Multifractal Cross-correlation Analysis Between Domestic And Foreign Stock Markets

With the acceleration of economic globalization, it gradually shows the linkage of the stock markets in various counties, also produces a risk conduction effect. So this paper mainly studies multifractal cross-correlation between domestic and foreign stock markets. We select stock index data of China, the USA, Germany, India and Brazil from January 2011 to September 2014, using the multifractal detrended cross-correlation method (MF-DCCA) to analyze the cross-correlations, multifractal characteristics and time scale consistency between Chinese stock market and developed/developing counties, and the time varying characteristics of the cross-correlations. Furthermore, we use the asymmetric MF-DCCA method based on uptrend/downtrend, and the asymmetric MF-DCCA method based on different risk conduction effect, to study the asymmetric cross-correlation between domestic and foreign stock markets.The empirical results indicate the presence of cross-correlation and multifractal characteristic between the domestic and foreign markets. And the persistent cross-correlation between China and developing countries is stronger than the persistent cross-correlation with the developed countries. Also, the cross-correlation significantly enhanced during the financial crisis. Furthermore, when a market is in different trends, the cross-correlations between domestic and foreign stock markets are asymmetric and having multifractal characteristics, and the degree of persistence of cross-correlation is stronger when a market in a downtrend than in a uptrend. And it exists bidirectional conduction effect between domestic and foreign stock markets, and the influence degree from foreign countries to domestic market is greater than the influence degree from domestic market to foreign countries.