提高PMI预判工业走势效果的探索与思考  

Assessing the Forecasting Power of PMI

作　　者：乔宝华[1] 徐光瑞[1] 

机构地区：[1]中国电子信息产业发展研究院工业经济研究所,北京100846

出　　处：《工业经济论坛》2015年第1期48-55,共8页Industrial Economy Review

摘　　要：本文首先剖析了我国PMI指标体系与发达国家的异同,然后采用相关性分析、时差相关分析和转折点分析等方法深入研究了PMI与工业经济走势的关系,结果发现我国制造业PMI与工业增速中度相关;与工业运行更趋同步,领先性相对较弱;用PMI转折点来预测工业经济走势转折点的可靠性并不稳定。为提高对工业走势的预判效果,应加大PMI调查的样本量,积极开展预测指标的探索,准确理解各项指标的内涵;在现有条件下应更多关注新订单和生产等分类指数的表现。This paper analyzed the details of China PMI index and compared China official PMI with developed countries; and then investigated the relation of PMI with the trend of industrial economy by correlation analysis, time difference correlation analysis and turning points analysis. It is shown that China＇s manufacturing PMI is moderately correlated with industrial growth. The relation between PMI and the trend of industrial economy is more coincidental than leading. It is not reliable enough to forecast the turning points of industrial economy by PMI turning points. We can improve the prognosis for the trend of industrial economy by increasing the sample size of PMI survey, finding better forecasting indicators, and catching the real meaning of each index. And in the current situation, we should pay more attention on the performance of PMI sub-index, such as new orders index and production index.

关 键 词：PMI 工业经济 先行性 预测 

分 类 号：F424[经济管理—产业经济]