Can systemic risk measures predict economic shocks? Evidence from China

Guojin Chen,Yanzhen Liu,Yu Zhang

DOI: https://doi.org/10.1016/j.chieco.2020.101557

IF: 4.744

2020-01-01

China Economic Review

Abstract:In this paper, we investigate how the systemic risk affects macroeconomic activity in China and examines the forecasting power of 12 different measures of systemic risk. Quantile regression is employed to capture the nonlinear relationship between the systemic risk and the distribution of future macroeconomic shocks. We find that the systemic risk skews the distribution of future shocks, which cannot be identified in the central tendency analysis within the traditional linear regression. In particular, when the systemic risk builds up, the risk of severe economic downturns increases while the risk of moderate economic downturns barely changes. When comparing the forecasting power of different systemic risk measures, we use both a fixed rolling window and a time-varying method to make the result robust. We find that, of the 12 widely used measures, 8 demonstrate significant predictability for subsequent shocks to economic growth in China and can thus serve as early warning signals.