摘要
基于2008—2019年中部某省卷烟销量数据,本文运用混频MIDAS模型,对该省卷烟销量进行样本内和样本外预测,探索混频模型对卷烟销量预测的适用性。研究发现:饮料烟酒类商品零售价格指数、全社会消费品零售总额、工业增加值增长率等宏观经济变量是预测卷烟销量的有效先行指标。同时,混频MIDAS模型在卷烟销量的样本内和样本外预测都具有较好的精度,预测的准确性和时效性强。因此,可以利用月度宏观经济变量对季度卷烟销量进行预测,为卷烟工商企业制定相关销售政策提供实证依据。
Based on the cigarette sales data of a central province in 2008-2019,this article uses the mixing MIDAS model to make insample and out-of-sample predictions of cigarette sales in the province,and explores the applicability of the mixing model in the prediction of cigarette sales.The study results show that retail price index of beverage,tobacco and alcohol products,total retail sales of consumer goods,growth rate of industrial added value,and other macroeconomic variables are effective leading indicators for predicting cigarette sales.At the same time,the mixing MIDAS model has good accuracy for both in-sample and out-of-sample predictions of cigarette sales,and shows strong timeliness.Therefore,monthly macroeconomic variables can be used to predict quarterly cigarette sales,providing an empirical basis for cigarette business enterprises to formulate relevant sales policies.
作者
石涛
李根栓
李鹏飞
郑飘
SHI Tao;LI Genshuan;LI Pengfei;ZHENG Piao(China Tobacco Henan Industrial CO.,LTD,Zhengzhou 450016,China;Economy institution,Henan Academy of Social Science,Zhengzhou 450002,China)
出处
《中国烟草学报》
CAS
CSCD
北大核心
2021年第4期92-99,共8页
Acta Tabacaria Sinica
作者简介
石涛(1986-),博士,助理研究员,主要研究方向:烟草经济管理,Tel:0371-69192920,Email:taoshi_ccp@163.com;通讯作者:郑飘(1990-),硕士,经济师,统计师,主要研究方向:烟草经济研究,Te:0371-69192652,Email:zhengp0311@163.com。
