Introducing stepwise causality in Global VAR (GVAR): Evidence from the European financial system
摘要
We develop a multistep GVAR causality framework to identify sequential causal linkages in highly interconnected financial systems. Building on the concept of stepwise / multistep causality, and the Global Vector Autoregressive (GVAR) framework, the proposed approach establishes the testing strategy and assess its finite-sample performance through Monte Carlo (MC) simulations. We apply the methodology to a panel of European Union (EU) economies in order to trace shock transmission between traditional banking and shadow banking sectors. The results reveal strong propagation. Shadow banking is tightly connected to the official banking system, while shocks do not remain confined within individual national financial systems but spread through the European financial network. The Italian shadow banking sector emerges as a particularly important transmission node, exerting statistically significant influence on the Eurozone and EU economies. These findings suggest that shadow banking constitutes a relevant channel of systemic risk and financial contagion in Europe. From a policy perspective, the results support the need for stronger macroprudential monitoring of non-bank financial intermediation and closer regulatory attention.