Translating financial integration into correlation risk: A weekly reporting's viewpoint for the volatility behavior of stock markets

We apply the multivariate extension of GARCH-type models in order to assess the systematic and systemic risks as well as the joint volatility behaviors of the U.S. and three European financial markets (Andersen et al., 2010). Therefore, we can appraise the co-movements of the four previous financial markets as well as the joint behavior of their respective volatilities (i.e. systemic risk). Moreover, the resulting conditional variance and covariance metrics allow for handling volatility spillovers (i.e. contagion risk in terms of transmitting volatility shocks from one market place to another market place). Indeed, results highlight the unprecedented high systemic risk levels (i.e. joint increased volatility levels) as well as a high contagion risk (i.e. volatility spillover) during the subprime mortgage market crisis. The transmission process of volatility shocks reveals to be simultaneous across financial markets due to a strong arbitrage activity and electronic trading practices among others. Most importantly, the estimated conditional correlations exhibit an upward sloping trend, which underlines an increase in the correlation risk between financial markets in the late nineties or early 2000. Thus, our major findings are twofold. First, we characterize the dynamic correlation risk across financial markets. Second, we also confirm the increasing and nonlinear trend in the correlation risk, which we are able to quantify.

► We study correlation risk across U.S. and European financial markets with GARCH model. ► The trend of financial markets is representative of the level of the systematic risk. ► Time-varying conditional covariances and correlations illustrate the systemic risk. ► We account for the asymmetric response of the variance-covariance matrix to bad news. ► We characterize the nonlinear and upward sloping trend in the conditional correlations.