本論文旨在探討市場波動度對新台幣兌美元匯率之預測能力,並進一步評估將此預測資訊導入利差交易策略後的實務績效表現。研究採用2007年4月至2024年12月的月資料,資料包含台灣與美國之利率差、消費者物價指數(CPI)年增率差、實質匯率、新台幣兌美元即期匯率,以及市場波動度指標,包括台灣波動度指數(TVIX)、美國市場情緒指標(VIX)及全球外匯波動度指數(VXY)。 在研究方法方面,本文首先以傳統經濟基本面為基礎模型,逐步加入市場風險指標,建立四個匯率預測模型(EF、MT1、MT2、MT3),並使用滾動視窗法進行樣本外預測分析,評估其在不同預測期(1個月、3個月、6個月)的精準度及方向預測正確率。進一步,將模型的預測結果應用於五種利差交易策略之模擬,包括傳統利差交易(CT)、基本面擴充策略(EF)與三種納入市場風險變數的擴充策略(MT1至MT3),並評估各策略在年化報酬率、波動度與風險調整後績效(Sharpe ratio)的表現。 研究結果顯示,市場風險因子顯著提升了匯率預測模型的樣本內解釋能力與樣本外預測績效,尤其以納入全球外匯市場波動度指標(VXY)之模型表現最為突出。此外,在利差交易策略績效分析上,加入市場風險指標之MT系列策略在各預測期間的報酬率與Sharpe值皆優於傳統CT策略,驗證市場風險指標在實務策略中的重要性。本研究成果在學術與實務層面皆具重要貢獻,並提供未來新興市場貨幣相關研究與投資操作的重要參考依據。 ;This thesis investigates the predictive power of market volatility indicators—Taiwan Volatility Index (TVIX), U.S. Volatility Index (VIX), and Global FX Volatility Index (VXY)—on the TWD/USD exchange rate and evaluates their effectiveness when integrated into currency carry trade strategies. Monthly data spanning from April 2007 to December 2024, including interest rate differentials, CPI inflation differences, real exchange rates, and volatility indices, are utilized. Forecasting models based on economic fundamentals are extended by incorporating these market volatility indicators. A rolling-window method is applied to evaluate out-of-sample forecasting accuracy across different forecast horizons (1-month, 3-month, and 6-month periods). The forecasting results are further employed to simulate several carry trade strategies. The empirical findings indicate that models incorporating market volatility factors substantially enhance exchange rate prediction accuracy and strategy performance, with the VXY-based model delivering the strongest overall results. This research underscores the practical value of market volatility indicators for both exchange rate forecasting and currency investment strategies in emerging market contexts.