近年來國際經濟環境迅速變動下,為了避免巨大的匯兌損失,投資人與投資機構因而有必要建構避險組合。本研究以1987年2月2日至2018年3月29日以紐約外匯市場(New York FX Market)的美元兌日圓之現貨與芝加哥商業交易所(Chicago Mercantile Market)的日圓之期貨為研究對象,利用移動視窗的架構,探討不同分配(常態分配與Student t分配)雙變量DCC COPLUA-GARCH(1,1)模型與ADCC COPLUA-GARCH(1,1)模型的避險績效。根據雙變量COPLUA-GARCH-Based模型可發現外匯市場現貨與期貨有高度相依性,而配適不同分配的COPULA GARCH-Based模型最能捕捉兩個市場的相依結構,建構出最小變異數避險組合,創造出最佳避險績效。研究結果發現Student t分配ADCC COPLUA-GARCH(1,1)模型優於常態分配ADCC COPLUA-GARCH(1,1)模型之避險績效。本研究的研究結果可供相關避險者的參考。
In the recent years, the global economic environment has rapidly changed. To avoid the huge exchange loss, it is critical for investors and investment institutions to build a hedging portfolio. We adopts window-rolling framework from February 2, 1987 and March 29, 2018 to examine hedging effectiveness of different distribution with(Normal distribution and Student t distribution) bivariate DCC COPULA-GARCH(1,1) and ADCC COPULA-GARCH(1,1), this study focuses on USD/JPY using the New York FX Market’s spot markets prices and the Chicago Mercantile Exchange futures markets prices. According to the bivariate COPLUA-GARCH-Based model, it can be found that there is a high degree of correlation between spot market and futures in the foreign exchange market. From the results, it is found that COPULA GARCH-Based models with different distribution can capture the dependent structures of the two markets and construct the minimum variance hedging portfolio to create the best hedging performance. The results reveal that hedging effectiveness of the Student t distribution with ADCC COPLUA-GARCH (1,1) model are better than the normal distribution with ADCC COPLUA-GARCH (1,1) model.