本文以日本股價指數爲主要之研究對象，利用大阪日經225指數期貨與新加坡日經225指數期貨，探討在空頭避險下，應用ARJI模型與GARCH模型進行避險之績效。由於價格在短期間存在跳躍(Jump)的現象，因此本文利用ARJI模型捕捉此不連續的行爲，進而與GARCH比較在不同避險期間下，利用移動視窗(rolling window)的方法探討樣本外(out of sample)避險的避險績效。實證結果爲當避險期間爲5天、10天與20天時，以ARJI模型進行避險的績效較GARCH模型優良；若當避險期間拉長爲40天與60天時，則以GARCH模型的避險績效較佳。另外，無論以ARJI模型或是GARCH模型，採用新加坡日經225指數期貨爲避險工具之避險績效恆優於以大阪日經225指數期貨爲工具之避險績效。
This paper investigates the optimal hedge strategies between Nikkei225 index futures market and the underlying cash market using ARJI model and GARCH model. We analyze that either hedge portfolios of OSE Nikkei225 index futures or SGX Nikkei225 index futures have better hedge performance to refrain from the loss of trading on the underlying spot market. At the same time, due to the behavior of price jump, we utilize ARJI model to capture price jump behavior in order to enhance the hedging performance in short hedging horizon and compare different hedging horizons with GARCH model. The empirical results show that ARJI acquires better hedging performance than GARCH model when the hedging horizon is 5 days l0 days and 20 days and the result is inverse when the hedging horizon is 4odays and 60 days. Additionally, whether employing ARJI model or GARCH model, the SGX markets have superior hedging performance than OSE markets.
