亞洲金融危機後大陸出口貿易之動態分析

Abstract

本研究應用時間數列模式來建構中國大陸出口金額的預測模式，探討亞洲金融危機對中國大陸出口貿易的衝擊效應，並在不同的期間，分析亞洲金融危機後中國大陸出口金額的動態過程，同時比較多變量模糊時間數列（二因子模式、引導式模式、馬可夫模式）、時間數列ARIMA模式及向量ARMA模式在預測的適用性。由本文的實証結果發現，中國大陸當期出口金額受到當期干擾項，落差十二期的干擾項和落差十三期的干擾項所影響，而亞洲金融危機的衝擊，應不是造成該期間出口減少的主因。如果選取的期間資料是亞洲金融危機的第二年或第三年後，對於中國大陸的出口金額的外在衝擊程度已逐漸減低。就預測模式而言，當分析的資料期間愈長，則時間數列ARIMA模式的預測誤差較小；分析的資料期間愈短，則多變量模糊時間數列模式有較佳的預測能力。

This study applied time series models to forecast the exports of Mainland China around the years of Asian Financial Crisis (1997) and then, compared the forecasting appropriateness of the multivariable fuzzy time series models and the ARIMA time series model. The results showed that the disturbing factors of the exports of the current term, together with the lag 12 and the lag 13, affected the current exports. Namely, the exports reflected the Asian Financial Crisis one year ahead and the aftereffects were much less in the second and third years of the Asian Financial Crisis. In respect to the forecasting models, the mean square error (MSE) was smaller with ARIMA time series model when the data period was longer; otherwise, the multivariable fuzzy time series models appeared to be a better forecasting model.