After large volatility: empirical study of Taiwan stock market

[[abstract]]本論文之研究以台灣股票市場自2003年1月2日至2009年12月31日期間內，共1741個交易日均有成交紀錄之432支個股，以單因子模式、三因子模式與四因子模式加上經由AIC(Akaike information criterion criterion)選取之最佳配適模式的四種模式，使用GJR-GARCH(1 1)模型，計算每支個股每日之超額報酬，並以此超額報酬做為價格衝擊規模的基準，檢驗後10天的累積異常報酬(CAR)，來觀察是否存在過度反應或報酬連續性(即反應不足)。實證結果顯示，四種模式下的正向衝擊均有反應不足現象存在，而在負向衝擊時，不同模式則會產生不同的情況，且在分割樣本後之再實證，也有很不同的結果產生，這部分是反應不足、過度反應及支持市場效率假說三種情況都存在的。[[abstract]]This study adopts 432 shares, which have transaction records in 1,742 days during from January 2, 2003 to December 31, 2009, in Taiwan Stock Exchange. Using four models including Single Beta CAPM model, Fama-French model, and Carhart Fama-French model with a best fit model through AIC(Akaike information criterion) to select, by GJR-GARCH(1,1) model to calculate daily excess return of each share in each day. In order to investigate whether overreaction or return continuation (or underreaction) exists, this research uses excess return as trigger value of price shock to observe cumulative abnormal return (CAR) in ten days after price shock. The empirical results indicate that positive of price shock lead to phenomenon of underreaction in four models. However, negative of price shock lead to different phenomenon such as underreation, overreaction, and supporting efficient market hypothesis in varies models, even after dividing into sub-sample have also three different results