我国财险企业欺诈类操作风险度量研究

发布时间：2018-03-08 01:05

  本文选题：欺诈类操作风险　切入点：PSD-LDA　出处：《保险研究》2017年02期 　论文类型：期刊论文

【摘要】：我国财险企业的操作风险管理起步较晚,目前尚未建立损失事件库,损失数据的缺失导致量化分析的研究较少。通过搜集整理并分析近15年来财险企业的欺诈类操作风险损失事件,发现数据具有"高频低损"和"低频高损"的特征,因此采用两阶段损失分布法(PSD-LDA)进行拟合。考虑到操作风险事件的属性以及数据搜集的不完整性,运用复合Poisson-Geometric分布来拟合损失频率从而度量欺诈类操作风险损失,并计提了相应的经济资本来抵御非预期的操作风险损失。结果表明,相对于设定损失频率为齐次泊松分布,基于复合PG分布的PSD-LDA模型能更好的拟合财险企业欺诈类操作风险损失,为我国财险企业操作风险的度量和管理提供了新思路。
[Abstract]:The operational risk management of property insurance enterprises in our country started relatively late. At present, a loss event bank has not yet been established. The lack of loss data leads to less research on quantitative analysis. By collecting and analyzing the fraudulent operational risk loss events of property insurance enterprises in the past 15 years, it is found that the data have the characteristics of "high frequency and low loss" and "low frequency and high loss". Therefore, the two-stage loss distribution method (PSD-LDA) is used to fit the data. Considering the attribute of operational risk events and the incompleteness of data collection, the composite Poisson-Geometric distribution is used to fit the loss frequency to measure the operational risk loss of fraud class. The corresponding economic capital is taken to resist the unexpected operational risk loss. The results show that the loss frequency is homogeneous Poisson distribution relative to the set loss frequency. The PSD-LDA model based on the compound PG distribution can better fit the operational risk loss of the property insurance enterprises, which provides a new idea for the measurement and management of the operational risk of the property insurance enterprises in China.
【作者单位】： 湖南大学金融与统计学院;
【基金】：中国人保财险灾害基金资助
【分类号】：F842.61;D924.35
，

本文编号：1581722