Uncertainty and global common volatility: an explainable artificial intelligence approach

Global financial markets are increasingly exposed to shocks that propagate rapidly across countries, sectors, and asset classes, generating synchronized volatility that reflects the systemic dimension of financial risk. While prior research documents the emergence of global volatility shocks, less attention has been paid to the factors associated with volatility synchronization. This paper investigates the uncertainty proxies associated with global common volatility (COVOL), combining SHAP values, Accumulated Local Effects (ALE), and Local Interpretable Model-agnostic Explanations (LIME) within a machine learning framework to capture potentially nonlinear relationships. The results show that COVOL is most strongly associated with uncertainty proxies linked to oil market volatility, broad equity market volatility, technology-sector volatility, and global investor risk aversion, with nonlinear and time-varying contributions across these predictors. These findings carry direct practical implications. Because oil market uncertainty (OVX) and equity market uncertainty (VIX and VXN) emerge as the proxies most strongly associated with global common volatility, monitoring these specific indicators can serve as an early-warning signal for the build-up of synchronized volatility across markets. This is particularly relevant for risk managers calibrating hedging strategies against systemic volatility, for investors managing global portfolio exposure, and for policymakers tracking financial-stability risks.