Natural rubber production forecasting of the world market with Box-Jenkins method

Abstract

             The objective of this study is to develop model and predict the number of natural rubber production in the world market with Box and Jenkins approach by using the monthly time series data from 2007 to 2016. The monthly time series data were separated into four periods that are from (1) 2007 to 2016, (2) 2009 to 201 6, (3) 2012 to 2016, and (4) 2014 to 2016 for forecasting. After that the study compares the results by root mean square error (RMSE) – the period which gets the lowest in the RMSE.

             The result of this study, the lowest of all periods in RMSE is the monthly data from 2014 to 2016, and the fitted model of this study is AR(1) AR(2) AR(4) MA(1) MA(3) SMA(12) or ARMA(4,3)(0,1)₁₂ that has RMSE in dynamic forecasting and static forecasting equal to 13.313 and 8.625, respectively. The average of yearly predicted future data for ten years from 2018 to 2027, the trend of number of natural rubber production decreases from 829.97 thousand tons in 2018 to 638.18 thousand tons in 2027, and the percent of change reduces by 23.12 percent.

               As a consequence, the supply of natural rubber in the world market tends to reduce the positive impact on natural rubber price of the world market. However, the sustainability of natural rubber market depends on the balance of natural rubber supply and demand in the world market in the future.