摘要
本文将半参数GARCH-M模型推广到多元情形,研究了一类多元部分线性GARCH-M模型。基于截面似然的方法,文章给出了模型参数和未知函数的估计。模拟结果显示,估计的表现良好。基于上证综合指数和香港恒生指数收益数据的实证研究结果也说明,模型在预测方面有一定的优越性,这表示我们考虑的模型在实际中有潜在的应用价值。
A multivariate extension of the semiparametric GARCH-M model is proposed. Based on the profile likelihood approach, estimators for both parameters and unknown functions are given. Simulation results show that our estimators perform well. When applied to the daily Shanghai Composite Index and Hang Seng Index returns, the proposed model shows certain superiority in terms of forecasting, which implies the model has potential applications in practice.
作者
朱华锋
李元
张兴发
ZHU Hua-feng;LI Yuan;ZHANG Xing-fa(School of Economics and Statistics,Guangzhou University,Guangdong Guangzhou 510006,China;Center for Statistical Science of Lingnan,Guangzhou University,Guangdong Cuangzhou 510006,China)
出处
《数理统计与管理》
CSSCI
北大核心
2018年第5期843-849,共7页
Journal of Applied Statistics and Management
基金
国家自然科学基金(11401123,11571148)
广东省普通高校创新团队建设项目(2015KCXTD014)
