摘要
介绍了非参数方法中的Bootstrap方法在估计样本分位点时的应用,其中在随机抽取子样时应用了MCMC方法,最后将该方法应用到了金融资产VaR的计算上.由于金融资产收益率的分布往往不能用参数方法准确的估计,Bootstrap作为非参数方法克服了这种局限,并改进了历史模拟方法.文章对欧元/人民币、日元/人民币两种汇率进行了VaR的计算的实证分析,计算了VaR的点估计和区间估计,并比较了几种计算方法,得到了一些有意义的结果.
In this paper the nonparametic Bootstrap method is introduced and the VaR of the financial capital is evaluated. Because the financial asset yield is not easy to be estimated by parametic method, as a nonparametic method, Bootstrap method overcomes those defects, and also improves the history simulation method. In this paper VaR of European dollar and Japanese yen exchange rates are calculated, both point estimator and interval estimator are given. By comparing those methods, some significative results are found.
出处
《系统工程学报》
CSCD
2004年第5期528-531,共4页
Journal of Systems Engineering
基金
国家自然科学基金资助项目(10071082)
教育部博士点基金资助项目
中国科学院和中国科技大学创新基金资助项目.