摘要
为了有效预测和监控我国价格数据,文章以CPI和PPI价格序列为例,利用商品价格大数据构建型(MIDAS)。结果显示,大数据下的混频MIDAS模型对CPI和PPI的动态预测效果优于传统的ADL和GARCH模型,证明大数据下的混频MIDAS模型对我国价格数据具有较好的监测效果。
In order to effectively predict and monitor China’s price data,this paper takes CPI and PPI price series as an example,and uses commodity price big data to construct high-frequency monitoring indexes of CPI and PPI,collects the high-frequency commodity big data from January 1,2009 to December 27,2019 to construct the MIDAS model.The results show that the mixed frequency MIDAS model based on big data has a better dynamic prediction effect on CPI and PPI than the traditional ADL and GARCH models,which proves that the mixed frequency MIDAS model based on big data has a better monitoring effect on China’s price data.
作者
纪尧
Ji Yao(School of Economics,Peking University,Beijing 100871,China)
出处
《统计与决策》
CSSCI
北大核心
2021年第7期36-39,共4页
Statistics & Decision
基金
北京市统计局科研项目
