在過去的股市相關性的探討,大多數的論文使用VAR(vector auto-regression model, 向量自我迴歸模型),VECM(vector error correction model, 誤差修正模型),以及Granger因果關係檢定來探討美國與台灣股市間的相關性分析。此篇論文使用分量迴歸法來建構模型,此外,我們使用美股日報酬為自變數,以台股的日報酬與隔夜報酬當因變數,檢驗美股對台股的影響。由分量迴歸的實證結果發現,台股受美股的影響顯著,特別是對台股大漲大跌時,有較大的影響。 Most researches use VAR, VECM, and Granger causality to estimate the links between US and Taiwan stock markets. This study use quantile regression approach to build model. Using this approach could investigate marginal effects of dependent variable at various quantiles. In addition, we also examine US daily return in previous days on Taiwan daily return and overnight return to explore the direct impact from US market. The empirical results indicate that Taiwan's stock market is significantly influenced by US stock market, particularly when the stock market rise and go down dramatically.
