Price comovement and market segmentation of Chinese A- and H-shares: Evidence from a panel latent-factor model

This paper examines the price comovement of cross-listed Chinese A- and H-shares using a panel model with latent factors and a heterogeneous long-run structure. Our model is more flexible than the cointegration system and is estimated using the data-driven Cup–Lasso method. The long-run H-share price discounts are heterogeneous across different groups of stocks. We have identified both stationary and nonstationary latent factors in the price differentials, which are driven by different economic variables. By analyzing the factor loadings of the nonstationary latent factor, we identify some trading-friction and information-friction variables that have effects on the price convergence between the A- and H-shares.