A complex network perspective of world stock markets: Synchronization and volatility

This paper studies the cross-correlations of 67 stock market indices in the past 5 years. In order to capture the interaction of the stock markets, we propose to take a complex network approach to analyzing the interdependence of the individual stock markets. Specifically, stock markets are considered as network nodes, and the network links (weights of links) are defined by the cross-correlations between market indices over a period of time (time window). Thus, the resulting network provides information about the interdependence of individual markets, with the network links representing the extents to which the markets are correlated. If we allow the time window to move in forward time and construct a network for each time window over a long period of time, we are able to capture the dynamics of the network. In our study, all networks are constructed from raw data of market indices, and our aim is to investigate how network properties can be used to infer market behavior. By examining the variation of the network parameters as time elapses, we show that stock markets of different countries have time-varying interaction, and that developed markets tend to demonstrate similar behavior while emerging markets are statistically independent of each other. Furthermore, we observe synchronization in the network of stock markets, which is an important universal property observed in many physical and man-made networks. Specifically, we show that stock markets of different countries generally behave in a synchronous manner when they experience fluctuation, which is especially notable in the developed markets. This work exposes the interdependence of stock markets in the world and proposes a complex network approach to identifying some salient global behavior of the interconnecting markets.