Nonlinear relationship between physical environment risks, investor attentions, and financial systemic risks: Evidence from mLSTM-CoVaR networks.

Abstract

How do physical environment risks affect financial market systemic risk? We use remoting data to measure physical environment risks and select 26 banks across 12 EU countries. We extend the CoVaR framework with the quantile-mLSTM algorithm, obtaining time-varying CoVaRs. We then use time-varying partial derivatives to calculate the banks' tail risk spillover effects. Next, we construct a panel Quantile-on-Quantile model to explore the nonlinear relationship between physical environmental risks and the systemic risk. We show that CoVaR significantly increases during COVID-19 and the Russia-Ukraine conflict. However, CoVaR does not rise significantly during the Israel-Hamas conflict. Besides, our results display a nonlinear relationship between physical environmental risks and systemic risk in banks. Higher levels of physical environmental risks generally increase systemic risk in most conditions. Under normal market conditions, rather than extreme market situations, increased investor attention amplifies the marginal impact of physical environmental risks on systemic risk. Our research provides deeper insights into the interplay between environmental factors, investor attention, and financial stability, highlighting the critical role of investor sentiment in amplifying the impact of physical environmental risks on systemic risk.