摘要
为了对中国经济波动及其经济增长驱动机制进行研究,先对1993—2016年的GDP增长率进行经验模态分析方法的分解,得出包含所有GDP增长波动特征的4个具有不同频率尺度的本征模态和1个平稳的残差项。然后根据波动周期特征对其进行重新组合,分别对其进行统计识别与经济解释,发现EMD在分解GDP增长率波动信息特征上具有有效性,同时分解重构后不同周期的中国经济波动均带有明显的经济意义。最后分别对各周期波动进行逐步回归分析从而研究其不同周期经济增长的驱动机制,发现中国经济增长的驱动机制主要为区制转换。
In order to study China＇ s economic fluctuation and its driving mechanism of economic growth, this article first decomposes GDP growth rate from 1993 to 2016 through the empirical modal analysis method, and then obtains four eigenmodes with different frequency scales and one stable residual term containing all the characteristics of GDP growth fluctuation. With recombining them according to the characteristics of fluctuation cycle, after statistical identification and economic interpretation respectively, it was found that EMD was effective in decomposing the fluctuation information characteristics of GDP growth rate. Meanwhile, the economic fluctuations of China in different periods after decomposition and reconstruction present obvious economic significance. Finally, the stepwise regression analysis of each cycle fluctuation is carried out to study the driving mechanism of economic growth in different cycles. It is found that the driving mechanism of China＇ s economic growth is mainly regime switching.
作者
陈双莲
李正辉
CHEN Shuang-lian;LI Zheng-hui
出处
《求是学刊》
CSSCI
北大核心
2018年第4期59-67,共9页
Seeking Truth
基金
全国统计科学研究项目(2017LY49)
关键词
EMD
波动特征
逐步回归法
驱动机制
EMD
fluctuation characteristics
stepwise regression method
driving mechanism
