透過您的圖書館登入
IP:3.145.201.71
  • 學位論文

連鎖值法下總體計量模型的建構

Build a Macro-econometric Model Based on Chain-Linked Method

指導教授 : 林建甫
本文將於2025/08/17開放下載。若您希望在開放下載時收到通知,可將文章加入收藏

摘要


行政院主計總處於2014年頒布了以連鎖值法取代定基法之消息,因此主計總處的總體資料此後都會使用此方法來進行計算。此方法雖有助於改善定基法的替代性偏誤,但也因為基期不斷變換的情況下,使得在連鎖值法下的實質國內生產毛額,並不再具有可加性,這對從支出面出發的大型計量總體模型來說,可說是相當大的缺點,因此本文將以大型計量總體模型為基礎,加入連鎖值法,建立一總體模型,並探討在連鎖法下,本研究所需作的模型設定上的改變,以及資料處理上會遇到什麼困難。 本文共計有60條方程式,其中結構方程式為38條,定義式則為22條。樣本前間則為2002年第一季至2019年第三季間。本研究將透過這些方程式以及變數樣本資料,建構出具有連鎖值法特色的總體計量模型。 建構模型完後,本研究將對其作配適度以及預測能力的評估,並進行檢討,是否模型還有改進的空間。另外本研究模型也進行了基準預測的部分,此部分是針對台灣的經濟所做的長期趨勢預測,而本研究透過連鎖值法建立的模型在此部分仍然有所不足之處,在預測結果中,雖然趨勢一致,但有稍微偏離的部分,而這導致預測結果的不精確,而這些預測結果可用來判斷模型的高估或是低估,可做為未來改善模型的各項變數使用的基礎。

並列摘要


The Directorate-General of Budget, Accounting and Statistics, Executive Yuan announced in 2014 that the chain-linked method replaced the fixed-base method. Therefore, the overall data of the Chief Accounting Office will be calculated using this method thereafter. The advantage of the chain-linked method is reducing the alternative bias because its based period is constantly changing. However, this character also makes it doesn’t additive when we calculate the real gross production. The real gross production cannot be calculated by the addition of real variables, is really disadvantageous to construct a macroeconomic model. Therefore, we try to construct a SEAs model that uses the data based on the chain-linked method and figure out how to fix this problem. The model will include 38 Structural equations, 22 definition equations. The sample space is from 2001 Q1 to 2019 Q3. We use this model to construct a SEAs model under the chain-linked method. After constructing the model, we will evaluate its goodness of fit and predictability and consider which part of the model has to be improved. In this model, we also do the benchmark forecast. This part is a long-term trend forecast for Taiwan’s economy. We can find some shortcomings of the model by the results of the benchmark forecast. In the forecast results, the trend of the model is consistent with reality, but there is a slight deviation, which leads to inaccuracy of the prediction results. These prediction results can be used to judge the overestimation or underestimation of the model and can be used as the basis for improving the variables which the model used in the future.

參考文獻


中文文獻
何金巡 (2005),「供需估測季模型9408號」,行政院主計總處第三局。
何金巡、林建甫、周麗芳 (2005),「油價、景氣與政府財政的總體經濟分析」,2005年總體經濟計量模型研討會,中央研究院經濟研究所及行政院主計總處。
何金巡、林建甫、周濟、周麗芳 (2007),「油價高漲年代政府油價政策對政府財政政策及總體經濟之影響與預測」, 2007年總體經濟計量模型研討會,中央經濟研究院。
林金龍 (2003),「利率政策的傳遞機制及其對總體經濟金融影響效果之實證分析」,《中央銀行季刊》,25(1),頁5-48。

延伸閱讀