本研究主要探討殖利率與總體經濟變數間關聯性，除了考慮先前學者所提出的物價與利率總體因素對殖利率有明顯影響力，本研究加入S&P500指數、採購經理人指數(PMI)以及消費者預期指數等具有代表景氣領先之變數，欲探討景氣循環指標是否可提高對殖利率的預測能力。另外，也考慮不同年期的殖利率間的交互影響，並且採用Vereda et al. (2007)所提出的改良式VAR模型為本研究的架構，研究此五變數對殖利率的預測能力。本研究的實證結果為，我們驗證先前學者所提出的消費者物價指數與利率資料是預測殖利率走勢的重要變數，若加入本研究所選取的三個代表景氣領先指標的變數可提升預測能力，其中消費者預期指數僅顯著影響3個月殖利率，不同天期殖利率資料亦透露出部分的訊息，故在進行殖利率預測的同時，仍可參考不同天期殖利率的走勢。 Some financial analysts emphasize the connections between macroeconomy and bond yields when they are making the strategic decisions about bond investments. This paper explores the relations between the term to maturities and the yields, and incorporates macro variables to predict Treasury bond yields of different maturities. This paper follows the work of Vereda et al. (2007) and explores the relations between US-Treasury bond yields and economic leading indicators. VAR models and Granger Causality Test are used to analyze the mutual influences among Treasury bond yields and macroeconomic variables. Our results support many authors’ opinions that Fed Fund rates and CPI index can predict movements of bond yields well. We additionally find evidences that the information of stock market index and PMI index can improve the forecasting performance for Treasury bond yields. Consumer side data is only useful in predicting 3-month yields in VAR models. Treasury bond yields of various maturities also reveal useful information when someone predicts the movements of yields.
