Ebook Testing Predicitive Ability of Business Cycle Indicators for the Euro Area
The analysis of business cycles and its characteristics has a long tradition in economic research. In addition, economic growth is beside unemployment and inflation one of the most important target variables in the decision process of policy makers. Due to increasing political and economical uncertainty about future developments, reliable forecasts become more and more important.
In our work, we concentrate on business cycle indicators for the Euro Area that have been developed by research institutes, banks and the European Commission in order to improve forecasts and reduce uncertainty. These indicators are important tools for enterprizes, central bankers and politicians to predict the future development of the economy. We analyze the predictive ability of seven indicators which are quite different regarding their conception. The empirical analysis shows that they also have significantly different forecasting performances. In contrast to related articles in this area we consider seven special economic indicators that are used in practice to conduct economic forecasting. These indicators are constructed with a specific focus on the European business cycle movements. However, the forecasting abilities of these indicators have not been analyzed so far from a comparative perspective to the best of our knowledge. Our comparison is conducted in two ways: the in?sample and the out?of?sample analysis. The former uses all available information to estimate cross correlations and to test against Granger causality. The latter tries to mimic a realistic situation where the future is unknown. We make use of the bivariate vector autoregressive framework to generate one?step ahead forecasts of year?over?year growth of industrial production which serves as the reference series. Our choice is motivated by the fact that this series is available at a monthly frequency implying a larger number of observations compared to the quarterly GDP series.
A great deal of literature concentrates on the forecasting abilities of other economic indicators such as the interest rate spread or indicators such as M2 growth (see Dovern and Ziegler (2008)). At forecasting real GDP our paper is related to rather extensive literature that assesses the forecasting properties of various leading indicators for Germany such as by Breitung and Jagodzinski (2001), Schumacher and Dreger (2004), Fritsche and Stephan (2002), Hüfner and Schröder (2002), Schumacher (2007) or Kholodilin and Siliverstovs (2006) and the Euro Area such as Forni et al. (2003).
This paper is organized as follows. Section 2 presents the data base and especially the components of different indicators. Section 3 covers the methods and results of the in sample analysis, while section 4 is dedicated to an out?of?sample experiment and recent tests for superior predictive ability. Main conclusions are drawn in section 5.
Download
PDF Ebook Testing Predicitive Ability of Business Cycle Indicators for the Euro Area
Posted in :
