| 摘要: | 本研究以美國芝加哥商業交易所(CME)的S&P500 指數期貨、COMEX黃金期貨及NTMEX西德州原油期貨來進行對其現貨市場的避險，研究期間取自2001年1月1日至2011年12月31日止。運用不同避險績效的衡量方法，包括變異數(Variance)與半變異數(semi-variance)、風險值(VaR)、條件風險值(CVaR)等來估計OLS、CCC-GARCH、DCC-GARCH、使用ARJI調整後之CCC-GARCH及DCC-GARCH等避險模型之樣本外避險績效。
實証結果顯示在S&P500或是西德州石油的日及週避險績效上，調整前之避險績效大多優於調整後，此與Hyde, Nguyen and Poon (2008)結果不一致。可能原因為研究的標的為股票的投資組合，且其估計的資產個數很多，為MSCI 34個不同國家的指數，其相關性較低。在黃金的避險上所得到得結果則與Hyde, Nguyen and Poon (2008) 結果一致，不管是在週或是日的避險績效上都是調整後的模型優於調整前。另在逐日避險跟採用週避險策略並無太大差別。但投資人考慮的下方風險，採週避險策略下，則調整後之模型會比沒調整前之模型，更能提供此類投資人更好的避險績效。
This study is to conduct hedging the cash markets of the S&P 500 index futures, COMEX gold futures of U.S. Chicago Mercantile Exchange (CME) and NTMEX West Texas crude oil futures. The study period is taken from January 1, 2001 as of December 31, 2011. Measurement methods of different hedge performance, including variance and semi-variance, VaR, etc. will be applied to estimate OLS, CCC-GARCH, DCC-GARCH, and out-of-sample hedge performance of CCC-GARCH and DCC-GARCH and other hedge models adjusted by using ARJI.
The results shown that the daily or weekly hedge performance of S&P 500 or West Texas Crude oil before the adjustment is more excellent than that after the adjustment, which is not consistent with the results of Hyde, Nguyen and Poon (2008). The possible reason is that the study object is stock investment group and the numbers of estimated capitals are many—index of MSCI 34 different countries, which has lower relevance. The results obtained on the hedge of gold are in consistent with the results of Hyde, Nguyen and Poon (2008). Both the daily and weekly hedge performance are better than those before the adjustment. In addition, day by day hedge policy and weekly hedge policy have no big difference.If the investors care more about the downside risk, and apply weekly hedge policy, the models being adjusted will provide better hedge performance for those investors before the adjustment. |
