摘要
笔者基于中国A股市场的分时高频交易数据,构建了连续交易操纵识别和测度模型。在此基础上结合2001—2017年证监会公布的市场操纵行政处罚案例,运用计数模型进行了上市公司特征对市场操纵影响的实证分析。研究结果表明,笔者构建的连续交易操纵测度模型识别成功率高达79%,且我国股票市场存在市值规模较小、经营绩效水平较差、前期发生市场操纵概率较高的上市公司股票更容易被市场操纵的特点。上述结果在对控制变量进行替换并截取不同样本区间进行反复实证检验后依然稳健。据此,笔者建议要提高上市公司质量,严厉打击市场操纵行为,并对不同类型上市公司股票采取更具有针对性的市场操纵监管措施。笔者首次采用分时高频交易数据有效监测了我国股票市场发生可疑连续交易操纵行为,这对打击市场操纵、维护金融市场安全具有重要意义。
Based on the high-frequency trading data of Chinese A-share market,this paper constructs a manipulation recognition model.Combined with the administrative punishment cases of market manipulation published by CSRC from 2001 to 2017,we use the counting model to analyze the influence of the characteristics of listed companies on market manipulation.The success recognition rate of the model is as high as 79%and China s stock market has the characteristics that listed companies which have smaller market value,poor operating performance,and higher probability of being manipulated in the early period are more easier to be manipulated.The above results are still robust after replacing the control variables and intercepting different sample intervals for repeated empirical tests.Accordingly,the author suggests improving the quality of listed companies,cracking down on market manipulation,and adopting more targeted supervision measures for diffe-rent types of listed companies.This paper not only uses time-sharing high-frequency trading data for the first time,the author used time-sharing high-frequency trading data to effectively monitor the suspicious continuous trading behaviors in China s stock market,which is of great significance to combat market manipulation and maintain the security of the financial market.
作者
李志辉
邹谧
LI Zhi-hui;ZOU Mi
出处
《中央财经大学学报》
CSSCI
北大核心
2018年第12期25-36,共12页
Journal of Central University of Finance & Economics
基金
国家社会科学基金重大项目"金融风险度量的新理论与新方法及其在中国金融机构的应用"(项目编号:14ZDB124)
国家自然科学基金青年项目"金融机构系统性风险敞口与贡献的度量及监管研究--基于金融网络视角的分析"(项目编号:71703111)
教育部人文社会科学研究青年基金项目"证券市场异常波动背景下价格限制类交易机制对市场质量的影响--基于效率与公平的视角"(项目编号:16YJC790049)
关键词
市场操纵
连续交易操纵识别模型
上市公司特征
分时高频交易数据
Market manipulation
Manipulation by actural purchases recognition model
Characteristics of listed companies
Time-sharing high-frequency trading data