| Summary: | 碩士 === 淡江大學 === 全球華商經營管理數位學習碩士在職專班 === 97 === This research applied various time series methodologies to fully investigate the long run and short run dynamic relationships among daily stock indices of the US NASDAQ, Taiwan stock, Japan Nikke 225 and Hong-Kong Hensen for the period of 2005/03/01 through 2009/03/31. Those time series methodologies applied include unit-root test, cointegration test, Granger causality, impulse response function and variance decomposition. The findings of this research are as follows. Unit-root test finds that all the four indices considered are I(1) series. Further test of cointegration shows that the four indices do not cointegrated together, which also implies that no long run equilibrium relationship exists among the four indices consided. The finding of on cointegration during the period of sub-prime financial crisis tells us that countries concentrate more about their own policies to defeat the downtrend of the bearish condition. Therefore the stock markets move apart among each other. This situation also allows diversification strategy when implementing the international portfolio. Results from Granger causality illustrate that the US sock index shows a significant leading position among all the indices considered. This also can say that the US financial market is among the leading role all over the world. The Granger causality also finds that Taiwan stock index leads Hong-Kong Hensen index. This might be reasoned by that both Taiwan and Hong-Kong are strongly related to China fast growing economic, however, Taiwan is more affected by China’s economic growth then Hong-Kong does. The overall ordering of the four indices form Granger causality is the US NASDAQ, Taiwan stock, Hong-Kong Hensen and Japan Nikke 225.
The analysis of impulse response function shows that all the indices are responded to their own shock more significant. However, the shock of the US index shows strong influences to all other three, which again supports the leading role of the US financial market. Finally, variance decomposition shows that all the indices are more likely self-explained, especially for the US index which explain its own variation higher then others do.
The purpose of this research is to lump sum the information about the dynamic relationships among international stock indices. The findings of this research offer a valuable information to the investors who’s investment strategies is not just for the interest bearing deposit, but collecting all the useful informations.
|
|---|
