Natural Rubber (NR) an agricultural product is an important raw material in a broad spectrum of industries such as automobile and pharmaceutical. NR sector has its local and global dimensions. Hence, understanding the dynamic nature of common trends in demand, supply and prices of NR and their interactions with other important factors is vital for the sustainable growth of this sector. This study aims at finding common trends in a multivariate time series and their interaction with explanatory variables using a technique called Dynamic Factor Analysis (DFA) since the conventional time series methods are not capable of handling this kind of analysis. To elaborate the applicability of DFA, four case studies are presented namely, i) world NR production, ii) world NR consumption, iii) NR prices in Sri Lanka and iv) NR prices at international market. Practical issues in selection of models based on AIC are discussed with solutions, which includes a test based on change in log likelihood. This study provides empirical evidences on selecting models based on the proposed method of model selection.Tropical Agricultural Research Vol. 24(1): 54-70(2012)DOI: http://dx.doi.org/10.4038/tar.v24i1.7989
Full Text Finder