本研究藉由Kose et al. (2008)所建構的動態因子模型,結合SVAR模型(Structural Vector Autoregression )探討美國貨幣政策對其經濟體背後因子的影響,並進一步剖析波動的來源,將之區分為總體因子、類別因子與個別因子,從中了解無法觀察到的隱藏因子(latent factor)對於經濟體潛在的演化影響,與其所引伸的經濟意涵。 本研究樣本架構依據Stock與Watson(2002)總體樣本類別,擷取其中75項總體變數並加以分類,分為工業、勞動、貨幣、匯率、物價、房地產六大類,藉由動態因子模型萃取其共同因子,再利用衝擊反應分析了解貨幣政策所帶來的影響。研究期間為2000年1月至2014年12月。研究結果發現,波動來源在類別因子上占了30%,較總體因子的12%顯著,此亦顯示經濟體中各市場所帶來的市場波動是為重要的,也因此以往文獻將焦點放在名目或實質經濟變數的影響,可能忽略市場波動所提供的重要信息。再者貨幣政策所帶來的衝擊,使得總體、工業、勞動、物價因子上皆呈現強度增強的效果,此效果亦顯示各市場所能承受衝擊的經濟彈性(resilience of the economy),且另一方面在貨幣政策的衝擊在總體、勞動、匯率、貨幣因子皆有延遲效果,也具以刻畫貨幣政策的落後時間。
This article combines dynamic factor model developed in Kose et al.(2008) and the Structural Vector Autoregression Model (SVAR Model) to analyze the impact of US monetary policy on its economic activity. We also take advantage of both models to decompose the fluctuation as macro-factor, classified-factor and individual-factor in order to realize the resource from the fluctuations. By doing so, we can clearly recognize the influence of latent factor on economic activity and the meaning of the impact. The dataset includes 75 updated macroeconomic indicators used by Stock and Watson (2002), which involve several measures of industry, labor, money, exchange rate, price and house. These indicators have been found to collectively contain useful information about the state of the economy for the appropriate identification of monetary policy shocks. Our main finding is that the macro-factor accounts for 12% of the fluctuation and the classified-factor accounts for 30% of the fluctuation. Therefore, the result implies that the information from the market is significant. The past paper might ignore the key. Moreover, the impacts of monetary policy suggest that the monetary shock would increase the strength of the macro-factor, industrial-factor, labor-factor and price-factor. Hence, we can understand the market’s resilience of the economy and the lag time of the monetary policy by impulse response analysis.