基于关系聚类的截面相关PVAR模型研究

Heterogeneity Panel Data Vector Autoregression Model under Cross-Sectional Based on Structural Relationships of Clustering

面板数据向量自回归模型(PVAR)研究中,相关性问题是热点研究问题。PVAR的相关性源于两个方面,一方面,模型设定中,变量受自身动态过程影响,变量间存在内生关系,另一方面指截面之间存在空间相关性。由于内生关系与截面相关性导致残差项之间存在相关性。本文研究存在截面相关性的PVAR模型,检测残差相关性,将数据从残差项相关性上进行分类,类内有相同或者相似的残差相关关系,研究每一个类内存在截面相关情形的模型估计,研究模型总体的参数估计以及格兰杰因果检验,本文提出的估计方法更有效,蒙特卡罗模拟结果显示,本文提出的估计方法有更好的拟合效果。

In the research of panel data vector autoregressive model,correlation is one of the hot topics.The correlation of PVAR stems from two aspects.On the one hand,in the model setting,variables are affected by their own dynamic process,and the relationship between variables is endogenous,on the other hand,it refers to the spatial correlation between sections,which are manifested as the correlation between residuals.In this paper,we study the PVAR model with cross-section correlation,and classify the data from the residual term correlation.the same or similar residual relation data are divided into the same classification.Based on Clustering,the parameter estimation,granger causality test are considered with cross-sectional dependence.Monte Carlo simulation shows that the estimation method proposed in this paper has a good fitting effect.