Anmerkungen:
Nach Informationen von SSRN wurde die ursprüngliche Fassung des Dokuments February 24, 2021 erstellt
Beschreibung:
The COVID-19 pandemic triggered an extreme variation in many key macroeconomic indicators. This paper documents that multivariate t-distributed errors are better equipped to capture this variation than common stochastic volatility in a Bayesian VAR. Diagnostics indicate that the data prefers to interpret the COVID-19 shock as a rare event. The paper shows that not accounting for heavy-tailed errors leads to imprecise density forecasts during the pandemic. Besides, this paper documents another source of parameter instability stemming from a mechanical update of the prior distribution. Particularly, the revised Minnesota prior leads to an extremely tight prior distribution for real variables and triggers an extreme loss of the in-sample fit. To mitigate this sensitivity, a COVID-19 robust prior calibration strategy is put forward