綜合同時指標是否同時？美國與台灣的實證比較

本文重新檢討美國及台灣的綜合同時指標對於景氣循環轉折點的認定之表現。不同於過去的相關文獻，我們首先討論綜合同時指標的不同的成長率定義、不同的資料頻率，是否會改變景氣轉折點的認定表現，以不同的成長率定義、不同的資料頻率進行實證研究可以讓我們瞭解景氣指標在認定或預測景氣轉折點的頑強性。我們亦針對編製綜合同時指標的各個變數，其是否爲編製綜合同時指標的優良指標進行初步探討？實證結果顯示美國官方所編製的綜合同時指標或是其各數列變數，都能夠準確地認定出美國的景氣轉折點。台灣經建會編製的綜合同時指標在認定景氣轉折點的表現常有『誤認』的情況發生，多變量馬可夫轉換因子模型的分析亦顯示台灣用來編製同時指標的各個變數都能與景氣循環具有一定程度的同時性，只是於1990年之後在認定景氣轉折點的能力並不理想。爲何會有如此結果，值得編製單位深思。

關鍵字

景氣轉折點；綜合同時指標；馬可夫轉換模型

並列摘要

This paper revisits the performance of the composite coincident indicator of the U.S. and Taiwan in identifying business cycle turning points. We consider various measurements of the growth rate of the indicator and different frequency of data for empirical study. The univariate and multivariate Markov Switching models are employed in empirical study. The estimation results provide considerable evidence that the composite coincident indicator, compiled by the NBER, performs very well in identifying the U.S. turning points. In case of Taiwan, the composite coincident indicator compiled by the CEPD, will produce false signal in identifying Taiwan's turning points, especially in the post-1990s.

並列關鍵字

Turning point ； coincident indicator ； Markov Switching model

參考文獻

Chen, S-W.,Lin, J-L.(2000).Identifying turning points and business cycles in Taiwan: a multivariate dynamic Markov-switching factor model approach.Academia Economic Papers.28,289-321.

Chen, S-W.,Lin, J-L.(2000).Modelling business cycles in Taiwan with time-varying Markov switching models.Academia Economic Papers.28,17-42.

Albert, J.,Chib, S.(1993).Bayesian analysis via Gibbs sampling of autoregressive time series subject to Markov mean and variance shifts.Journal of Business and Economic Statistics.11,1-5.