Forecasting Availability and Consumption of Rubberwood in Peninsular Malaysia

Rubberwood sawntimber has become one of the major contributors to the Malaysian furniture export and economic growth in the past two decades as its demand from wooden furniture manufacturing has increased tremendously. However, the availability of rubberwood sawntimber would not meet industry’s demand. Therefore there is a need to examine and to identify the availability of rubberwood sawntimber in fulfilling the demand. For this purpose, up-to-date information on availability and consumption of rubberwood sawntimber in Malaysia is important. Factors affecting availability and consumption of rubberwood sawntimber are identified through the multiple regression model. The availability and consumption of rubberwood sawntimber are forecasted for five years ahead from 2008 to 2012 through three models that have been identified among the best forecasting technique which is multiple regression, Box-Jenkins and composite model that combine both models. The composite model is applied to balance the limitation of each model in order to get the best estimate forecasts. With each of the strengths of these two methods, using them together as one method will generate the most reliable forecast value. The results of the regression analyses have been statistically diagnosed and modified to generate the best models. Availability of rubberwood sawntimber is influenced by replanted area, price of rubberwood log, latex price and previous production of rubberwood sawntimber. Consumption of rubberwood sawntimber is influenced by national income and prices of other light hardwood species that are mainly consumed by sawmills, i.e. Dark Red Meranti and Red Meranti In forecasting availability and consumption of rubberwood sawntimber statistical evaluation of all of the three models shows an acceptable result which means these models could give reliable forecasts. Both multiple regression and Box-Jenkins model in forecasting availability and consumption of rubberwood sawntimber in Malaysia are reliable due to the smaller error and a closer value between forecasts and actual values compared to previous forecasts. The composite model makes the forecast values more reliable and stable and this proves that this model balance the strength and weaknesses of both forecasting techniques and is able to generate the best estimated forecasts