摘要
选取同行拆借IBO007利率数据,采用拟似然函数估计构建的CKLS类四个模型,研究结果表明:含有跳和异方差的CKLSGJ模型是此类模型中最佳的短期利率模型,模型能恰当的解释我国短期利率均值回复性、水平效应及跳跃行为.蒙特卡洛模拟结果表明:模型能较好的拟合数据,有很强的预测能力.
In this paper we choose the IBO007 as the interest rate data, after putting forward four CKLS models we estimate them by employing quasi maximum likelihood method. As a result we find that the CKLSGJ model which consists of CKLS, jump and GARCH term is more fitable. This model can properly account for the mean reversion, the level effect and the jump behavior in the process of interest rate movement. In the end by using Monte Carlo simulation, this study further demonstrates that the CKLSGJ model be some commendable in fitting the interest rate data, meanwhile having more powers to forecast it.
出处
《数学的实践与认识》
CSCD
北大核心
2013年第9期28-36,共9页
Mathematics in Practice and Theory
基金
教育部人文社科一般研究项目(09YJA790028)
辽宁省教育厅创新团队项目(2008T054)
科技部创新方法(2009IM010400)
国家自然科学基金(71273044)
关键词
异方差
跳过程
拟似然估计
CKLS模型
heteroscedasticity
jump process
quasi likelihood function
CKLS model