摘要
鉴于商业银行的操作风险管理具有样本量小、结构复杂等特点,较难运用传统方法构建风险预警系统,本文采用贝叶斯网络,研究了商业银行操作风险预警系统的建模过程并根据我国商业银行操作风险的外部数据给出了算例。通过构建由关键风险指标和关键风险诱因组成的商业银行操作风险拓扑结构,本文分析了各类风险指标对操作风险的作用形式,在对各类节点赋值的基础上,通过贝叶斯推理,建立起商业银行操作风险的预警系统,以便在出现可能导致巨额损失时,商业银行能够及时采取措施化解操作风险。
Because operational risk of commercial banking has special characteristics with few samples and complex structure, it is difficult to build up pre-warning system by means of conventional financial risk pre-warning approach. With the Bayesian networks, by constructing operational risk's topology networks being consist of KRIs and KRDs, this paper analyzes many indicators' action forms of operational risk. After evaluating every node, we calculate these indicators' influence on operational risk, and building up the pre-warning system of operational risk. Thus, once discovering some operational risk which may result in huge losses to commercial bank, commercial bank may immediately take action to alleviate risk.
出处
《管理工程学报》
CSSCI
2008年第4期56-61,共6页
Journal of Industrial Engineering and Engineering Management
基金
全国统计科学研究计划资助项目(2006C04)
关键词
贝叶斯网络
操作风险
风险管理
bayesian networks
operational risk
risk management