本研究旨在探討四個金融危機期間，八個亞洲新興國家股市的全球與區域波動蔓延效果；採納Bekaert ＂et al＂.（2005）以「超額相關性」為蔓延效果之定義，並應用Baele and Inghelbrecht（2010）的兩因子結構狀態轉換模型與蔓延效果檢定模型，對Beirne ＂et al＂.（2008）建議的條件變異數進行檢測。經實證發現：（1）本文設定的兩因子動態報酬模型，可以捕捉所有樣本國家股市報酬的長期基本面與短期週期性的聯動關係，而且全球與區域風險會隨潛在狀態變數與結構變數之變動而變動；（2）在四個金融危機期間，除網路泡沫化期間外，均有部份國家股市存在全球或區域的波動蔓延效果；（3）整體而言，東協國家股市存在全球或區域的波動蔓延效果，而非東協國家股市則未發現；（4）部份國家股市面臨正負報酬衝擊時，存在不對稱的波動蔓延效果。

The purpose of this paper is to explore global and regional volatility contagion on eight Asian emerging markets during past four financial crises. Accepting the definition as excess correlation of contagion by Bekaert ＂et al＂. (2005), this paper applies two-factor structural regime-switching model and contagion test model modified by Baele and Inghelbrecht (2010) to test the conditional variance as volatility contagion suggested by Beirne ＂et al＂. (2008). The empirical results show: (1) the two-factor dynamic return model of this paper can correctly characterizes the long-term fundamental linkages and short-term cyclical variations between markets, and global and regional exposures of all markets are time-varying with changes in latent regime variables and structural instruments; (2) global and regional volatility contagion on some individual markets are observed during each financial crisis period except Nasdaq rash; (3) general evidences for global and regional volatility contagion on ASEAN countries are found, but not on non-ASEAN countries; (4) some markets have significantly asymmetric volatility contagion between positive and negative return shock.