CPI波动分析的SARIMA模型拟合及预测

The Fitting and Forecasting of CPI Fluctuation under the SARIMA Model

通货膨胀是经济发展过程中存在的一个重要问题,而CPI又是衡量通货膨胀程度最重要的指标。该文采用1990年1月至2011年12月中国消费者价格指数月度数据,通过对消费者价格指数(CPI)自相关函数与偏自相关函数的统计识别以及ADF平稳性检验,建立SARIMA时间序列模型,对消费者价格指数进行分析和短期预测。实证结果表明:中国消费者价格指数的发展变化情况具有明显的趋势性和季节性,SARIMA模型能较好地拟合和预测我国CPI变化情况。

Inflation is an important problem in the process of economic development,while the CPI is one of the most important indicators of inflation.In this paper,a SARIMA time series model was built to analyze and forecast the short-time CPI,which was based on the monthly data from 1990 to 2011.Meanwhile,the statistical recognition of autocorrelation function and partial autocorrelation function and ADF stationary test were used to inspect their natures.The research results show that the CPI of China has obvious tendency and seasonal fluctuation.Furthermore, the SARIAM model adequately describes the Chinese CPI fluctuation and forecasts the future short-time CPI.