本文提出以誤差修正指數加權移動平均（bias-corrected exponentially weighted moving average, Bias-corrected EWMA）模型取代Harris and Shen（2003）使用冪指數加權移動平均（power exponentially weighted moving average, Power EWMA）模型估計避險比率，其目的除了仍保有指數加權移動平均（exponentially weighted moving average, EWMA）模型在動態避險策略上易於估計的優點外，另外可避免使用Power EWMA模型時需事前主觀假設資產報酬率的分配型態而可能降低避險的效果。為驗證Bias-corrected EWMA模型的避險績效，本文以四種股價指數期貨為實證研究的對象，檢測三個不同研究期間下Bias-corrected EWMA模型、EWMA模型及Power EWMA模型的績效。整體而言，實證結果顯示 Bias-corrected EWMA模型相較於EWMA模型與Power EWMA模型有較佳的避險績效。

We propose the Bias-corrected EWMA model to estimate optimal hedge ratios for stock index futures. Our proposed method not only retains the easy usage characteristic of the EWMA model on dynamic hedge strategies but also captures the non-normality situations of the returns on assets. Using four stock index futures and three sample periods, we compare the optimal hedge ratios and hedge performances with the Bias-corrected EWMA model, EWMA model and Power EWMA model. Our empirical results show on average that the Bias-corrected EWMA model outperforms the EWMA and Power EWMA models.