無模型設定隱含波動性模型預測績效一台指選擇權市場實證

Forecasting Performance of Model-free Implied Volatility for the Taiwan Option Market

This paper aims to examine an estimator of model-free implied volatility (MF-IV), as derived by Britten-Jones and Neuberger (2000), investigating its information content within the index options market of Taiwan. Based upon a comparison between the forecasting performance of the MF-IV model and that of other volatility forecasting models, such as Black-Scholes implied volatility (BS-IV), historical volatility and GARCH volatility, our empirical results demonstrate that the MF-IV model possesses higher information efficiency, subsuming all information contained within the historical and GARCH (1,1) volatility models in the forecasting of fixture realized volatility for a weekly forecasting horizon. 本文目的旨在以Britten-Jones與Neuberger（2000）所推導之無模型設定隱含波動性（model-free implied volatility, MF-IV）模型，檢定台指選擇權之資訊内涵。藉由MF-IV模型與其他波動性模型如Black-Scholes隱含波動性（BS-IV）、歷史波動性模型與GARCH（1,1）模型之預測績效比較，本文發現MF-IV模型完全包含歷史波動性模型與GARCH（1,1）模型在預測未來5天買權到期時現貨市場真實波動性所具有之資訊。

本文目的旨在以Britten-Jones與Neuberger（2000）所推導之無模型設定隱含波動性（model-free implied volatility, MF-IV）模型，檢定台指選擇權之資訊内涵。藉由MF-IV模型與其他波動性模型如Black-Scholes隱含波動性（BS-IV）、歷史波動性模型與GARCH（1,1）模型之預測績效比較，本文發現MF-IV模型完全包含歷史波動性模型與GARCH（1,1）模型在預測未來5天買權到期時現貨市場真實波動性所具有之資訊。