Market Linkages, Variance Spillovers and Correlation Stability: Empirical Evidences of Financial Contagion

Abstract

We propose a simultaneous equation system with GARCH errors to model the contemporaneous relations among Asian and American stock markets. On the estimated residuals, we evaluate the correlation matrix over rolling windows and introduce a correlation matrix distance, which allows both a graphical analysis and the development of a statistical test of correlation movements. Furthermore, we introduce a methodology that can be used for identifying turmoil periods on a data-driven basis. We employ the previous results in the analysis of the contagion issue between Asian and American stock markets. Our results shows some evidence of contagion and the proposed statistics identifies, on a data-driven basis, turmoil periods consistent with the ones currently assumed in the literature.

Keywords: Financial market contagion, Market linkages, Variance spillovers, Dynamic correlations, Rolling correlations, Transformed correlations

JEL Classification: G15, F3, C22, C32