| 摘要: | 近年來有學者提出了幾個由證券交易資料萃取出的資訊不對稱衡量指標。這些指標的用途主要是探討資產訂價的問題, 但是也有可能應用在公司理財或會計的研究上。然而這些指標是否可以有效地衡量資訊不對稱,仍然是個有待探討的實證議題。本計畫旨在提供此議題的部分答案,作法是探討這些指標與傳統上作為資訊不對稱或流動性的公司特徵間的關係。本計畫在幾個部分異於前人關於資訊交易機率及逆向選擇價差的類似研究。首先, 本計畫關注探討近年來發展的新指標, 包括 Durate and Young (2009) 所提出的修正資訊交易機率(AdjPIN), Hwang and Qian (2010) 所提出的資訊交易誤差修正係數(ECIN), 以及Easley, de Prado and O’Hara (2010) 所提出的交易量同步化資訊交易機率(VPIN)。其次, 前人僅研究資訊指標與資訊代理變數的關係。本計畫引進一些與資訊不對稱無關的流動性代理變數, 希望這些變數與資訊指標的關係有助解釋這些指標。最後, 本研究以主成分分析協助解釋公司特徵變數, 並以主成分回歸減少自變數個數, 以期釐清資訊指標的意義。Quite a few measures of information asymmetry extracted from security trading data have been developed recently, which are aimed to answer the questions in asset pricing and may have further applications in corporate finance and accounting. Whether or how well they measure information asymmetry, however, is an empirical question. This project will shed some lights on this question by examining the relations between some information asymmetry measures and firm characteristics which are commonly used to proxy for information asymmetry or liquidity. It extends previous similar studies on the probability of informed trading (PIN) and on the adverse selection component of bid-ask spreads in several ways. Firstly, this project examines a few newly developed measures for information asymmetry, namely, the adjusted probability of informed trading (AdjPIN) by Durate and Young (2009), the error-correcting coefficient of the potentially informed trades (ECIN) by Hwang and Qian (2010), and the volume-synchronized probability of informed trading (VPIN) by Easley, de Prado and O’Hara (2010). Secondly, unlike previous studies that only investigate the relationship between information asymmetry measures and information asymmetry proxies, this project adds liquidity variables unrelated to information asymmetry to the analysis, in the hope that the relationship of these variables and the information measures, or the lack of it, may help clarify the meanings of the measures. Thirdly, the principal component regression is proposed to reduce the number of independent variables and to help with their interpretations. 研究期間：10008 ~ 10107 |
