The supremacy of ARIMA over GARCH and ARFIMA models in modelling of Autocorrelated Kerosene-Type Jet Fuel Prices

Abstract

In response to the increasing and unpredictable volatility in energy markets, particularly in the price of jet fuel, robust predictive models have become essential for effective decision-making and risk management. This study, therefore, offers an in-depth comparison of various models, including ARIMA, GARCH-family (such as GARCH, TGARCH, and EGARCH), and ARFIMA, for analyzing and predicting jet fuel prices like kerosene, by utilizing monthly data from February 1990 to June 2025. The methodology commences with time series plots and statistical descriptions to evaluate trends, normal distribution, and market volatility. Findings from unit root tests indicate that the raw data is non-stationary, which leads to the need for log transformations and differencing. Significant autocorrelation is identified through ACF and PACF plots and LjungBox Q-tests, highlighting the presence of time dependency. To investigate long memory properties, the Geweke and Porter-Hudak (GPH) test was applied, which reveals short memory behavior in returns, thus supporting the use of ARIMA and GARCH models. Among these, ARIMA (3,1,3) excels in AIC/BIC metrics, making it the best choice for forecasting. On the other hand, GARCHfamily models offer a more profound understanding of volatility. The analysis of News Impact Curves demonstrates that GARCH is symmetric, while TGARCH and EGARCH effectively capture the leverage impact of adverse shocks. Notably, the EGARCH model, adept at handling asymmetry and nonlinearity, is found to be the most effective for volatility assessment. These insights are valuable in practice: ARIMA models are ideal for short-term price forecasting, whereas EGARCH offers essential insights into risk. This study equips energy marketers, aviation fuel users, and transportation companies with reliable tools for accurate price and risk forecasts amidst market fluctuation. Additionally, policymakers can leverage these results to enhance market stability and lessen the economic repercussions of price shocks