The role of early indicators in export forecasting

Macroeconomic variables can be forecasted by a large number of econometric methods, such as structural time series models, vector-autoregressive models, estimates of single equations and simultaneous multi-equation models. Experience in export forecasting has shown that the inclusion of additional, current information can improve forecasting accuracy. In this article, data series are discussed that can be used for short-term appraisals of German exports as well as econometric estimation methods that allow for a quantification of the current growth of exported German goods using these early indicators. The current analysis shows that these leading indicators point to a weakening in export activity for the first quarters of 2005. The forecast for real goods exports for the second quarter of 2005 stands at €198-206.7 billion and at €188.5-195.3 billion for the third quarter of 2005.