This paper focuses on the forecasting process at the Czech National Bank with an emphasis on incorporating expert judgments into forecasts and addressing data uncertainty. At the beginning, the core model and the forecasting process are described and it is presented how data and the underlying uncertainty are handled. The core of the paper contains five case studies, which reflect policy issues addressed during forecasting rounds since 2008. Each case study first describes a particular forecasting problem, then the way how the issue was addressed, and finally the effect of incorporating off-model information into the forecast is briefly summarized. The case studies demonstrate that a careful incorporation of expert information into a structural framework may be useful for generating economically intuitive forecasts even during very turbulent times, and we show that such judgements may have important monetary policy implications.
JEL codes: C53, C54, E17
Keywords: DSGE models, forecasting, Kalman filter, monetary policy
Issued: October 2013
Download: RPN No. 2/2013 (pdf, 1 MB)