments.