Time-Varying Network Connectedness of G-7 Economic Policy Uncertainties: A Locally Stationary TVP-VAR Approach

Abstract

This work analyzes the frequency-dependent network structure of Economic Policy\rUncertainties (EPU) across G-7 countries between January 1998 and April 2021.\rWe implement an approach that builds dynamic networks relying on a locally stationary\rTime-Varying Parameter-Vector Autoregressive model using Quasi-Bayesian\rLocal Likelihood methods. We compute short-, medium-, and long-term network connectedness\rof G-7 EPUs over a period covering several economic/financial turmoils.\rFurthermore, we structure short-term network topologies for the Global Financial\rCrisis (GFC) and the COVID-19 pandemic periods. Findings of the study indicate\ramplified interdependencies between G-7 EPUs around well-known economic/geopolitical\rincidents, frequency-dependent connectedness networks among them, and stronger\rinterdependencies than the medium-, and long-term linkages. Finally, we find that\rshort-term spillovers are not persistent in the long-term for both turmoil periods.