基于马尔可夫机制转换动态因子模型对我国经济周期拐点的识别

林秀梅 李青召 历姿彤

摘要：基于1993年1月至2017年6月我国工业增加值、货币供应量（M1）、社会消费品零售总额、全社会固定资产投资等宏观经济指标，本文利用两步估计法的马尔可夫机制转换动态因子模型（MS-DFM）对我国经济周期进行了测度，并对经济周期的拐点进行了识别。结果显示，经调整后的月度动态因子与季度GDP高度相关，其可以作为月度GDP的替代变量。MS-DFM能够很好地识别我国经济周期的拐点。相较于仅仅利用季度GDP构建MS-AR模型，MS-DFM对经济周期和经济周期拐点的刻画更加灵敏。研究发现，目前我国经济依然处于紧缩期，而供给侧结构性改革将成为经济保持稳定增长的关键所在。

关键词：经济周期 拐点MS-DFM

中图分类号：F064.1 文献标识码：A

Identifying the Turning Points of China's Business Cycles Based on Markov-Switching Dynamic Factor Model

Lin Xiumei Li Qingzhao Li Zitong

Abstract: Using the industrial added value, money supply（M1）, total retail sales of consumer goods, total fixed asset investment from January 1993 to June 2017, this paper based on the two-step method of Markov-Switching Dynamic Factor Model（MS-DFM）measure the business cycle and identity the business cycle turning points of China. The results show that the adjusted monthly dynamic factors are highly correlated with the quarterly GDP, which can be used as a replacement variable for the monthly GDP. MS-DFM can identify the turning point of China's economic business cycle very well. Compared to the MS-AR model which only use the quarterly GDP, MS-DFM is more sensitive to describe the business cycle and the it's turning points. It is found that China's economy is still in a tight period, and the structural reform of the supply side will be the key to maintain steady economic growth.

Keywords: Business Cycle Turning Points MS-DFM