| Summary: | 碩士 === 國立高雄應用科技大學 === 金融系金融資訊碩士在職專班 === 106 === This research explores the important variables of TSEC, T50, T50ETF and TAIEX, analyzing the relevance of each. The information being researched was spanning from June 30, 2003 to July 30, 2017 on the basis of unit root test, co-integration, vector error correction model, causality test and autoregressive model without stationary time-series. Hereby summarizing the empirical research conclusion of this paper as followings:
1. The co-integrated analysis of variables between TSEC, T50ETF, T50 index, TAIEX shows a long-term stable equilibrium equation relationship exists.
2. The variable of TSMC is leading T50 ETF reaction in one way, but shows two way leading reaction with T50 and TAIEX mutually. As for T50 index, which is leading T50 ETF reaction in one way and T50 ETF is leading TAIEX in one way as well. Yet T5o index and TAIEX are in relationship of Granger causality.
3. The empirical of impulse response validates : When the impulse occurs, the response of TSEC return rate was not only be affected the lagging period itself, but also impulse the response of other index.
4. Forecasting the error variables to analyze the share price of TSMC can make self-interpretation to the highest extent: The result shows the share price of TSMC is most likely to affect other index but however not easily affected by others.
The lowest extent to make self-interpretation is the variables of TAIEX: The results shows TAIEX can most easily be affected by other variables, yet not easily affect others.
|
|---|
