Maulidya Maghfiro, Trianingsih Eni Lestari
Palm oil plants process crude Palm Oil (CPO). CPO is an important commodity in Indonesia because Indonesia has the highest production rate in the world. The problem is that the price of CPO in Indonesia tends to fluctuate, leading to price instability. It is necessary to see the pattern of data from CPO prices with forecasting. In the analysis of time series data, there is a long memory phenomenon, so one of the methods used is the ARFIMA method. The ARFIMA method develops the ARIMA method with fractional value differencing. The ARFIMA method has residuals that meet white noise, normal distribution, and homogenity. However, some financial data have inhomogeneous residuals, so there is a need to test heteroskedasticity's effects. One of the proper methods of dealing with the effects of heteroskedasticity is the GARCH method. Another phenomenon that often occurs in the GARCH method is an asymmetric effect. FIGARCH is a development of the GARCH method for addressing data with asymmetric effects. The FIGARCH method is also very good in forecasting long memory effects. The aim of this research is to model and forecast the price of CPO using the ARFIMA-FIGARCH method in order to determine the future prices of CPO and facilitate decision-making for exporters and importers. The ARFIMA-FIGARCH method with the best model is ARFIMA(1, d, 3)-FIGARCH (1,d,1) with a differencing value (d) of 1.1355451 which is capable of forecasting with an AIC value of 16,290. © 2024 Author(s).
Mathematics Department, Faculty of Mathematics and Natural Sciences, Universitas Negeri Malang, Malang, Indonesia
