【摘要】 操作风险因其厚尾性使其测度更具复杂性,极值理论能够解决此一问题。但在银行内部数据不充分的情况下,只有少数点进入尾部,尾部数据不够大,必然会影响参数的确定,继而影响测度精度。引入信度理论,将外部数据构建的操作风险损失尾部与内部数据建模进行整合,通过估算合适的信度因子,利用外部数据弥合内部数据不充分引致的测度误差,以期获得更精确的测度值。更多还原

【Abstract】 Owing to the thick tail,the operational risk measurement is more complicated,which can be settled by using the extreme value theory.However,there are a few dots in the tail area and the tail data is not enough when the bank inside data is less sufficient,which will affect the determination of the parameter.Then we introduce the credibility theory to integrate the operational risk loss tail built on the outside data with the model based on the inside data and estimate the proper credibility factor,which can bridge over for the measurement error caused by the insufficiency of inside data,hoping to achieve a more accurate measurement.更多还原