A GARCH Model for Forecasting Volatility of Oil and Gas Exports in Indonesia

Abstract

This research aims to analyze the dynamics of oil and gas exports in Indonesia during the period from 2012 to 2022 using the GARCH approach for forecasting volatility. The data utilized in this study encompass the monthly published volumes of Indonesia's oil and gas exports, sourced from the Indonesian Central Statistics Agency's website. The analysis involves a substantial amount of data, comprising 132 monthly time series spanning a significant timeframe. The findings indicate that the most suitable model for predicting oil and gas volumes is the GARCH (1,1) model. The GARCH approach is employed to model the volatility within the data of oil and gas exports. The results reveal the utilization of information criteria, including Akaike (14.73), Bayes (14.86), Shibata (14.73), and Hannan-Quinn (14.79). Moreover, the forecast analysis for the next ten periods depicts a consistent upward trend. Generally, these forecast results suggest that while the mean values of the data remain relatively stable, the volatility levels are anticipated to increase over the forthcoming periods. The implications of this research are crucial within the context of economic and international trade, as the volatility in oil and gas exports can significantly impact national economic policies and corporate decisions.