China Markets and Global Stock Return Predictability

Abstract

This paper explores the China equity market's impact on the global equity markets through the lens of asset pricing. We study the predictive properties of the lagged China returns for global stock returns and find that the lagged China returns can significantly predict other markets' stock returns, but not vice versa. We augmented three selected global equity pricing models (Chen, Roll, and Ross's model (CRR), Ang and Bekaert's model (AB), and Rapach, Strauss, and Zhou's model (RSZ)) with the lagged China returns respectively, and find that these popular global stock return predictors cannot explain the predictive power of the lagged China returns. Moreover, we find that only the lagged China returns can lead the international returns at the one-month horizon but not the lagged US returns. Further evidence reveals it is due to that the information diffusion of the China market occurs at the longer-horizons (e.g. one-month) while the information diffusion of the US market occurs at the shorter-horizons (e.g. one-week). Lastly, we documented that the lagged China returns during the low investor-attention period have stronger predictability compared to the predictive performance during the high attention period, which is in line with the information friction theory that when the attention-constrained investors make investment decisions they fail to allocate attention to certain economic state variables, which causes the slow information diffusion across markets and leads to the cross-market return predictability. Overall, our results indicate that the lagged China returns should be regarded as a global state variable that helps us predict future stock returns.