本文使用月資料。樣本期間自2000年01月至2011年08月,應非參數的Triples檢定探討台灣18個分類指數與匯率相關性是否存在不對稱效果,並進一步檢視股票報酬與匯率貶值相關性的不對稱形式。實證結果發現,以30筆移動視窗建立的相關係數皆為不顯著,沒有證據支持各分類股價指數與匯率的相關性存在不對稱效果;以60筆及90筆移動視窗建立的相關係數檢定發現紡織纖維和鋼鐵工業都存在顯著負的陡度不對稱效果,說明兩者的相關係數在不好的消息時,下降速度相對好消息時上升速度快。
This study uses monthly data spanning the period 2000-2011, applies Triples test of Randles et al. (1980) to investigate asymmetric behavior in correlation between eighteen industrial stock index and exchange rate and its form of asymmetry. The empirical result shows no significant evidence under the 30 moving window. Instead, under 60 and 90 moving windows, the result shows significant negative steepness effect in the industry of Textiles and Steel & Iron. It implies that correlation coefficient fall faster in bad news than rise in good news for Textiles and Steel & Iron industries. The results are relevant for investor, modeler and policymaker.