基于logistic模型的证券公司客户流失预警分析

Customer Churn Warning Analysis on Securities Companies Based on Logistic Model

伴随着中国经济的高速发展和经济全球化的不断加深,客户流失问题比争夺客户更需要证券公司的高度关注。文章从反映客户交易情况的指标出发,采用K-均值聚类获取客户流失状态;接着通过6种逐步回归方法进行变量筛选,并建立logistic客户流失预警模型;再对模型的泛化能力进行检验并基于证券公司的业务特点给出分析。研究结果表明:反映客户交易活跃度的指标是证券公司实施客户流失预警的关键,进而为证券公司有针对性地挽留客户提供有效的方法和可行的建议。

With the rapid development of China＇ s economy and the deepening of economic globalization, customer chum has become more important than grabbing customers for securities companies. Starting from the index reflecting the customer transactions, K -means cluster is used for obtaining customer chum state. Through 6 kinds of stepwise regression method to variable selection, a logistic customer churn warning model is set up in this paper. Moreover, the generalization ability of the model is tested and analysis based on business characteristics of securities companies＇are given. The results show that ：the customers＇trading activity index is the key to the implementaion of customer chum warning of the securities companies. Furthermore, an effective model and feasible suggestion are put forward to targeted customer retention for the securities companies.