Change-Point Detection in the Covariance Structure of Multivariate Volatility Models

Abstract

We propose semi-parametric CUSUM tests to detect a change point in the covariance structure of non-linear multivariate models with dynamically evolving volatilities and correlations. The asymptotic distributions of the proposed statistics are derived under mild conditions. We discuss the applicability of our method to the most often used models, including constant conditional correlation (CCC), dynamic conditional correlation (DCC), factor, asymmetric DCC and BEKK. Our simulations show that, even though the near-unit root property distorts the size and power of tests, de-volatizing the data by means of appropriate multivariate GARCH models can correct such distortions. We apply the semi-parametric CUSUM tests in the attempt to date the occurrence of financial contagion during the great recession.