本文以2001年12月至2008年5月間之台灣加權指數 (TX)、台指選擇權 (TXO)與台指期貨(TXF)日內及日資料，進一步於包含式模型中分別加入歷史波動率 (history volatility, HV)、真實波動率 (realized volatility, RV)、價平選擇權隱含波動率 (at-the money option impilied volatility, ATM)、決定性波動率函數 (deterministic volatility function, DVF) 等波動率預測模型之預測能力，並驗證TXF或台指指數何者可做為投資TXO市場之參考標的物。 實證結果顯示，由波動率預測模型之調整後R2以及相關係數分析指出，TXF之波動率預測模型對於TXO隱含波動率有較佳的解釋能力，其中以TXF之移動平均天數為十天的歷史波動率以及取樣頻率為9分鐘之真實波動率的解釋能力較高；由此證據顯示，相對於台指指數，TXF可為對於投資TXO市場較佳的參考標的物。投資績效顯示，無論是買賣買權或賣權的投資策略皆可獲得正向報酬；若在考慮交易成本下，則是以持有期間為五天之投資績效多能獲得正向的報酬。 This paper puts the volatility-forecasting models, including history volatility(HV), realized volatility(RV), at-the money option implied volatility(ATM) and deterministic volatility function(DVF) into the encompassing regression to examine the forecasting ability of volatility model for TX, TXO and TXF during December 2001 to May 2008. Also, we verify that TXF or TX which is the better reference object for investing in TXO market. Based on the results of the adjusted-R2 and correlation coefficient of volatility-forecasting models, we find that the forecasting ability for the implied-volatility of TXO is better with volatility-forecasting models of TXF than TX. For the forecasting-models of TXF, the history volatility which is estimated by 10-day moving average and the realized volatility which is estimated by 9-minute intraday sampling frequency have the higher explanation ability. By above-mentioned, we can prove that TXF is the better reference object for investing in TXO market than TX. However, in the analysis of investment strategy, the results indicate that the strategy has positive performance, regardless of call or put option strategy. Our results also show that only the investment strategy of 5-day holding period has the positive performance with the consideration of transaction cost.