| 摘要: | 本研究嘗試探討月度頻率下，各項總體經濟如何影響短、中、長期公
債券殖利率，為此，選取了五個重要的總體市場變數：「美國失業率」、
「美國製造業採購經理人指數」、 「美國核心消費者指數」、「 VIX 波
動率指數」、「標普 500 指數報酬率」，藉由美國三個月、六個月、一年
期、二年期、三年、五年、十年及二十與三十年的公債殖利率利用主成分
分析法找出殖利率曲線「 Level 」、「 Slope 」及「 Curvature 」三因子
後並研究各控制變數對於短、中、長期公債殖利率之變化。樣本取樣區間
為 2014 年 1 月 1 日至 2023 年 12 月 31 日共 10 年 120 筆月資料，進行
深入統計分析進行實證研究，結果發現，殖利率曲線三因子對於各天期公
債殖利率之解釋力已達 99%，但若能加入總體經濟變數將會使得整體模型
效果更佳，另外也發現對於短期殖利率變化解釋效果更為優秀，解釋能力
隨天期變長而下降。;This empirical research attempts to explore how various macroeconomic
factors affect short-term, medium-term, and long-term bond yields on a
monthly frequency. For this purpose, five significant market variables were
selected: "U.S. Unemployment Rate," "U.S. Manufacturing Purchasing
Managers Index," "U.S. Core Consumer Index," "VIX Volatility Index," and
"S&P 500 Index Returns". Using the principal component analysis to derive
the "Level", "Slope", and "Curvature" of the yield curve from U.S. three-
month, six-month, one-year, two-year, three-year, five-year, ten-year, and
twenty and thirty-year bond yields, this study then examines how each
control variable influences the changes in short-term, medium-term, and long-
term bond yields. With a sample period from January 1, 2014 to December
31, 2023, comprising 120 monthly data points over ten years, an in-depth
statistical analysis was conducted for this empirical research. The results
revealed that the three factors of the yield curve can explain up to 99% of
the yield of bonds of various periods. Yet, incorporating macroeconomic
variables would improve the overall model significantly. Additionally, we
discovered that the model explains the changes in short-term yields better,
with its explanatory power decreasing as the term lengthens. |
