روند احتمالی مشترک و نوسانات در بازارهای سهام در آسیا و اقیانوس آرام

This paper uses Johansen's cointegration test and a modified cointegration test with generalized autoregressive conditional heteroskedasticity (GARCH) effects to examine linkages between the U.S. and five Asian-Pacific stock markets (Australia, Hong Kong, Japan, Malaysia, and Singapore) during the period from 1988 to 1994. The modified cointegration test with GARCH effects is used to assess whether these stock price series share common time-varying volatility. The results indicate that the six stock markets are highly integrated through the second moments of stock returns but not the first moments.

This paper uses a modified cointegration test with GARCH effects to test for common volatility in six stock markets, including Australia, Hong Kong, Japan, Malaysia, Singapore, and the United States. Unlike that of previous studies that examine short-term mean return and volatility spillovers, the focus of the current study is on whether international stock markets have long-term, common time-varying volatility. The empirical results from examining the data for the period from 1988 to 1994 indicate that the six stock market indexes are not cointegrated in the first moments. The findings of this study also indicate the presence of ARCH effects in most stock price series, suggesting that testing for cointegration needs to account for time-varying volatility. The modified cointegration test with GARCH effects results reveal that the six daily stock market index series do have the same volatility process. However, for weekly data, statistically significant common time-varying volatility exists only in the data system that exhibits ARCH effects. In short, the modified cointegration analyses generally find the six stock markets to be integrated through the second moments of stock returns but not the first moments. The empirical results of this paper therefore suggest that volatility transmissions among international stock markets exist not only in the short term (volatility spillover), as documented in prior studies, but also in the long run (common time-varying volatility).