| Summary: | 博士 === 國立臺灣科技大學 === 管理研究所 === 95 === This paper adopts GM(1,1) and GARCH/TGARCH time series model to predict the return and volatility of nine major index futures among American and Eurasian markets. In a further step, by means of grey relational theory, GARCH and GM(1,N) model, it can be found the volatility relationship and the main factor among the targeted markets mentioned above. Finally, the comparison about the prediction performance between GARCH/TGARCH and grey theory will be discussed. The findings reveal that the GARCH/TGARCH outperformed the GM(1,1), consisting of the optimal GM(1,1｜α) model. In addition, according to the simulated GARCH/TGARCH model aimed at the prediction concerning about the volatility of daily return of American and Eurasian stock price index futures, the DJ, NASDAQ and S&P in America, CAC and FTSE in Europe, TX, TE and TF in Asian market all have two-way daily return spillover effect. Besides, CAC and Nikkei, CAC and S&P, S&P and TF, NASDAQ and TX also have significant mutual spillover effect of daily return. Meanwhile, it is found the GARCH effect indeed exists, that is to say, there is some volatility clustering situation in the aforementioned markets. It also reveals by GM (1,N) that the daily return of Dow Jones index future is the main influence factor on other index futures rate of returns, which is similar to the Gannon’s (2005) results. That is, the GM(1,N) model can work in financial spillover analysis. In conclusion, the American market firmly dominates global stock markets and forward markets.
|
|---|
