Multiscale network for 20 stock markets using DCCA

Abstract

The aim of this paper is to analyze the stock exchanges for a large set of countries (20 in total) before and after the subprime crisis, identifying which markets are the most central and if the linkage pattern changed after the crisis. We started by calculating the correlations between stock markets’ returns, using the ρDCCA, in order to identify if there is some variation in the scale between the links in the different stock markets of the network, in both periods. Additionally, a cross-correlation filtering process will be performed with the intention of identifying which countries have stronger relationships according to the used time scales. The results show the central role of European markets among the world’s main financial markets, mainly France, Germany and the United Kingdom. Moreover, after the subprime crisis we find the formation of two large communities, one of European and American countries and the other formed by Asian countries plus Australia, while in the pre-crisis period three communities could be identified. It is possible to conclude that after the 2008 crisis the connectivity and integration of the network for the whole set of analyzed timescales increased.