Estimation methodology of short-term natural rubber price forecasting models


Zainalabidin, M. and Mohamed, A. Fatimah and Chong, E. C. F. and Mad Nasir, S. and Khin, A. A. (2011) Estimation methodology of short-term natural rubber price forecasting models. Journal of Environmental Science and Engineering, 5 (4). pp. 460-474. ISSN 1934-8932

[img] Text
Restricted to Repository staff only

Download (218kB)


This study developed a short-term econometric model of world natural rubber price Standard Malaysia Rubber Grade 20(SMR20). Both single and simultaneous equations were utilized using monthly data from January 1990-December 2008 as estimation period and data from January 2009-June 2009 was used as an ex-ante forecast. The data were tested for unit root and Vector Error Correction and co-integration method was used to estimate the parameters of the model. The models specifications were developed in order to discover the inter-relationships between NR production, consumption and prices of SMR20 and to determine forecast price of SMR20. Comparative analysis between the single-equation specification and simultaneous supply-demand and price equation were made in terms of their estimation accuracy based on RMSE, MAE and (U-Thile) criteria. Ex-ante forecasts was carried out for the period of January 2009-June 2009. The results revealed that the values of the RMSE, MAE and U of simultaneous supply-demand and price equations model were comparatively smaller than the values generated by the single-equation model. These statistics suggest that the simultaneous equation of supply-demand and price model is more accurate and efficient measure in terms of its statistical criteria than the single-equation model in predicting the price of SMR20 in the next 6 months.

Item Type: Article
Subjects: H Social Sciences > HB Economic theory. Demography
Divisions: Faculty of Management (FOM)
Depositing User: Ms Rosnani Abd Wahab
Date Deposited: 13 Feb 2014 03:55
Last Modified: 13 Feb 2014 03:55


Downloads per month over past year

View ItemEdit (login required)