MULTIPERIOD FORECASTING ANALYSIS OF A DYNAMIC MODEL FOR A SMALL SAMPLE

In this paper we derive the formulae for the bias and mean squared forecast error (MSFE) of the least squares forecast several periods ahead in the context of a dynamic model. Since the expressions are in terms of integrals, we have also obtained the numerical value of the bias and MSFE for different values of parameters and different disturbance structures. The results confirm some earlier studies (based on the AR(1) model), for example Lahiri (1975) and Hoque et al. (1988).     

THE EFFICIENCY OF THE COCHRANE-ORCUTT PROCEDURE

It is well known that the ordinary least squares (OLS) estimator, though unbiased, is inefficient in the presence of autocorrelated disturbances. Further, it is also widely accepted that the Cochrane-Orcutt (C-O) estimator is more efficient than the OLS estimator. However, Kadiyala (1968) and Maeshiro (1976, 1978) have argued that OLS is more efficient than C-O when the independent variable is trended and the autocorrelation coefficient is positive. We re-examine this issue and show that C-O is more efficient than OLS for the model without an intercept term.