基于混频模型的外商直接投资对中国宏观经济波动预测研究

Prediction of Foreign Direct Investment on China Macroeconomic Fluctuations Based on MIDAS Model

引入外商直接投资(foreign direct investment,FDI)和“三驾马车”,基于混频数据抽样模型预测中国季度GDP增长率,考察新冠疫情背景下FDI对中国宏观经济波动的影响。研究发现,单变量混频模型的均方根误差(root mean square error,RMSE)小于基准模型,预测精准度更高;多元混频模型引入FDI后,模型的实时预报RMSE比值为0.13,低于未引入的0.16,预测季度GDP增长率更精准;基于FDI预测的季度GDP增长率较“三驾马车”更高,表明FDI促进经济增长。

The Mixed-frequency Data Sampling model introduced foreign direct investment(FDI)and the“troika”,predicted China's quarterly GDP growth rate,and examined the impact of FDI on China's macroeconomic fluctuations under the background of COVID-19.The research shows that the root mean square error(RMSE)of the unit mixing model is smaller than the RMSE of the benchmark model,and its prediction accuracy is higher.After the introduction of FDI into the multivariate mixing model,the RMSE ratio of real-time forecast is 0.13,which is smaller than 0.16 without the introduction of FDI.The model predicts quarterly GDP growth rate more accurately.The quarterly GDP growth rate forecast based on FDI is higher than that based on“troika”,which proves that FDI promotes economic growth.