结构突变趋势平稳过程与随机趋势过程的虚假回归研究

统计研究 ›› 2011, Vol. 28 ›› Issue (5): 105-110.

结构突变趋势平稳过程与随机趋势过程的虚假回归研究

张凌翔张晓峒

出版日期:2011-05-15 发布日期:2011-05-20

Spurious Regression between Broken-Trend Stationary Process and Stochastic Trend Process

Zhang Lingxiang Zhang Xiaotong

Online:2011-05-15 Published:2011-05-20

摘要/Abstract

摘要： 内容提要：在已有研究的基础上，本文更为深入的研究含有结构突变的趋势平稳变量与随机趋势变量间的虚假回归问题。本文推导出OLS估计下DW统计量、F统计量以及R2的极限分布，并且将回归模型扩展到动态情形下，推导出用于Granger因果检验的F统计量的极限分布；采用Monte Carlo模拟方法分析了数据生成过程的各项参数对各统计量有限样本分布的影响；最后，本文分析了在有限样本下，数据生成过程的各项参数对虚假回归及虚假Granger因果关系发生概率的影响。

关键词: 关键词：结构突变, 随机趋势, 虚假回归

Abstract: Abstract: We study the phenomenon of spurious regression between two random variables, when the generating mechanism of one of individual series is assumed to follow a stationary process around a trend with multiple breaks in the level and slope of trend and the other is stochastic trend. We develop the relevant limit theory of the DW statistics, F statistics and R2 of the OLS regression and the F statistics of Granger Causality tests. Furthermore, we analyze the distributions of these statistics in finite samples and the impact of parameters of the generating mechanism on these finite sampling distributions via Monte Carlo simulations. Finally, in finite samples, we reveal the effects of the parameters of the generating mechanism on spurious regression and spurious Granger Causality relationship.

Key words: Key words: Structural Breaks, Stochastic Trends, Spurious Regression

张凌翔张晓峒. 结构突变趋势平稳过程与随机趋势过程的虚假回归研究[J]. 统计研究, 2011, 28(5): 105-110.

Zhang Lingxiang Zhang Xiaotong . Spurious Regression between Broken-Trend Stationary Process and Stochastic Trend Process[J]. , 2011, 28(5): 105-110.

[1]	王晓军赵明. 中国高龄人口死亡率随机波动趋势分析[J]. 统计研究, 2014, 31(9): 51-57.
[2]	项后军潘锡泉. 汇率变动、货币政策与通货膨胀[J]. 统计研究, 2011, 28(5): 3-12.
[3]	张卫平等. 制度变迁背景下人民币实际汇率趋势研究：1986—2009[J]. 统计研究, 2011, 28(1): 36-42.
[4]	白仲林. 同期相关面板数据结构突变单位根检验的统计性质 ——中国CPI指数平稳性的经验证据* [J]. 统计研究, 2008, 25(10): 86-91.