本研究主要在探討證券市場中私有資訊交易。由於Easley et al. (1996)建立資訊交易機率模型(probability of informed trading, PIN)，作為資訊不對稱的直接測度。Aslan et al.(2008)更證明了資訊相關的公司特徵變數與資訊交易機率的關聯性。但是也有學者持不同的意見，Duarte and Young (2009)就認為PIN不完全是衡量資訊不對稱的指標。因此我加入與資訊有關和非資訊不對稱相關的公司特徵，利用實證迴歸去說明這些公司特徵因子對於PIN的影響性，檢驗PIN是否除了衡量資訊不對稱外，也隱含了其他與資訊無關的因素。此外，為了解決公司特徵間常有強烈相關性的問題，納入主成分分析去萃取出公司特徵變數的重要成分，釐清公司特徵變數與PIN的關係。This thesis investigates informed trading in security market. Easley et al. (1996) built the model of probability of private information-based trade(PIN) as a direct measurement of informed trading. Aslan et al.(2008) considered that some firm characteristics could provide measures of a firm's information environment, and then proved the relationship between firm characteristics related to information and a firm's PIN. However, others have different opinions on this issue. Duarte and Young (2009) thought the PIN was not a total measure of asymmetric information. Therefore, this thesis attempts to add more firm characteristics related and unrelated to information into regressions, and test whether those variables have influences on PIN. In addition, a high correlation between firm characteristics may result in the problem of multicollinearity, and thus I use the principal component analysis to determine the influential components to detect the relationship between PIN and firm characteristics.
