Co-movements of international equity markets: a large-scale factor model approach

We analyze the comovements of a set of country-sector indexes from 45 different countries studying their factor

decomposition based on a PCA analysis for a large cross section framework. We derive a measure to analyze the

comovements over time based on the part of variance explained by the main extracted factors and we apply the

method from Bai and Ng to study the relevant number of factors. We conduct rolling estimations for the period 1994-

2006 focusing on the set of emerging markets. We show that both, emerging and developed equity markets

experienced increasing comovements over the period of study, reflecting the integration of those markets. We have

estimated that the main factor accounts for 30\% and 20\% of the whole variation of each data set. We use the

comovements to gauge integration in two different ways, both indicating higher integration for developed markets.

Finally, we relate the comovements to a measure of diversification and we conclude that it is only possible to reduce

85\% of the average risk of an equity index by diversification at the end of the period compared to 95\% at the

beginning for the set of emerging markets.