Stock return predictability and model instability: Evidence from mainland China and Hong Kong

This study examines the predictability of the Shanghai Composite, Shenzhen Composite and the Hang Seng China Enterprise index returns during the period 1993 to 2010, with emphasis on whether considering structural breaks in model parameters improves the stock return predictability. Results indicate higher linear stock return predictability for the Hong Kong market than for the Chinese markets. However, the results differ when model instability is considered. Specifically, using Bai and Perron's (1998, 2003) approach, the results indicate the presence of structural breaks particularly for the Shenzhen market, which appear to coincide with major economic events or political and institutional changes. The predictable component in stock returns is also time-varying when re-estimating the model over different subsamples defined by the break. Overall, results highlight the importance of considering breaks in forecasting stock returns, and suggest that the Hong Kong market is a relatively ideal haven to park wealth for risk-averse investors whereas the Shenzhen market offers enhanced opportunities for risk-seeking investors. (C) 2017 Board of Trustees of the University of Illinois. Published by Elsevier Inc. All rights reserved.