Timely available indicators that allow an assessment of the current economic situation are indispensable for a fast and evidence-based response by economic policymakers. Especially when the situation might change abruptly as in the beginning of the 2020s: Covid-19, Energy-Crisis, War in Ukraine.
This project makes use of novel high-frequency data sets from the field of transportation that enables us to assess very recent changes in economic activity. Other high-frequency data are also included. We consider four different kinds of reference series for forecasting: 1) aggregate monthly industrial production, 2) industry-specific industrial production, 3) monthly total goods exports and imports of Austria, and 4) monthly bilateral trade flows with Austria’s most important trading partners. After checking the lead/lag structure of the novel indicator set, we will consider a wide variety of standard univariate and multivariate models for the forecast evaluation. To assess the performance of the monthly indicators, we will compare statistics of the pseudo out-of-sample forecast errors between the benchmark and augmented models (which include the novel data). In addition to the models based solely on monthly indicators, we will employ models for mixed frequencies to assess the informational content of weekly updates of the indicators. In the end, this should help us to improve our nowcast for of the Austrian economy.