摘要
准确测度信息不对称是研究金融市场信息不对称问题的基础。在详细考察经典的信息交易概率模型的建模过程和假设条件的基础上,针对信息交易者和非信息交易者在不同信息状态下的订单提交状况,提出考虑外生信息状态概率、异质期望和交易活跃度的改进的信息交易概率模型;通过带机制转换的向量自回归模型估计外生信息状态概率,综合考虑多方面的市场交易信息,提高模型估计的准确性。实证结果表明,改进的信息交易概率模型可以较好地刻画证券市场的信息不对称程度。从截面上看,信息交易概率与标的资产的垄断优势和话题性相关,垄断优势越强的股票信息交易概率越低,话题性越强的股票信息交易概率越高;从时间上看,信息交易概率的变化受公告信息的影响,公告信息发布前的信息交易概率较高,公告信息发布后的信息交易概率较低。因此,在防范信息风险时,应重点关注话题性较强的版块,特别是股价比较容易受舆论影响波动的个股;同时,在公告信息发布之前的时段,应对市场波动和信息动向给予更多关注。
Measuring information asymmetry accurately is the research basis of financial market's information asymmetry issue. Based on the modeling and assumption of the classical PIN model, this paper examines different order submission behavior of in- formation traders and non-information traders under different information states. Herein, an improved PIN model, which takes the information state probability as the exogenous variable and considers the heterogeneous beliefs and trading activity, is built up. The exogenous information state probability is estimated by Markov-switching vect()r autoregressive (MSVAR) model, which takes various market trading information into consideration, so the accuracy of the model could be improved. The empirical re- sults indicate that the advanced model is more accurate in measuring the information asymmetry of financial market. First, the probability of informed trading (PIN) is related to the monopolistic advantage and topicality of underlying asset. On one hand, strong monopolistic advantage always results in a lower PIN; on the other hand, high topicality is usually accompanied by high PIN. Furthermore, the PIN vibration is influenced by reports. The PIN is always in a high level before the information being opened to the public, and subsequently decreases significantly after being reported. Therefore, for preventing infomiation risk, it is needed to focus on the stocks that are easy to be influenced by public opinion and to pay more attention on the market fluctuations and the trends of information before announcement of related information released.
出处
《管理科学》
CSSCI
北大核心
2014年第6期121-131,共11页
Journal of Management Science
基金
国家自然科学基金(71171144
71471129)
高等学校博士学科点专项科研基金(20130032110016)~~
关键词
信息风险
PIN模型
交易活跃度
异质期望
信息状态
information risk
PIN model
trading activity
heterogeneous beliefs
information state