基于马尔科夫区制转换模型的中国经济周期识别  

Identification of China’s Business Cycle Based on Markov Switching Model

作　　者：虞栋杰 郑静 Yu Dongjie;Zheng Jing(College of Economics,Hangzhou Dianzi University,Hangzhou 310018,China)

机构地区：[1]杭州电子科技大学经济学院,杭州310018

出　　处：《统计与决策》2022年第8期38-42,共5页Statistics & Decision

基　　金：国家社会科学基金资助项目(21BTJ071)。

摘　　要：经济周期研究既能了解经济运行的真实状况,也是经济危机监测和国家政策调控的基础。文章采用一种能够综合利用高频数据和低频数据的经济周期计量模型,即马尔科夫区制转换混频数据回归(MS-MIDAS)模型,利用月度货运量数据以及季度GDP数据对我国经济周期进行识别与预测,并提供了一种新的方法估计模型参数。结果表明:中国经济周期波动具有非对称性;MS-MIDAS模型在识别与预测经济周期方面具有准确性和简便性。同时,实证结果证实了巴菲特的观点:货运量是经济状况的晴雨表。Research on business cycle can not only understand the real state of economic operation, but also be the basis of economic crisis monitoring and national policy regulation. This paper adopts an econometric model of business cycle which can use both high frequency and low frequency data synthetically, namely, Markov switching mixed frequency data regression(MS-MIDAS) model, then uses monthly freight volume data and quarterly GDP data to identify and forecast China’s business cycle, and finally provides a new method to estimate model parameters. The results show that the fluctuation of China’s business cycle is asymmetrical, and that MS-MIDAS model is accurate and simple in identifying and predicting business cycles. At the same time, empirical results confirm Buffett’s view that freight volumes are a barometer of economic health.

关 键 词：经济周期 货运量 马尔科夫区制转换模型 混频数据回归 

分 类 号：F123.5[经济管理—世界经济] F503