基于贝叶斯推断的操作风险度量模型研究  

Study on measuring model of operational risk applying Bayesian inference

作　　者：卢安文[1,2] 任玉珑[1] 唐浩阳[1] 

机构地区：[1]重庆大学经济与工商管理学院,重庆400044 [2]重庆邮电大学经济管理学院,重庆400065

出　　处：《系统工程学报》2009年第3期286-292,349,共8页Journal of Systems Engineering

摘　　要：在度量操作风险时,由于数据不足的问题可能造成监管资本配置的偏差,进而导致防范和控制风险的能力降低.本文在对操作风险度量模型分析的基础上,应用贝叶斯推断来度量操作风险.结合实际数据,对操作风险的损失频率和损失金额的分布函数进行了估计,并利用蒙特卡罗模拟方法对操作风险损失值进行了模拟.实证分析表明:应用贝叶斯推断来度量操作风险,可以较好地解决目前操作风险损失事件数据不足的问题.The key problem is scarce of data about operational risk event in measuring operational risk, and it will lead to deflection of regulating capital collocation and low capacity of precautioning and controlling risk. This paper measures the operational risk applying Bayesian inference based on analysis of measuring model of operational risk. The distribution of loss frequency and loss amount of operation loss event were estimated by using actual data. The loss amount of operational risk was simulated by means of Monte Carlo simulation. By the empirical analysis the result shows that the problem of scarce of data about operational risk event can be solved well by applying Bayesian inference to measure the operational risk.

关 键 词：操作风险 贝叶斯推断 蒙特卡罗模拟 

分 类 号：F830[经济管理—金融学]