Nowcasting in a pandemic using non-parametric mixed frequency VARs

This paper develops Bayesian econometric methods for posterior inference in non-parametric mixed frequency VARs using additive regression trees. We argue that tree models are ideally suited macroeconomic nowcasting the face of extreme observations, instance those produced by COVID-19 pandemic 2020. is due to their flexibility and ability model outliers. In an application involving four major euro area countries, we find substantial improvements performance relative a linear VAR.