| 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 === The major purpose of this study is researching the relationships among S&P 500, US dollar index, commodity and high yield bond. We used unit root test, Johansen co-integration test, vector error correction model, causality test, and vector auto-regression model to examine the variables respectively whether they have the stable and balanced relationship over a long period, and then analyzed the adjustment process of short-term dynamic imbalance of all variables, and explored the causal relationship with in every variable. The research data of this thesis are the monthly data from January 1994 to August 2017. Now the main empirical results of this research are summarized and analyzed as follows:
1. From the empirical result of Johansen Co-integration Test, we know that:
According to the Johansen co-integration relationship, the S&P 500 index is positively correlated with the U.S. dollar index and the U.S. high-yield bond index; however, it is inversely related to the CRB commodity index. Further explanation, with other variables unchanged, the increase in the US dollar index and the US high-yield bond index caused the stock market to soar and the S&P 500 index to rise, but the rise of the CRB commodity index will also bring about the short of the stock market, which is, the decline of the S&P 500 index.
2. From the empirical result of the Granger Causality Test, we know that:
From the Matching Causality Check we found that the CRB commodity index and the S&P 500 index are mutual feedback relationships, and the S&P 500 Index leads the U.S. dollar index and the U.S. high-yield bond index. Therefore, it can be used to predict changes in the US dollar index and the US high-yield bond index. In addition, the CRB Commodity Index has the leading relationship with the US High Yield Bond Index, which can also be used to forecast changes in the growth of the US high-yield bond index.
3. From the empirical result of the Impulse Response Function, we know that:
When an impact occurs, the U.S. dollar index, S&P 500 stock index, CRB commodity index, and the U.S. high-yield bond index are affected by their own shocks. It is the largest from the beginning and it converges as the number of period increases.
4. From the empirical result of Forecast Error Variance Decomposition, we know that:
The S&P 500 Index, the U.S dollar Index and the U.S. High Yield Bond Index get the strongest self-explanation. It means that the three indexes are the most exogenous. The CRB Commodity Index gets the weakest self- explanation and is affected other indexes.
|
|---|
