摘要
对于商业银行而言,操作风险已经成为与市场风险和信用风险同样重要的风险。本文利用极值理论超越样本的估计能力,采用极值理论中对数据要求量较少,可以进行单步预测的超阈值(POT)方法对我国商业银行操作损失极端值分布进行估计,以均值超额函数图和拟合直线的交点确定阈值,估计出给定置信水平之下操作风险损失的分位数,从而使得国内商业银行操作风险监管资本的计算成为可能。
During the recent years, the operational risk has become an important research field in banks risk management. It has been recognized as the major risks for commercial banks together with the market risk and the credit risk. By using peak over threshold method (POT), which requires less data than other methods, and based on the extreme value theory method, this article estimates the extreme loss distribution and the p-percentile operational loss under certain belief range for Chinese commercial banks, then calculates the operational risk supervise capital. The results are encouraging and the proposed method may be a useful tool for estimating the operational risk.
出处
《运筹与管理》
CSCD
2007年第1期112-117,共6页
Operations Research and Management Science
基金
国家自然科学基金资助项目(70531040)
中国科学院基金资助项目(yjjz946)
科技政策与管理科学研究所所长基金(0600281501)
关键词
金融学
POT方法
极值理论
操作风险
均值超额函数图
finance
peak over threshold method
extreme value theory
operational risk
mean excessfunction plot