The main objective of this study is to estimate petroleum products demand using a random trend approach with aim of deriving improved and more robust estimates of price and income elasticities. The study specifies the random trend model of petroleum products demand as a two-step stochastic process. The estimates of model parameters for each petroleum products, in Nigeria are obtained by applying maximum likelihood in conjunction with Kalman filter. The study revealed that the introduction of random trend reduces the estimate of the coefficient of the lagged dependent variable in the three petroleum products relative to no trend model. As a result, price and income elasticities of petroleum product demands are higher in the short run and long run relative to constant intercept model. The introduction of random trend leads to improvement in the mean square errors of within sample forecasts
Keywords: Elasticity, Kalman Filter, Maximum Likelihood And Energy Demand
