Abstract: The financial globalization facilitates the flow of capital and good across the countries, and thus improves the operational efficiency of the financial markets. At the same time, the financial globalization enhances the risk transmission and thus leads to high volatility of financial markets.Therefore, the co-movement of financial markets is an important issue for risk managements, asset pricing and portfolio management.
A MHAR-DCC model with a Markov regime switching structure is developed to investigate the volatility co-movement of stock markets by employingthe high-frequency data from the six major stock markets in the world from 1 January 2010 to 31 December 2013.
The structural changes in co-volatility are identifed by introducing an exogenous stochastic component that follows a Markov regime switching process. There are more and longer-duration break periods identified by the model with the normal distribution as compared to that with the student-t distribution. In addition, the positive impact that scales up the covariance dominates the breakdown periods.There exhibit significantly positive correlations between the stock market within the same geographyregions as well as significantly positive correlations between the US stock market and the other markets, indicating the leading position of the US stock market in the co-movement of world stock markets.

Key words: volatility co-movement, MHAR-DCC model, high frequency data, multiple stock markets