Forecasting China's inflation in a data-rich environment

Abstract

Inflation is one of the most recent critical issues facing China. To improve inflation forecasts within China, this study investigates the predictive ability of three dimension reduction techniques used in a data-rich environment: Principal Components Analysis (PCA), Sliced Inverse Regression (SIR), and Partial Least Squares (PLS) applied in the Factor-Augmented Autoregression (FAAR) model proposed by Stock and Watson (2005). Varied macroeconomic data from China between January 1998 and December 2009 are obtained to construct factors for use by three different techniques. The performance of different dimension reduction methods depends on forecasting horizons, the number of factors chosen, and the number of slices for SIR. The empirical study finds that the FAAR model with an optimal number of PCA factors outperforms the other model in out-of-sample inflation forecasting in China.