指数平滑转换的协整回归模型检验及其实证分析

Testing Linearity in Cointegrating Exponential Smooth Transition Regressions and Its' Empirical Analysis

在解释变量内生和误差序列相关的指数平滑转换协整回归模型中,研究协整关系属于线性协整还是指数平滑协整的Wald检验方法。在用泰勒展开替代指数函数得到的辅助回归模型中,应用超前和滞后方法解决误差序列的相关性和解释变量的内生性,得到Wald统计量的极限分布收敛于标准的χ2分布。模拟计算验证了Wald检验具有良好的有限样本性质。实证分析发现多种期货价格和现货价格之间具有非线性协整关系。

This paper develops Wald tests that can be uesd to test lineartiy in cointegrating exponential smooth transition regression models, which regressors and errors are assumed to be dependent both serially and contemporaneously. Replacing the exponential function by its Taylor approximation, the auxiliary regression model is estimated by ordinary least squares techniques. Based on leads and lags approach, the Wald statistic is proposed and standard chi-square distribution is derived. Simulated experiments indicate that the tests have reasonable finite sample properties. Empirical applications to spot and futures prices illustrate the paratical usefullness of the tests.