Coarse Bayesian VARs - Correcting BVARs for Incorrect Specification: Florian Huber, joint with Massimiliano Marcellino - Economics Seminar

Event Date: 15 March 2023

Speaker: Florian Huber, University of Salzburg, joint with Massimiliano Marcellino, Bocconi University

Time: 4-5:15pm

Location: Please contact Rachel Hill (r.hill@strath.ac.uk) for Zoom details. On campus: CW506b

Abstract: "In this paper, we propose a simple method for addressing issues related to model miss-specification in multivariate time series models. Our method, called coarse Bayesian VARs (cBVARs), replaces the exact likelihood with a coarsened likelihood that takes into account that our model might be miss-specified along important dimensions. This results in a computationally simple model. As opposed to more flexible models, our model avoids overfitting, is simple to implement and estimation is fast. The resulting cBVAR  performs well in simulations using highly non-Gaussian data generating processes. Applied to US data, cBVARs improve point and density forecasts compared to standard BVARs. Additionally, we demonstrate the use of cBVARs for analyzing the effects of uncertainty shocks."

Published: 9 February 2023