摘要
在风险管理中,风险量度估计的稳定性对金融机构的经济资本确定及风险分配起着重要的作用.本文从预期短缺(ES_α)估计的稳定性角度分析重要性抽样技术和 Monte Carlo 模拟在估计信用资产组合 ES_α方面的差异.结果表明,由于组合损失分布尾部事件的稀有性,与传统的 Monte Carlo 模拟方法相比,运用重要性抽样方法估计的 ES_α比较稳定,且生成的风险贡献能够明显地体现出资产间不同的风险特征.
In the risk management, the stability of risk-measure estimation plays an important role in determining the economic capital and assigning risk in the financial institution. From the point of view of the stability of the expected shortfall estimation＇, the difference between importance sampling （IS） and Monte Carlo simulation in estimating the expected shortfall of the credit portfolio is analysed in the paper. A numerical example is given to show that, due to the rarity of the tail events in the loss distribution of the credit portfolio, the expected shortfall estimated by using IS is more stable than one by using Monte Carlo simulation. Also the risk contributions generated by IS can clearly show the different risk features among the assets.
出处
《系统工程学报》
CSCD
北大核心
2008年第5期526-531,共6页
Journal of Systems Engineering
基金
国家自然科学基金资助项目(70573076)
高校博士学科点专项科研基金资助项目(20050056057)
关键词
信用风险
预期短缺
重要性抽样
风险贡献
credit risk
expected shortfall
importance sampling
risk contributions
