Past

Risk Contagion Under Extremes: interplay of Heavy-tailedness and Tail dependence

Abstract

Risk contagion (RC) has been shown to play an important role in explaining financial crises in recent decades. This paper contributes to quantitative risk management by modelling extreme dependence and extreme risks in an integrated manner via Extreme Value Theory (EVT). A flexible RC measure is proposed to tailor a pair of risks with extensive dependence structures. Its asymptotic behaviour is well established under a mild dependence condition. Besides the consistency of the empirical estimators, we introduce an extrapolation method based on EVT to estimate the proposed RC for high quantile levels where little data is available. The findings of the interplay between extremes and tail dependence structures are well illustrated by intensive numerical studies and an empirical study on the RC between financial markets. Moreover, we apply the RC to 30 global systemically important financial institutions (G_x0002_SIFIs), which perform well in ranking systemic risks during systemic events.