Real-time macroeconomic monitoring using mixed frequency data: Evidence from China

Abstract

Timely and effective monitoring of macroeconomic dynamics is crucial for the management and stabilization of the macroeconomy. However, existing studies tend to rely on low-frequency macro-level data to monitor economic fluctuations and fail to capture changes that take place in real time. This study therefore develops a novel real-time monitoring system of China's macroeconomic prosperity by incorporating monthly and quarterly macroeconomic data with high-frequency daily Internet search data from January 1, 2012 to March 31, 2019. We find that our daily prosperity indexes can provide real-time monitoring signals with daily updates, significantly improving the accuracy of key macroeconomic indicators for nowcasting. Moreover, our monitoring system incorporates leading, coincident, and lagging indexes, and enables simultaneous monitoring of the synergistic and asymmetric characteristics of cyclical fluctuations, depicting a comprehensive picture of the business cycle. Our findings have practical implications for enhancing the quality and timeliness of macroeconomic monitoring and early intervention. • We develop a novel real-time monitoring system of macroeconomic prosperity. • Our model incorporates macroeconomic data with daily Internet search data. • Leading, coincident, and lagging indexes are integrated into one monitoring system. • The constructed daily indexes improve the accuracy of macroeconomic monitoring. • Our findings provide practical references for timely macroeconomic management.