摘要
将Pls-Logit模型引入到寿险公司财务风险预警中,并运用我国57家寿险公司的财务数据进行实证分析。首先利用Pls方法完成因子提取及成分确定,进而将该方法与Logit方法结合,从而建立基于Pls-Logit方法的寿险公司财务风险预警模型,并与使用Lasso方法建立的寿险公司财务风险预警模型相比较。实证分析结果表明Plslogit模型把2种模型相结合解决了logit模型检验和估计不精准的问题并且基于Pls-Logit方法的财务风险预警模型的预警准确率优于Lasso预警模型。
By introducing The Pls-Logit model into the financial risk warning of life insurance companies,and using the financial data of 57 life insurance companies in China for empirical analysis,the paper firstly used the Pls method to complete the factor extraction and the component determination,and then combined the method with the Logit method to establish the financial risk early warning model of the life insurance company based on the Pls-Logit method and compared it with the financial risk early warning model of the life insurance company established by Lasso method. The empirical results show that the Pls-logit model combined with the two models can solve the problem of logit model checking and estimating inaccuracies and the early warning accuracy of the financial risk warning model based on the Pls-Logit method,which is better than the Lasso early warning model.
出处
《技术与创新管理》
2018年第1期102-110,共9页
Technology and Innovation Management
基金
国家自然科学基金项目"基于结构化大数据深度挖掘的非寿险保险公司经营风险模型研究"(61502280)
