基于时间序列模型的货币供应量预测

Forecast of Money Supply Based on Time Series Model

货币供给是重要的金融变量和重要的宏观经济变量,对稳定宏观经济运行以及金融市场起着不可忽视的作用。本文对2010年1月~2020年4月的货币供应量月度数据拟合SARIMA(2,1,1)(1,0,0)12模型,并预测2020年5~9月的数据,将预测值与实际观察值进行对比。根据SARIMA(2,1,1)(1,0,0)12拟合结果发现,第一,我国广义货币供应量变动规律主要受上两期货币供给的影响和上一期货币供给随机干扰的影响。第二,2020年5月~9月份数据的平均相对误差为0.228%,说明模型的预测效果良好。第三,选择的模型不仅能很好地拟合序列趋势,对未来长期货币供应量的预测也有很好的借鉴意义。

Money supply is an important financial variable and an important macroeconomic variable, which plays an important role in stabilizing macroeconomic operation and financial market. In this paper, the monthly data of money supply from January 2010 to April 2020 is fitted with SARIMA(2,1,1)(1,0,0)12 model, and the data from May to September 2020 are predicted. The predicted value is compared with the actual observed value. According to SARIMA(2,1,1)(1,0,0)12 fitting results, it is found that, firstly, the change law of broad money supply in China is mainly affected by the last two periods of money supply and the random interference of the last period of money supply. Second, the average relative error of the data from May to September in 2020 is 0.228%, which shows that the prediction effect of the model is good. Third, the selected model can not only fit the trend of the series well, but also provide a good reference for the prediction of the future long-term money supply.