摘要: | 本研究以EURUSD、USDJPY 、 與AUDUSD為研究對象,探討外匯選擇權隱含波動率對於即期外匯市場波動度之預測能力。本研究以Garman and Klass (1980)波動度估計方法做為即期外匯市場波動度之替代變數,進而比較GARCH模型及implied volatility對波動度的預測能力。研究結果發現一個禮拜、一個月期implied volatility的預測能力皆優於GARCH模型。另外,不論在模型樣本內及樣本外之預測能力表現上,一個禮拜、一個月implied volatility的預測能力也優於GARCH模型。 This thesis uses EURUSD, USDJPY, and AUDUSD to study the forecasting ability of currency option implied volatility. Using the Garman and Klass (1980) estimator as the proxy of spot market volatility, we compare the fitting performance of implied volatility and GARCH in a regression framework. The results show that 1-week and 1-month implied volatilities fit better than the volatility estimated from the GARCH model. Moreover, 1-week and 1-month implied volatilities outperform GARCH forecasts in both in-sample and out-sample forecasting evaluation. |