摘要
在异方差线性回归模型中,当模型误差项的协方差阵未知时,对异方差模型进行估计目前还没有比较好的方法。基于此,提出一种异方差模型的两阶段估计—基于异方差一致协方差阵估计,该方法将异方差一致协方差阵估计HC5m和广义最小二乘估计法结合起来,综合使用全部样本的信息,并对异方差模型进行估计。通过大量的蒙特卡洛数值模拟和实证分析,结果表明该方法具有一定的可行性和有效性。
In the heteroskedastic linear regression model,when the covariance matrix of the error term of the model is unknown,there are no better methods to estimate the heteroskedastic model.Based on this,a two-stage estimator of the heteroskedastic model is proposed,based on heteroskedastic consistent covariance matrix estimates(HCCME).This method combines the heteroskedastic consistent covariance matrix estimation HC5 m and the generalized least squares(GLE)estimate method,using the entire information of the sample to estimate the heteroskedastic model.According to a large number of Monte Carlo numerical simulations and empirical analysis,the results show that the method has certain feasibility and effectiveness.
引文
[1]Box G E P,Cox D R.An Analysis of Transformation(With Discussion)[J].Journal of the Royal Statistieal Society:Series B,1964(26).
[2]Su L,Zhao Y,Yan T,et al.Local Polynomial Estimation ofHeteroscedasticity in a Multivariate Linear Regression Model and Its Applications in Economics[J].Plos One,2012,7(9).
[3]何其祥,郑明.一元线性模型异方差的局部多项式回归[J].系统管理学报,2003(2).
[4]Zhang X,Hao H,Liang J.A New Nonparametric Estimation Method of the Variance in a Heteroscedastic Model[J].Hacettepe Journal of Mathematics and Statistics,2015,44(1).
[5]张荷观.基于分组的异方差检验和两阶段估计[J].数量经济技术经济研究,2006(1).
[6]李顺勇,钱宇华,张晓琴,等.基于变量选择和聚类分析的两阶段异方差模型估计[J].应用概率统计,2018(2).
[7]White H.A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity[J].Econometrica,1980,48(4).
[8]Hinkley,David V.Jackknifing in Unbalanced Situations[J].Technometrics,1977,19(3).
[9]Horn S D,Horn R A,Duncan D B,et al.Estimating Heteroscedastic Variances in Linear Models[J].Journal of the American Statistical Association,1975,70(350).
[10]Long J S,Ervin L H.Using Heteroskedasticity:Consistent Standard Errors in Linear Regression Model[J].American Statistician,2000,54(3).
[11]Cribari-Neto F.Asymptotic Inference Under Heteroskedasticity of Unknown Form[J].Computational Statistics&Data Analysis,2004,45(2).
[12]Cribari-Neto F,Silva W B D.A New Heteroskedasticity-Consistent Covariance Matrix Estimator for the Linear Regression Model[J].Asta Advances in Statistical Analysis,2011,95(2).
[13]Li S,Zhang N,Zhang X,et al.A New Heteroskedasticity-Consistent Covariance Matrix Estimator and Inference under Heteroskedasticity[J].Journal of Statistical Computation&Simulation,2016,87(1).
[14]格林,张成思.计量经济分析[M].北京:中国人民大学出版社,2011.
(1)此评价指标进行的是横向比较,即在相同的情况下不同方法间的优劣比较。
(2)同上。
(1)此处假设是为了满足使用分组两阶段法时的需要。