本文主要之研究目的，即是要以時間數列分析法 (Time Series Analysis) 中之 介入模式（ Intervention Model，或稱干預模式），分析 l997 年 4 月以來亞洲各國在面 臨金融風暴的襲擊時，央行之干預此一事件點對匯率時間數列之影響。其次，探討各國央行 在棄守匯率後，不同的投機性投資攻擊模式對匯率時間數列不同之衝擊型態。最後，比較單 純之時間數列 ARIMA 模式以及干預模式之預測績效。 研究結果顯示，各國央行在金融風暴 中對匯率之干預的確有效，有延後金融風暴對匯率衝擊之效果。不同的投機性投資攻擊模式 造成亞洲國家顯著但不同型態的匯率貶值型態。

　　　　　The major purpose of this paper is to apply the intervention model of time series analysis to investigate the impacts of the financial crises and the intervention of Central Banks on the exchange rates of Asian countries since April, 1997. Additionally, this paper also examines the impacts of event points of financial crises on the time series of exchange rates after the abandonment of the policy of stable exchange rates by the Central Banks. Results indicate that the intervention of Central Banks is an effective tool on exchange rates during the financial crises. The intervention delays the impacts of financial crises on the exchange rates. As the reasons of speculative portfolio attacks being different, the financial crises result in a different significant devaluation pattern of currencies of Asian countries.