养老保险基金风险预警指标体系研究——基于神经网络思想的分析

Study on Early Warning Indicators of Pension Risk——Analysis of Neural Network Ideas

养老保险基金的财政补贴压力日益显现,构建一套有效可靠的预警指标体系可以加强其风险预警。但现行的风险预警指标体系还存在着缺少对我国养老保险体制特殊性的考虑、在管理运营风险中定性指标较少、缺乏动态分析以及忽略了个人账户与统筹账户的差异化筹资形式等问题,因此,必须在了解中国"统账结合"的养老保险特点基础上,进行"统账分离"风险指标提取,运用神经网络思想,探索构建非线性的网状养老保险基金风险预警指标体系,以进一步厘清养老保险基金风险的内在关系,从而有助于搭建更为有效实用的养老保险基金风险监管体系。

Pressure of financial subsidies of basic old-age insurance is becoming more evident. To build an effective and reliable early warning indicator system can enhance its risk warning. But the current risk warning indicator system still lacks consideration of the special nature of our pension system,qualitative indicators in operational risk management and dynamic analysis. Therefore, we must understand the basic characteristics of pension insurance and risk identification in China＇s ＂combined account system ＂to build a risk index system in ＂a separated account system ＂,and use the neural network ideas to explore how to develop the non-linear mesh pension risk warning indicator system,in order to further clarify the relationship between the internal pension fund risks,clear overall objectives and principles of the pension system reform.