ARIMA Model in Predicting Jakarta Composite Index

Abstract

This study discusses stock price modeling using ARIMA model. We apply to model to the Jakarta Composite Index (JCI) as it represents all stock performances listed in Indonesia Stock Exchange. In this study, we propose several ARIMA models based on the daily from June 10th, 2019 until December 6th, 2019. The parameters among the models are estimated by using RStudio. We chose the best model by considering its AIC and RMSE. The best model that is ARIMA (21, 1, 2) with 99% confidence interval. This model is then used to predict the next 15 days (December 09, 2019 to January 02, 2020).