股指期货操纵预警的Logistic模型实证研究

发布时间：2018-09-12 05:08

【摘要】：通过建立Logistic回归模型判别股指期货被操纵的可能性,将股指期货是否被操纵的问题转化为根据一定时期内股指期货市场的波动性以及流动性指标计算股指期货在一定时间内被操纵的概率,建立了股指期货操纵事件的预警模型.选取香港恒生指数期货操纵时段和对应的非操纵时段,进行了预警模型的实证检验,证明了该模型可以对股指期货操纵事件起到较好的预警作用.研究发现,在股指期货市场中,操纵期与非操纵期的交易量和空盘量显著不同,而反映市场波动性的收益率指标的变化并不显著,对操纵行为是否发生可以使用空盘量和交易量构建模型进行判别.Logistic模型可以很好的识别股指期货市场上的操纵行为,在进入操纵期以后,根据模型计算出的操纵事件的发生概率发生了显著的变化.
[Abstract]:By establishing Logistic regression model to judge the possibility of stock index futures being manipulated, The problem of whether the stock index futures are manipulated or not is transformed into calculating the probability of the stock index futures being manipulated in a certain time according to the volatility of the stock index futures market and the liquidity index in a certain period, and the early warning model of the manipulation events of the stock index futures is established. This paper selects Hong Kong Hang Seng Index futures manipulation period and corresponding non-manipulation period to carry out an empirical test of the early warning model, which proves that the model can play a better role in early warning of stock index futures manipulation events. It is found that in the stock index futures market, the trading volume and the short market volume of the manipulation period and the non-manipulation period are significantly different, but the change of the return index reflecting the market volatility is not significant. Logistic model can be used to identify the manipulation behavior in the stock index futures market. After entering the manipulation period, the logistic model can be used to judge whether the manipulation behavior occurs or not, and the logistic model can be used to identify the manipulation behavior in the stock index futures market. The probability of manipulation events calculated according to the model has changed significantly.
【作者单位】：天津大学管理与经济学部;银行业监督管理委员会山东监管局;
【基金】：教育部新世纪优秀人才支持计划项目(NCET-07-0605) 国家自然科学基金(70971096,70801043) 中国期货行业协会联合研究计划(GT200702) 天津社会科学基金(TJ05-TJ003)
【分类号】：F832.51

【参考文献】