论文部分内容阅读
经纪人制度的实施对证券行业未来发展起到了极大的推动作用,但证券经纪人风险则随着经纪人制度的普遍实施而凸显。传统的经纪人风险侦测是根据经验建立相关检测模型,通过设定阀值确定是否存在风险,该方法是一种事后的检测方法,主观性较强且误判和漏判的概率较大。本文结合客户的背景资料及其交易行为建立logistic回归定量分析模型,计算客户被“代客操作”的概率,反向核查其经纪人的违规行为。实证结果表明,该模型能极大提高风险侦测的准度与深度,能较好克服主观性的弊端,有助于发展完善的风险管理体系。
The implementation of the broker system has played a significant role in promoting the future development of the securities industry, but the risks of securities brokers are highlighted with the broadest implementation of the broker system. Traditional broker risk detection is based on experience to establish the relevant detection model, by setting the threshold to determine whether there is a risk, the method is an ex post detection method, the subjectivity is strong and the probability of miscarriage of justice and missed judgment. In this paper, we establish a quantitative analysis model of logistic regression based on the customer’s background information and its trading behavior to calculate the probability that the customer is “valedictorized” and reverse check the agent’s irregular behavior. The empirical results show that the model can greatly improve the accuracy and depth of risk detection, can better overcome the drawbacks of subjectivity and contribute to the development of a sound risk management system.