论文部分内容阅读
本文结合多机制平滑转换回归模型和半参数平滑转换回归模型,提出多机制半参数平滑转换回归模型。对模型转换函数中的未知光滑有界函数采用级数估计,并给出了结合Back-fitting算法和非线性最小二乘法估计模型参数的具体执行步骤,随机模拟结果说明了本文模型和估计算法的可行性和灵活性。应用本文模型和估计算法对我国宏观经济运行周期的实证研究表明,我国经济增长的非线性结构可以分为四个显著不同的增长机制:扩张阶段、衰退阶段、收缩阶段、恢复阶段,并且宏观经济政策的作用有三到四个季度的迟滞效应。
Combining the multi-mechanism smooth transition regression model and the semi-parametric smooth transition regression model, a multi-mechanism semi-parametric smooth transition regression model is proposed in this paper. The estimation of the number of series is given to the unknown smooth bounded function in the model transfer function. The steps of estimating the model parameters by Back-fitting algorithm and nonlinear least square method are given. The results of stochastic simulation show that the model and the estimation algorithm Feasibility and flexibility. The empirical study on the macroeconomic operation cycle of China using this model and the estimation algorithm shows that the nonlinear structure of China’s economic growth can be divided into four distinct growth mechanisms: the expansion phase, the recession phase, the contraction phase and the recovery phase, and the macroeconomic The role of policy has three to four quarters of hysteresis.