本文以 GJR-GARCH 為分析模型，針對所選八家台灣上市公司股價所計算之每日對數報酬率(daily log returns)，對於各種不斷到達的新增訊息所引起的波動反應。所納入條件變異數方程式的訊息到達(解釋變數)分別為：(1)同日成交數量(2)成交量變動率(3)星期一與星期五之日曆效應(4)不同權值規模(size-based)投資組合間的波動外溢效果。研究結果發現(1)同日成交量對於台股權值較低的小公司，有能力捕捉其波動性，但是對於權值偏高的大公司，其解釋能力顯有不足(2)成交量變化普遍會導致公司報酬率的波動(3)臺灣股市波動性並不具有星期五效應，至於星期一效應也只出現在部分的小公司(4)不同規模的投資組合間雖然互有波動外溢現象，但其不對稱性非常明顯， 亦即訊息到達後，先造成大公司股價的波動，此波動再進而影響到小公司，引起小公司股價的波動。 / Applying the GJR-GARCH model to the daily returns of eight selected firms from Taiwan stock market, this paper examines response of variance volatility to various information arrivals which separately include (1) concurrent trading volume (2) change in trading volume (3) calendar effects, especially Modnay and Friday effects, and (4) asymmetric volatility spillover between two sized-based portfolios. The results find that concurrent trading volume as a proxy of information arrival dramatically reduces volatility persistence of the small firm's conditional variance, but has little influence on large firm's, and change in trading volume cause significant change in conditional variance. Although there is a conjecture that the volatility in stock markets may be higher on Monday and Friday, it can't be found in this study. The results also strongly support that the volatility spillover effect from larger to small portfolio is more significant than that from smaller to large portfolio.