美、日股市巨幅波動下的股市連動效果－美國、日本與亞洲四小龍股市實證結果

This study theoretically adopts Markov-switching models (hereafter SWARCH models) by Hamilton and Susmel (1994). Examine the weekly return variability of stock market indices including Dow Jones (price-weighted), Nikkei (price-weighted), and measures for the four Asian tigers. Via a SWARCH setting, model stock return variances as being state varying, we document greater correlations when index returns are more volatile. Specifically, we specify American and Japanese stock markets as the home markets and include observations for Singapore, Hong Kong, South Korea and Taiwan in the set of foreign markets, exploring the correlations between the home market group and the foreign market group. Our empirical findings support the following notions. First, the correlation appears to be the greatest when both US and Japanese markets are in a high variance state, whereas the first runners up is the correlation for settings one and only home market being based on a high variance state. And the correlation is the lowest when both US and Japan markets are in a low variance state. Our results show that the global financial epidemic is most significant when the major developed markets are volatile. Moreover, the findings suggest that accompanying the rapid growth in global fund transfers, deregulations, and international divisions of labor in the electronic industry, the global market epidemic effect has been significantly increasing since 1990. The research design and test results of this study may add to the literature and practices of measuring cross-country market correlation patterns, implementing risk management systems, and establishing diversification strategies.