Prediction of CPI in China based on MIDAS

Abstract

Over the past few decades, accurate forecasting of CPI is crucial for government policy making and corporate strategic planning. The purpose of this study is to utilize the mixed data sampling (MIDAS) regression method to forecast the CPI in China and to compare its forecasting performance with the traditional autoregressive moving average (ARIMA) model.  We demonstrate that the MIDAS regression model is a promising approach for capturing and forecasting the complex dynamics of macroeconomic indicators. With this approach, policymakers and entrepreneurs can obtain more accurate and comprehensive CPI forecasts and thus make more informed decisions