摘要
笔者利用概率统计以及计量经济学方法对我国许可证申领数据库进行发掘和建模。由于许可证申领数据序列领先于我国商品实际进出口量值序列,因此本文采用概率密度函数、ARMA、回归、VAR、聚类分析等统计及计量方法,从定量角度分析许可证申领数据对实际进出口数据的领先性特征。在此基础上,笔者构建了中国进出口许可证综合领先指数,并利用该指数对中国重点商品进出口状况进行监测预警。我们尝试系统地对我国许可证申领数据库进行数据挖掘、统计计量建模、领先指数构建并进行监测预警,这在国内外均属首次,具有开创性意义。通过本课题的研究发现:许可证数据库具有构建许可证综合领先指数并进行监测预警的数据基础。许可证申领数据对我国对外贸易具有领先性,但不同商品的领先规律不同。建立在单个商品上的领先指数的预警监测功能要优于分类指数及综合指数。
This paper uses the method of probability statistics and econometrics to mining and modeling the database of China＇s import-export license. Since the license data is the leading index of the actual import- export data, this paper uses statistical and econometric method, such as probability density function, ARMA model, regression model, VAR model, cluster analysis, to analyze this leading rule. Based on the leading rule, this paper also construct a comprehensive leading index for a group of products and using the index to monitor and forewarn China＇s actual trade. This research is the first time to systematically mine and analyze China＇s import-export license database. We find that different products have different leading characteristics, but it shows stable for the same product. The more product of the leading index is combined, the less precise the index shows.
出处
《中央财经大学学报》
CSSCI
北大核心
2015年第4期64-72,共9页
Journal of Central University of Finance & Economics
基金
商务部配额许可证事务局与中央财经大学国际经济与贸易学院联合课题"中国进出口许可证综合领先指数及其监测预警研究"之研究成果
由本文作者担任执笔人
感谢所有课题组成员对该课题的贡献
关键词
进出口许可证
领先指数
聚类
分层
监测预警
Import-export license Leading index Cluster Stratification Monitoring early warning
