本文基于分位数回归理论与非线性回归方法,提出一个包容性较强的模型:非线性参数异质Phillips曲线模型,并给出估计、检验与条件密度预测方法。该模型不仅可用于刻画Phillips曲线的非线性与非对称等典型特征,而且还可以揭示在不同经济环境下通货膨胀的完整分布变动规律,从而能够准确掌握通货膨胀的不确定性,便于科学决策。最后,将该模型应用于中国Phillips曲线特征研究,结果显示:该模型在拟合优度、结构分析、预测能力等方面均优于其他Phillips曲线模型。
A new model with strong inclusive, named Nonlinear Heterogeneous Parameters Phillips Curve that is proposed in this paper based on quantile regression theory and nonlinear regression method. Furthermore, the modeling methodologies for estimation, test and conditional density forecast are studied in detail. The new model can be used not only to describe the stylized facts, such as nonlinearity and asymmetry, but also to discover the dynamic feature of the distribution of the inflation under different economic circumstances, through which we will be able to accurately capture the inflation uncertainty. In the end, the new model is applied to the practice in China. The empirical evidence tells us the fact that the new model is superior to other Phillips curve model in terms of goodness of fit, structural analysis and predictive ability.