Do Sentiment Indicators Help to Assess and Predict Actual Developments of the Chinese Economy?

Abstract

This paper evaluates the usefulness of business sentiment indicators for forecasting developments in the Chinese real economy. We use data on diffusion indices collected by the People's Bank of China for forecasting industrial production, retail sales and exports. Our bivariate vector autoregressive models, each composed of one diffusion index and one real sector variable, generally outperform univariate AR models in forecasting one to four quarters ahead. Similarly, principal components analysis, combining information from various diffusion indices, leads to enhanced forecasting performance. Our results indicate that Chinese business sentiment indicators convey useful information about current and future developments in the real economy. They also suggest that the official data provide a fairly accurate picture of the Chinese economy.

Keywords: forecasting, diffusion index, VAR, China

JEL Classification: E32, E37, P27