Short- and medium-term car registration forecasting based on selected macro and socio-economic indicators in European countries

Abstract

The automotive industry plays a key role in the European economy. In this paper, we determine which macro and socio-economic indicators have significant predictive power on car registrations - a proxy to automotive sector performance - across European countries. Contrary to the current literature which mainly focuses on long-term forecasting, we built our models on the highly seasonal monthly data of a medium-term period to make short-term forecasts. Our approach utilises predictors identified by the literature review. Presented models are built on the Vector Autoregressive models and are accompanied by formal tests, such as the Granger causality test. We have found mixed evidence about the importance of selected predictors as no general patterns were identified. We have found that the most useful predictor is the total number of registrations from the strongest export partner and past registration figures in the analysed country. Car registrations are virtually inelastic to the change of public transportation costs, fuel prices and short-term interest rates offered to households in most of the analysed countries. We have received mixed results about household sentiment across countries. Countries with a higher level of GDP are more sensitive to the changes in unemployment. © 2019 Elsevier Ltd