Markov机制转换模型研究——在中国宏观经济周期分析中的应用

Research on the Markov Switching Model

本文首次引入反映我国经济增长周期模式改变和状态转移机制变迁的虚拟变量,对传统Markov机制转换模型进行了修正,由此解决了将Markov模型应用于中国年度宏观经济数据研究中国经济周期问题的难题。运用修正后的Mark-ov模型,本文对我国1953～2005年的年度实际产出增长率的数据进行了拟合,研究表明,该模型较好地刻画了我国实际产出增长的周期性变化。根据分析我们发现,改革前后我国经济周期的非对称性特征比较明显,并且经济增长周期模式和经济周期性变化机制存在显著差异。

For the first time, this paper take a dummy variable into the traditional Markov Switching Model to depict the change of Chinese economic cycle pattern and Regime-Switching mechanism. We resolve the problem that how to study Chinese business cycles with the Markov Switching model based on annual macroeconomic data. Fitting the data of Chinese real GDP growth from 1953 to 2005 with our model, we find that the model perfectly describes Chinese real GDP growth＇s periodical movement. Chinese Business cycle pattern has changed after the Chinese Economic Reform. The Regime-Switching mechanism also has changed after the Chinese Economic Reform. Asymmetry of the Chinese economic cycle is remarkable. Before the Chinese Economic Reform , the expansion period is longer than contraction period but it is reversed after the Chinese Economic Reform.