COVID-19, Macroeconomic Dynamics and Fear in Europe: A Network Global VAR Approach
Alessandro Celani,Paola Cerchiello,Paolo Pagnottoni
DOI: https://doi.org/10.2139/ssrn.3933917
2021-01-01
SSRN Electronic Journal
Abstract:The COVID-19 pandemic has exerted severe impacts on the socioeconomic fabric of worldwide countries, spreading financial and economic instability, affecting sentiment of market participants and institutions, and causing an unprecedented epidemic-related economic crisis. Interconnectedness is a key determinant for the spread of a crisis in the network structure of entities, as large shocks tend to propagate to the economy in its complexity. In this paper we combine the statistical and econometric literature concerning generalized forecast error variance decomposition (GFEVD) with network science to study the topology of spillover indices across countries, macroeconomic variables and economic sentiment data. To this aim, we propose a dynamic network framework derived from a Global Vector Autoregressive model (GVAR), a suitable macroeconometric model to consider simultaneous multi-level interdependencies across variables, which we analyse through the lens of network theory. We apply our methodology to study the interrelated dynamics of a set of European countries' by looking at the Industrial Production, the Retail Trade and the Economic Sentiment indexes over the period 01/2000-03/2021. Our main results show how different crises exert diverse impacts on real economy, economic sentiment and on the entire countries' macroeconomic network structure. While the aggregate impact of the Global Financial Crisis on the network interrelationships in Europe is approximately the same in terms of global spillover magnitude as that induced by the COVID-19 outbreak, the latter seems to be more persistent over time, and with starkly diverse implications on the temporal network structure. We also demonstrate that, overall, economic sentiment tends to be an informationally dominant node in the network over real economic variables, even during the Global Financial Crisis period. However, the two components have recently swapped their roles due the COVID-19 outbreak, meaning that such a large, unexpected, and non-financial shock has allegedly caused the real economy to dominate macroeconomic expectations.