本文分別對實證期間的擴張時期與緊縮時期建立一個new Keynesian的小型開放dynamic stochastic general equilibrium (DSGE) 模型，考慮價格與工資僵固性及金融市場之摩擦性；為一個前瞻性總需求總供給模型，特性如下：(1) 各變數設定為相關變數之實際觀察值與其長期趨勢值，或目標值，之離差的隨機變數；(2) 短期利率、實質匯率及貨幣總量均為內生變數；(3) 最適貨幣政策法則之設定與共整合檢定結果及一般化最小平方法估計式一致，貼現率為政策工具變數。本文先將各時期所建立的DSGE模型設定為state space model；其次，以最大概似法利用Kalman filter依特定遞迴步驟實證估計模型，部份係數值隨著時間的經過而改變。最後，由所估計的模型探討貨幣政策波動之動態效果。實證顯示貨幣政策波動對總體經濟具有穩定作用並且在擴張時期相對有效。 Under the new Keynesian framework, this article presents a small open dynamic stochastic general equilibrium (DSGE) model with price and wage rigidities and financial market frictions for the periods of expansion and contraction in business cycles, separately. The features of the forward-looking aggregate demand-aggregate supply model are (1) each variable in the model is expressed as the stochastic component of its actual counterpart (the deviation from a deterministic long-run trend, or a target value); (2) the short-run interest rate, the real exchange rate, and the monetary aggregate are all endogenous; (3) the best monetary policy discretion is consistent with the results of the cointegration tests and generalized least-squares estimates; rediscount rate is the policy operating target. With time varying coefficients, we employ a maximum-likelihood procedure with a Kalman filter to estimate the corresponding state space model specified by each DSGE model. Finally, the best fit is used to analyze the dynamic effects of a monetary policy shock. The estimates reveal that a monetary policy shock can stabilize the economy and is more effective during the expansionary period.
