Abstract:
In this paper, we propose a discrete version of the short-term and long-term component model of the agricultural futures prices. The maximum likelihood estimate of each parameter is obtained using an adaptive filtering algorithm. The diagnostics statistically support the specification of the model. The short-term components exhibit no causal relationship with economic fundamentals such as inflation rate and economic growth rate. These components, therefore, seem to be driven mainly by fads rather than market fundamentals. On the other hand, the long-term components show conitegrating relationship with only one cointegrating vector among the three futures contracts examined. Copyright Springer Science+Business Media, LLC 2006