New nonparametric measures for instantaneous and granger-causality tail co-dependence

We propose a new methodology to asses risk spillovers in time-series framework. Firstly, we introduce an explicit nonparametric measure of cross-sectional conditional tail co-movement, which is intuitively comparable the Conditional Value-at-Risk (CoVaR). show that nonlinear CoVaR (NCoVaR) able capture even highly dependence structures. Secondly, for purpose potential contagion analysis, adapt be informative about causality direction between variables Granger sense. By showing natural estimators two metrics are U-statistics, construct formal tests independence and non-causality. Numerical simulations confirm common situations have better size power properties than their parametric counterparts. The illustrated empirically by assessing transmissions sovereigns banking sectors euro area, observed irregular co-movements asset prices after global financial crisis. measures seem less susceptible these irregularities analogues, providing clearer overview underlying sovereign-bank feedback loops.