「避險」乃是期貨市場及選擇權市場最主要的功能之一，避險者只要持有適當的期貨合約或選擇權合約，便能減少其現貨部位所面臨的價格波動風險；至於避險成效的大小，則取決於避險比率是否適當而定。因此如何決定最適避險比率便成為學術界及實務界一個重要的課題。在過去避險的實證研究中，大多只著重於如何經由歷史資料，觀察期貨價格與現貨價格變動的相關性，並據以得出一個靜態的最適避險比率。事實上在一個充滿變數的經濟社會中，此一靜態避險比率之穩定性，一直受到相當大的質疑。因此本研究乃應用計量分析方法，提出五種最適避險比率之動態調整策略，並以S&P500股價指數期貨及選擇權作為避險工具，說明如何應用本文所建議之動態調整策略，以增進股票投資組合之璧顯成效。尤其要強調的是本文在評估各種避險策略之成效時，除了比較一般文獻上新採用的報酬率變異數減少百分比 ( R^2 ) 外，並比較各種避險策略在平均報酬率以及報酬率機率分配在偏度上的差異，以期提供更廣泛而公平的評估基礎。

With the changing nature of the real world it is unrealistic to assume that the hedge ratio estimated from historical data over a long period is constant. The dynamic nature of the hedge ration presents a practical problem in performing OLS to compute the optimal hedge ratio. Five econometric models for handling the nonstationary optimal hedge ratio are analyzed. They are: the pure random coefficient model, the Bayesian model, the univariate ARCH model, the bivariate ARCH model, and the generalized changing parameter model. In addition, these methods are applied to examine the dynamic hedging potentials for stock portfolios. It is found that the return-risk relatives from the dynamic strategies are substantially higher than those of the static hedging strategies while the return distributions of the former are also more positively skewed than the latter. The findings demonstrate the superiority of the more sophisticated dynamic hedging strategies.