Future price information is crucial not only for producers but also for other agribusiness actors. Price is a signal for them to make a decision regarding what to produce and when to sell including for natural rubber. For this reason, forecasting and selecting the best model becomes important. This study is aimed to analyze and select the possible forecasting methods for monthly natural rubber prices in Indonesia and World Markets. The univariate model of Double Exponential Smoothing, Decomposition, and ARIMA models are applied to forecast price data from 2012:1 – 2016:12. The selection of an accurate model is based on the lowest value of MAPE, MSD, and MAD. ARIMA is the possible methods for world rubber price forecasting while Double Exponential Smoothing should be applied for predicting domestic rubber prices because it allows for a better predictive performance.
Volume 11 | Issue 7
Pages: 387-395