Detecting turning points in the Danish economy in real time

This working paper documents an econometric model for detecting turning points in the Danish economy in real time. The model is a mixedfrequency model using both monthly and quarterly data, which can be estimated on an unbalanced panel of data and be updated immediately as data comes through. The model allows the user to provide both a statement about the probability of the Danish economy being in a recession or an expansion as well as in terms of a quantitative estimate of the growth rate on a monthly basis. Further, the model provides a measure of the uncertainty surrounding such a nowcast. The properties of the model are shown and the realtime performance of the model is analysed. It is shown that the model gives correct signals about the activity in the Danish economy, also in real time. The model has been in use in Danmarks Nationalbank since the beginning of 2020, and nowcasts and the uncertainty surrounding this nowcast produced by the model have been published since summer 2020.