基于Copula-GJR-Skewt模型的投资组合风险预测研究 |
| |
引用本文: | 陈玲俐. 基于Copula-GJR-Skewt模型的投资组合风险预测研究[J]. 数学的实践与认识, 2014, 0(18) |
| |
作者姓名: | 陈玲俐 |
| |
作者单位: | 成都理工大学商学院; |
| |
基金项目: | 国家自然科学基金(71171025);国家社会科学基金(12BGL024);教育部人文社科青年基金(10YJCZH086) |
| |
摘 要: | 针对多元投资组合的风险预测,采用GJR-Skewt模型刻画单资产的厚尾、有偏特征,以及Copula模型刻画多元投资组合的非线性相关结构,用Monte Carlo方法模拟金融资产的随机分布,并结合滚动时间窗法,对投资组合的未来风险进行样本外动态预测.实证结果表明,Copula-GJR-Skewt模型对资产收益的风险预测能取得满意的效果;在VaR预测性能上,以GJR-Skewt模型作为边缘分布函数时,即使存在系统偏差,也能取得最优预测结果;预设残差服从有偏学生分布时,VaR的预测结果优于正态分布;传统的Garch-Guassian模型预测能力最差.
|
关 键 词: | Copula GJR-Skewt Monte Carlo模拟 风险预测 |
A Study on Portfolio Risk Predicting Based on Copula-GJR-Skewt Model |
| |
Abstract: | This paper aimed at portfolio value at risk forecasting,to effectively capture the characteristics of market risk factors,like leptokurtic fat tailed and skewed distribution,as well as leverage effect,a GJR-Skewt model is used to depict these stylized facts,and a Copula model to depict the nonlinear portfolio correlation.Numerical results of slide time windows method and Monte Carlo simulation method show that use G JR-skewt as marginal distribution performs best of dynamic out-of-sample VaR predicting,even system bias exists.Besides,risk factors model under skewed student distribution hypothesis perform better than norm hypothesis,while Garch-norm model performs the worst. |
| |
Keywords: | Copula GJR-Skewt Monte Carlo Simulation VaR forcasting |
本文献已被 CNKI 等数据库收录! |
|