Performance of periodic error correction models in forecasting consumption data

Periodic time series models have become an appealing tool for the analysis of seasonal time series. Since these models condition the data-generating process on the season they are able to cope with periodic generalisations of economic models with seasonal preferences and seasonal technologies. This paper examines the forecasting performance of alternative specifications of a periodic error-correction model for the relation between consumption and income. Estimation and forecasting exercises are performed for data from the United Kingdom, Sweden, Germany and Japan.