摘要
论文引入Markov切换模型探讨中国PMI指数波动的持续性、波动幅度以及波动频率等特征。对此,首先通过ADF方法和JB统计量检验PMI时间序列的平稳性和非对称性,随后通过CUSUM方法检验得到时间序列不存在结构变点特征,采取极大似然估计方法确定Markov切换模型的相关参数。结果得到,PMI时间序列的低波动性持续的时间最长,进而得到波动形态的切换主要集中于中等波动和高等波动之间;并且PMI时间序列的波动存在明显的"聚集波动"和"分段波动"现象。最后,加入其他经济体PMI指数进行对比分析得到,中国PMI指数除了受到自身影响外,还会受到其他经济体的PMI指数的影响,而美国PMI指数对中国PMI指数的影响最为显著。
This paper has introduced Markov Switching Model to study the fluctuation persistence,fluctuation extent and fluctuation frequency of Chinese PMI index.Regarding this,the paper firstly analyzes stability and asymmetry of PMI series by using JB Statistics and ADF test,and then analyzes the mechanism feature by CUSUM test method,and finally takes the maximum likelihood estimation method to determine the parameters of Markov switching model.The result shows that,the low fluctuation of PMI time series lasts the longest time,and the state switches between medium fluctuation and high fluctuation;further,the fluctuation of PMI series is stable at low state,and there is obvious"clustering fluctuation"and"stage fluctuation".Finally,we compare and analyze the PMI index of other countries with the PMI index of China.The result shows that,PMI index of China is affected by the PMI index of other countries,but the impact of America's PMI index on China's PMI index is the most significants.
作者
王路加
郭亚妮
WUANG Lu-jia;GUO Ya-ni(Xingjiang University of Finance and Economics,Urumqi 830012;Xingjiang Branch,National Development Bank,Urumqi 830001)
出处
《财务与金融》
2018年第2期88-94,共7页
Accounting and Finance
基金
2017年自治区研究生教育创新计划科研创新项目(项目编号:XJGRI2017115)
新疆财经大学研究生科研创新项目(项目编号:XJUFE2017D011)
