Autoregressive distributed lag modelling for Malaysian palm oil prices

Abstract

Modelling food commodities prices has become the area of interest in financial time series. This study aims to model Malaysian average monthly prices of crude palm oil using dynamic regression approach. The sample period covers from January 2000 until December 2013. The model investigated is Autoregressive Distributed Lag (ARDL) model. The model uses multivariate analysis with monthly prices, productions, imports, exports and closing stocks of crude palm oil as the variables. The ARDL model is selected using Akaike Information Criteria (AIC) and Schwartz-Bayesian Criteria (SBC). The capabilities of this model in estimating the crude palm oil prices is compared to Box-Jenkins Autoregressive Integrated Moving Average (ARIMA) model by using Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE). The process of modelling is done by using Eviews and Microfit statistical software. This study concluded that ARDL model is a better model in modelling the palm oil prices. The ARDL model selected by using AIC produce better estimation than the ARDL model selected by using SBC. Furthermore, there exist long-run relationship between crude palm oil prices and its determinants.