Dissecting Financial Connectedness in Spillovers and Correlations

This paper proposes a novel econometric framework to estimate connectedness, its components spillovers and correlations and their power to explain the variation of returns and volatility. The results based on a sample of large stocks and global stock market indices show that spillovers have comparatively low explanatory power in normal periods but high explanatory power in crisis periods. The estimates indicate that past information is more quickly incorporated in stock prices in normal times than in times of extreme price movements. A comparison of the COVID-19 outbreak with the 2008 financial crisis also suggests the COVID-19 shock led to more spillovers and thus was more disruptive.