本研究提出一個結合Power EWMA估計式與歷史模擬法的兩階段估計風險值模型,以克服歷史模擬法潛藏無法捕捉與時改變波動的行為而導致風險預測能力降低的問題。此外,本文並提出利用峰態係數法來推估隨時間改變的波動行為之Power EWMA估計式,可避免運用GARCH模型時必須事先設定條件分配以求解高度非線性參數估計的問題,而保有歷史模擬法容易估計的特性。本文經失敗率分析及Kupiec(1995)統計檢定結果,證實本研究所提出之方法具有簡單操作以及正確預測風險值的能力。
This study proposes a two-stage approach combining the Power EWMA estimator with the historical simulation when estimating Value-at-Risk. Our method can avoid estimating parameters to forecast the variance when using GARCH model and retains the easy usages characteristic of the historical simulation approach. In addition, we also use the kurtosis coefficients to estimate the distribution form for capturing the time-varying volatility. In the light of results of the failure rates and Kupiec test, the empirical result shows that the proposed method can considerably enhance the estimation accuracy of Value-at-Risk.