A revisit to efficient forecasting in linear regression models

This paper deals with the improved forecasts for the values of the study variable in linear regression models utilizing the minimum risk approach. It considers the simultaneous forecasting of actual and average values of the study variable and reports the performance properties of the classical unbiased forecasts and two biased forecasts with respect to the criteria of the bias vector, mean squared error matrix and forecast risk, employing the small disturbance asymptotic theory.