保险资产管理公司违约风险预警模型的构建与应用  

Construction and Application of Default Risk Warning Model for Insurance Asset Management Companies

作　　者：刘晓瑜 杨晓东 LIU Xiaoyu;YANG Xiaodong

机构地区：[1]国寿投资保险资产管理有限公司,北京100020 [2]中国人寿保险(集团)公司,北京100033

出　　处：《保险职业学院学报》2024年第1期19-27,共9页Journal of Insurance Professional College

摘　　要：近年来市场上信用风险事件不断爆发,保险资产管理公司对信用风险的监控和预警的需求也在逐步增加。然而保险资产管理公司内部在信用风险模型开发上面临数据不足的先天劣势。本文使用二级市场债券交易的公开数据进行拟合,解决了数据来源不足的问题,创新地提出以“违约风险预警”概念置换“实质违约”概念,解决了债券市场上实质违约债券不足、违约率过低、不能真实反映市场信用风险状况等问题,在投前尽调、投后管理中都有较好的实际应用作用。本文主要使用逻辑回归模型对大样本数据进行拟合,拟合结果较好,在统计上有较强的解释意义。In recent years,credit risk events have been constantly erupting in the market,and the demand for credit risk monitoring and early warning by insurance asset management companies is gradually increasing.However,insurance asset management companies face an inherent disadvantage of insufficient data in the development of credit risk models.This article uses public data from bond trading in the secondary market for fitting,solves the problem of insufficient data sources,and innovatively proposes replacing the concept of Substantive Default with the concept of Default Risk Warning to solve the problem of insufficient substantive default bonds,low default rates,and inability to truly reflect the market credit risk situation in the bond market.It has better practical application in pre-investment due diligence and post-investment management.This article mainly uses logistic regression model to fit large sample data,and the fiing results are good,which has strong explanatory significance in statistics.

关 键 词：信用风险 量化模型 逻辑回归 单变量分析 WOE分组 

分 类 号：F840.323[经济管理—保险]