空间滞后面板平滑转换模型的估计及数值模拟

摘要

将空间滞后项引入面板平滑转换模型，构建了空间滞后面板平滑转换模型，通过综合应用拟极大似然法和非线性最小二乘法，构造了该模型的参数估计方法，并通过蒙特卡洛数值模拟探讨了参数估计方法的小样本性质；数值模拟结果显示，提出的估计方法在小样本条件下表现良好，参数估计值随着样本容量的增大而收敛到参数的真值。%This paper introduced the spatial correlation into panel smooth transition model , constructed the spatial lag panel smooth transition model ,designed estimation method of model parameter by applying the quasi‐maximum likelihood method and the nonlinear least square method synthetically ,and explored the small sample properties of parameter estimation method by Monte Carlo numerical simulation .Simulation results showed that the proposed estimation method performed well in small samples .T he parameter estimation value converges to the true value as the sample size increases .