Identifying the drivers of economic uncertainty perception in China: a news-based approach

Abstract

Relying on news media as a reflection of public sentiment, we build the Uncertainty Perception Indicator (UPI) to measure perceived economic uncertainty in China. Applying Latent Dirichlet Allocation to n = 5,600 news articles, we decompose UPI into three components: Uncertainty relating to the real economy, financial markets, and politics. We find that on average, overall uncertainty perception is mainly driven by uncertainty about the real economy. This is especially true since 2018, when around 60 percent of the corpus was related to the real economy. We also find that stock price exposure is highest for real economic uncertainty.