检验n和T都很大的固定效应动态面板模型的条件异方差性

研究了动态面板数据模型的条件异方差性检验问题.对于n和T都很大的固定效应动态面板数据模型,通过残差的一阶差分的平方序列,建立一个人工自回归模型,并基于该人工自回归模型系数的最小二乘估计构造检验统计量,检验误差序列的条件异方差性.研究表明在一定的假设条件下,得到的检验渐近服从卡方分布,计算简单方便,通过一些模拟试验研究了检验的小样本性质.模拟研究表明该检验表现很好.

Testing for Conditional Heteroscedasticity of Dynamic Panel Models with Fixed Effects when both n and T are Large

In this paper, some tests are investigated for conditional heteroscedasticity of dynamic panel data models. For the dynamic panel data models with fixed effects when both 71 and T are large, by first-order difference of the residuals, some artificial autoregressive models are considered and some tests are constructed on the basis of the ordinary least squares (OLS) estimators of the coefficients of the artificial autoregressive models. The results show that the tests are asymptotically distributed as χ~2 and easy to implement. The small sample properties of the tests are investigated by means of Monte Carlo experiments. The evidence shows that the tests perform well.