改进货币总量统计能否改善我国通货膨胀预测精度?——基于GMDH神经网络模型的分析  

Can Improving Monetary Aggregate Statistics Improve the Accuracy of China’s Inflation Forecast?Analysis Based on GMDH Neural Network Model

作　　者：王勇 周潮[2] WANG Yong;ZHOU Chao(School of History,Northwest University,Xi’an 710028,China;Sub-branch of Zhangye City,The People’s Bank of China,Zhangye 734000,China)

机构地区：[1]西北大学历史学院,陕西西安710028 [2]中国人民银行张掖市中心支行,甘肃张掖734000

出　　处：《西安财经大学学报》2020年第6期13-20,共8页Journal of Xi’an University of Finance and Economics

摘　　要：文章基于我国1990年1季度至2019年4季度统计数据,利用GMDH神经网络模型,对比分析了加权货币总量、简单货币总量、市场利率、GDP增速4种测度指标在预测通货膨胀时的具体表现。研究结果表明:两类货币总量指标对通货膨胀的预测效果要优于GDP增速和市场利率,能够成为中国人民银行通货膨胀预测的重要测度因子;加权货币总量的预测精度要明显优于其他指标,简单货币总量也有较高的预测精度,而GDP增速的预测精度相对最低。自我国宏观经济进入下行周期以来,近期国内消费品价格出现的持续快速上涨已引发社会广泛关注,这就要求中国人民银行进一步完善货币总量统计,不断提高通货膨胀预测能力和预测精度,并前瞻性地调整货币政策操作,及时消除宏观经济的波动性,为经济高质量发展创造条件。Based on data of the first quarter of 1990 to the fourth quarter of 2019,using GMDH neural network model,this paper comparative analyzed the specific performance of four measurement indicators of weighted currency total,simple currency total,market interest rate,and GDP growth rate in forecasting inflation.The results show that the two types of monetary aggregates index forecast for inflation effect is superior to the GDP growth and market interest rates,inflation is forecast to become the People’s Bank of China’s important measure factor;The forecast accuracy of weighted monetary aggregate is obviously better than other indexes,the forecast accuracy of simple monetary aggregate is also higher,and the forecast accuracy of GDP growth is relatively the lowest.Since the macroeconomic of China entered the downward cycle,the recent domestic consumer prices have continued rapid rise has caused widespread concern,this requests the People’s Bank of China should further improve the monetary aggregates,rising inflation prediction ability and prediction accuracy,and the adjustment of the forward-looking monetary policy operation,timely eliminate macroeconomic volatility,create conditions for high quality and economic development.

关 键 词：中国人民银行 通货膨胀 货币总量 GMDH模型 

分 类 号：F822.5[经济管理—财政学]