On univariate forecasting comparisons: the case of the spanish automobile industry

This paper investigates the forecasting ability of a new univariate models family of unobservable components, when compared with other more standard univariate methodologies. A forecasting exercice is carried out with each method, in monthly time series of automobile sales. The accuracy of the differents methods is assesed by comparing several measures of forecasting performance on the out of sample predictions for various horizons as well as differents assumptions on the models parameters