本文探討美國貨幣政策衝擊對拋補利率平價（Covered Interest Parity， CIP）偏差之影響，並分 析此一影響是否因金融海嘯前後、市場別與天期別而有所差異。本文以 G10 與新興市場國家 （emerging market economy， EME）的貨幣為研究對象，採用 1 年期、5 年期與 10 年期政府 債券 CIP 偏差作為應變數，並依循 Jarociński (2024) 之高頻識別架構，將美國聯邦公開市場 委員會 (Federal Open Market Committee， FOMC) 宣告拆解為傳統貨幣政策衝擊、路徑衝擊、 帶規模資產購買 (Large-Scale Asset Purchase， LSAP) 衝擊與資訊效果。實證分析方法上，本 文採用縱橫資料固定效果模型 (panel data fixed effect model) 與個別貨幣時間序列模型，並控 制交易商 (dealer) 槓桿變動與廣義美元指數 (Broad US Dollar Index) 之變動。 實證結果顯示，金融海嘯後 G10 國家的較長天期政府債券基差明顯轉為負值，且 G10 與 EME 國家之 CIP 偏差波動皆有所擴大。迴歸結果指出，金融海嘯前資訊效果為較穩定且一 致的顯著來源；金融海嘯後，傳統貨幣政策衝擊對 CIP 偏差之影響較不穩定，LSAP 衝擊與資 訊效果則主要反映在 G10 國家之 5 年期與 10 年期政府債券基差。個別貨幣分析亦顯示，1 年 期結果較為分散，而 5 年期與 10 年期結果較能支持長天期基差對資訊效果與 LSAP 衝擊之反 應。 本文結果說明，美國貨幣政策對 CIP 偏差之影響並非僅透過短期美元融資成本或傳統利 率管道發生，也可能透過長天期殖利率、期限溢酬、全球風險偏好與跨境資產配置進行傳導。 整體而言，本文發現美國貨幣政策衝擊對 CIP 偏差之影響具有明顯的時期別、市場別與天期別 差異。;This thesis examines the effects of U.S. monetary policy shocks on covered interest parity (CIP) deviations and investigates whether these effects differ across periods， markets， and maturities. The analysis focuses on G10 and emerging market economy (EME) currencies and uses gov- ernment bond CIP deviations at the 1-year， 5-year， and 10-year maturities as the dependent vari- ables. Following the high-frequency identification framework of Jarociński (2024)， this study decomposes FOMC (Federal Open Market Committee) announcement shocks into conventional monetary policy shocks， path shocks， LSAP (Large-Scale Asset Purchase) shocks， and central bank information shocks. The empirical analysis employs panel data fixed effects models and individual-currency time-series regressions， while controlling for changes in dealer leverage and the broad dollar index. The results show that G10 longer-maturity government bond bases turn more negative after the global financial crisis， while CIP deviations in both G10 and EME currency markets become more volatile. Regression results indicate that information shocks are the most stable and consistent source of significant effects before the crisis. After the crisis， conventional mon- etary policy shocks do not show stable effects on CIP deviations. In contrast， LSAP shocks and information shocks have more pronounced effects on G10 government bond bases at the 5-year and 10-year maturities. Individual-currency regressions further suggest that 1-year results are more heterogeneous， whereas the 5-year and 10-year results provide stronger support for the response of longer-maturity bases to information and LSAP shocks. Overall， the findings suggest that U.S. monetary policy affects CIP deviations not only through short-term dollar funding costs or the conventional interest rate channel， but also through long-term yields， term premia， global risk appetite， and cross-border portfolio allocation.
