響盈餘宣告後股價飄移(post-earnings-announcement drift, PEAD)的因素有許多，本 文從中整理出三大原因分別為資訊不對稱、流動性以及投資人的行為偏誤， 並進一步探 討三大因素對飄移的影響，同時驗證三大因素都考慮的情況下對股價漂移的影響，檢驗 哪項因素是造成漂移最具解釋能力的原因。 整理出三項主要因素下， 在每項因素中各選取兩項變數來對飄移做驗證。 股價飄移 以宣告後 60 天的累積超額報酬(CAR)為觀察， 並將其依正負號分為 CAR+ 及 CAR- 以 利於觀察股價飄移方向。資訊不對稱部分選擇PIN和AIV兩變數作為代替， Realize spread (Realsp) 和 Illiquidity 則為流動性因素的代理，而行為偏誤為過度自信及注意力不集中 兩類，又分別以 Turnover 和 Friday (公司盈餘宣告日是否為星期五) 兩變數代理。本文 主要研究 2014 年全年公司盈餘宣告後的股價飄移，並依各項變數所需不同之數據抓取 2013 至 2015 年的資料做計算。 大多數公司為季宣告，因此每三個月為一次研究事件， 實證研究顯示在 CAR+ 或 CAR- 的結果中，只有流動性這項因素是兩個變數 (Realsp 及 Illiquidity) 都顯著的，因此可以說在這三項因素中流動性是對 PEAD 最具 解釋能力的， 其中 Realsp 這項變數的顯著水準更高，顯示本文結果表示其比起 Illiquidity 能夠更好的衡量流動性。;Many factors cause post-earnings-announcement drift (PEAD). This thesis points out three major causes as information asymmetry, liquidity, and investor behavioral bias, and further explores the impact of these factors on the drift of stock price. This impact is then verified and tests are conducted to find out which factor is the most crucial cause of PEAD. After summarizing three main factors, I select two variables in each factor to examine their effect on drift. The drift of stock price is observed based on cumulative abnormal return (CAR), which is divided into CAR+ and CAR- to facilitate the observation of the direction of stock price drift according to the sign. For the information asymmetry part, the probability of informed trading (PIN) and abnormal idiosyncratic volatility (AIV) are selected. Realize spread (Realsp) and Illiquidity are proxies for liquidity factors. Overconfidence and inattention represent behavioral biases and I use Turnover and if firms announce on Friday or not as proxies. This thesis mainly studies the drift after the companies’ earnings announcement for the whole year of 2014, and as most companies announce quarterly, every three months is a research event. Empirical research shows that in the regression results of CAR+ and CAR-, only the liquidity factor (Realsp and Illiquidity) is significant for both. Therefore, it can be said that liquidity is the most important reason for PEAD among these three factors. Furthermore, the significant level of the variable Realsp is higher, showing that the results of this thesis indicate that it can better measure liquidity than Illiquidity
