Granger因果检验的非线性进展及应用研究
发布时间：2018-09-07 16:07
【摘要】：Granger因果检验是计量经济学的重要组成部分,也是现代经济、金融学分析的重要工具,近年来Granger因果检验在非线性检验方向有了较大进展。本文在线性Granger因果检验的基础上,阐述了Granger因果检验的非线性进展,重点总结了针对一阶矩的基于回归模型、非参函数和信息理论的三大类非线性方法以及针对二阶矩的基于残差交叉相关系数和多元条件方差模型下的两大类非线性方法,讨论了不同非线性Granger方法中数据要求、核心模型、建模关键以及模型优缺点,提出了Granger因果检验"线性-非线性"的整体框架和研究范式.通过模型分析和比较,本文可为因果检验的非线性理论和模型研究提供参考,并对因果检验在经济和金融领域的更广泛应用提供支持。
[Abstract]:Granger causality test is an important part of econometrics and an important tool of modern economic and financial analysis. In recent years, Granger causality test has made great progress in the direction of nonlinear testing. In this paper, on the basis of linear Granger causality test, the nonlinear development of Granger causality test is expounded, and the regression model based on first-order moments is summarized. Three kinds of nonlinear methods based on nonparametric function and information theory, and two kinds of nonlinear methods based on residual cross-correlation coefficient and multivariate conditional variance model are discussed. The data requirements of different nonlinear Granger methods are discussed. The core model, the key to modeling and the advantages and disadvantages of the model are presented. The overall framework and research paradigm of the Granger causality test "linear-nonlinear" are proposed. Through model analysis and comparison, this paper can provide a reference for the study of nonlinear theory and model of causality test, and support the wider application of causality test in the field of economy and finance.
【作者单位】： 山东师范大学管理科学与工程学院;中央财经大学管理科学与工程学院;中国科学院数学与系统科学研究院;
【基金】：国家自然科学基金(71203247,71271223) 北京市自然科学基金(9152017) 广西石化资源加工及过程强化技术重点实验室开放课题基金的资助
【分类号】：O212
本文编号：2228744
[Abstract]:Granger causality test is an important part of econometrics and an important tool of modern economic and financial analysis. In recent years, Granger causality test has made great progress in the direction of nonlinear testing. In this paper, on the basis of linear Granger causality test, the nonlinear development of Granger causality test is expounded, and the regression model based on first-order moments is summarized. Three kinds of nonlinear methods based on nonparametric function and information theory, and two kinds of nonlinear methods based on residual cross-correlation coefficient and multivariate conditional variance model are discussed. The data requirements of different nonlinear Granger methods are discussed. The core model, the key to modeling and the advantages and disadvantages of the model are presented. The overall framework and research paradigm of the Granger causality test "linear-nonlinear" are proposed. Through model analysis and comparison, this paper can provide a reference for the study of nonlinear theory and model of causality test, and support the wider application of causality test in the field of economy and finance.
【作者单位】： 山东师范大学管理科学与工程学院;中央财经大学管理科学与工程学院;中国科学院数学与系统科学研究院;
【基金】：国家自然科学基金(71203247,71271223) 北京市自然科学基金(9152017) 广西石化资源加工及过程强化技术重点实验室开放课题基金的资助
【分类号】：O212
【相似文献】
相关期刊论文 前2条
1 黄亚桥,答军;信息理论的拓展——建立指标学的构想[J];统计与决策;1991年03期
2 郝明,范贻昌,陈通;熵─信息理论在多品种小批量企业日常生产作业动态管理中的应用初探[J];辽宁师范大学学报(自然科学版);1996年01期
，本文编号：2228744
本文链接：https://www.wllwen.com/kejilunwen/yysx/2228744.html
