Which factors in China? A pre-registered study

We evaluate the ability of the asset pricing models of Liu, Stambaugh and Yuan (2019), hereafter LSY, and Fama and French (2018) to price Chinese stocks. Following the methodology specified in the pre-registered report (Gharghori and Nguyen, 2025), we employ spanning regressions and maximum Sharpe ratios to compare models. Our results show that the LSY models exhibit higher Sharpe ratios and outperform both conventional and modified Fama-French models in spanning regressions. In contrast to developed markets, the investment factor carries a negative premium in China. Additionally, the LSY models do not span this factor. Consequently, the model that delivers the highest Sharpe ratio is the one that augments the LSY four-factor model with the investment factor of Fama-French.