美元指數可以預測美國主要股票市場嗎?

本研究主要以台灣經濟新報資料庫，2018年 1月 2日到 2020 年10 月 23日間之美元指數、道瓊工業股價指數、標準普爾500股價指數、那斯達克股價指數之日資料計660筆，並運用多種時間序列計量方法，驗證美元指數與美國主要股票市場之間的因果關係，探討美元指數是否可以預測主要的美國股票市場？實證結果發現四個指數之間不存在共整合關係，表示變數間不存在長期穩定關係，因此迴歸模型採用向量自我迴歸模型(VAR)來做分析。在Granger因果檢定中，整體結果顯示美元指數報酬預測能力最強，可以影響道瓊工業股價指數報酬、標準普爾500股價指數報酬與那斯達克股價指數報酬；反觀美國主要股價指數報酬，只有道瓊工業股價指數報酬可以影響美元指數報酬，而標準普爾500股價指數報酬與那斯達克股價指數報酬，完全無法影響美元指數報酬。

關鍵字

美元指數；股價；向量自我迴歸模型；因果關係

並列摘要

This article is based on the Taiwan Economic News Database (TEJ), the date of the US dollar index, Dow Jones industrial stock index, S&P 500 stock index, and Nasdaq stock index from January 2, 2018 to October 23, 2020 A total of 660 transactions, and using a variety of time series measurement methods to verify the causal relationship between the US dollar index and major US stock markets, explore whether the US dollar index can predict major US stock markets? The empirical results found that there is no co-integration relationship between the four indexes, which means that there is no long-term stable relationship between the variables. Therefore, the regression model uses the vector autoregressive model (VAR) for analysis. In the Granger causality test, the overall result shows that the dollar index has the strongest ability to predict returns, which can affect the returns of the Dow Jones Industrial Stock Index, the S&P 500 Stock Index and the Nasdaq Stock Index. In contrast, the major US stock index returns are only Dow Jones Industrial stock index returns can affect dollar index returns, while S&P 500 stock index returns and Nasdaq stock index returns cannot affect dollar index returns at all.