管理评论

›› 2018, Vol. 30 ›› Issue (1): 3-13.

• 经济与金融管理 •    下一篇

基于多元传导模型的物价指数预测新方法——2018年中国物价展望

骆晓强1, 鲍勤2, 魏云捷2, 杨博宇1,2   

  1. 1. 中国科学院大学经济与管理学院, 北京 100190;
    2. 中国科学院数学与系统科学研究院, 北京 100190
  • 收稿日期:2018-01-06 出版日期:2018-01-28 发布日期:2018-01-24
  • 通讯作者: 魏云捷(通讯作者),中国科学院数学与系统科学研究院助理研究员,博士。
  • 作者简介:骆晓强,中国科学院大学经济与管理学院博士研究生;鲍勤,中国科学院数学与系统科学研究院助理研究员,博士;杨博宇,中国科学院数学与系统科学研究院博士研究生。
  • 基金资助:

    国家自然科学基金应急管理项目（71642006）；国家社科基金重大项目（15ZDA011）；中国科学院预测科学研究中心资助。

A Multivariate-transmission-based New Approach for Forecasting China's Price Indexes in 2018

Luo Xiaoqiang1, Bao Qin2, Wei Yunjie2, Yang Boyu1,2   

  1. 1. School of Economics and Management, University of Chinese Academy of Sciences, Beijing 100190;
    2. Academy of Mathematics and Systems Science, Chinese Academy of Sciences, Beijing 100190
  • Received:2018-01-06 Online:2018-01-28 Published:2018-01-24

PDF (PC)

1194

摘要：

精准的物价预测是科学制定宏观经济调控政策的决策前提和依据，如何提高物价预测的准确性是宏观经济政策实践面临的重要问题之一。本文以工业生产者购进价格指数（PPIRM）、工业品出厂价格指数（PPI）、居民消费价格指数（CPI）为研究对象，基于Granger因果检验分析价格指数构成要素之间的价格传导关系，提出一种新的多元传导模型对三种价格指数的环比数据进行预测，并在此基础上计算翘尾因素以推算物价的同比数据。实证结果表明，本文所提出的新的预测方法对物价环比数据和同比数据的预测效果在水平预测和方向预测的角度上，均优于基准模型。基于这一方法对2018年我国物价进行了预测，主要结果包括：2018年我国物价整体稳定，CPI将温和上涨，涨幅比2017年提高，全年上涨2.1%，其中翘尾因素影响1个百分点；PPI和PPIRM将继续上涨，但涨幅比2017年回落，全年分别上涨3.6%和4.3%，其中翘尾因素分别影响2.4和2.8个百分点。

关键词: PPIRM, PPI, CPI, 物价预测, 多元传导模型, 2018年中国经济

Abstract:

Accurate forecasting of price indexes would provide solid support for effective macroeconomic policy decisions. Thus, one of the most significant issues in economic policy practice is to improve the forecasting performance of price indexes. In this paper, a new approach is proposed to forecast three main price indexes: Purchase Price Index for Industrial Products (PPIRM), Producer Price Index for Industrial Products (PPI) and Consumer Price Index (CPI). The proposed multivariate transmission method is based on the price transmission mechanism among the three price indexes by using the Granger causality test. According to this method, different econometric forecasting models are selected for different components of the three price indexes and the results are properly integrated. The empirical results indicate that the integrated model based on the multivariate transmission method outperforms the benchmark model in terms of both level and directional predictive accuracy. Furthermore, the method is used to forecast China's CPI, PPI and PPIRM in 2018. The results suggest that in 2018, CPI will modestly increase by 2.1 percent from 2017 with 1 percent caused by tail-raising factor, and PPI and PPIRM will increase by 3.6 percent and 4.3 percent respectively with 2.4 percent and 2.8 percent caused respectively by tail-raising factor.

Key words: PPIRM, PPI, CPI, inflation forecasting, multivariate transmission method, China's economy in 2018