风险价值(value at risk,VaR)是国际金融界广泛支持和认可的一种度量金融风险的工具。分别利用Bootstrap、MOVER(method of variance estimates recovery)和Fiducial方法给出正态总体下VaR的区间估计方法,并进行了模拟比较。模拟结果发现基于Fiducial思想的广义区间估计在覆盖率和区间等尾性上具有更稳健的性质。最后对上证180重点指数的对数收益率VaR进行了分析。
【英文摘要】
Value at risk( VaR) is a tool used for measuring financial risk,which is widely supported and recognized by international finance. To study the interval estimation method in normal population,we used three methods such as Bootstrap、MOVER( method of variance estimates recovery) and Fiducial,and numerically compare the performance of them. The simulation results showthat generalized interval estimation based on Ficucial is more robust in coverage and balanced tail error. In the end,the VaR of logarithmic yiel...
网络出版地址:http://www.cnki.net/kcms/detail/37.1389.N.20161012.0954.002.html0引言在金融风险管理中,风险价值VaR(value at risk)是估计和管理资产组合风险的一个重要方法[1],它是指在市场的正常波动下,给定置信水平下资产投资组合在未来特定一段时间内最大可能的损失[2]