Evaluating asset pricing models: A revised factor model for China

Abstract

We develop a revised factor model, accounting for unique features of Chinese markets, and evaluate the performance of competing asset pricing models. Extant literature reveals that eliminating the smallest 30% of stocks improves the performance of factor models. The revised factor model excludes firms with a high expected probability of becoming shells, which are companies valued as shells in reverse mergers serving as an alternative way to go public. Our revised model has the smallest model specification errors and the best explanatory power among various constructed portfolios. This new finding suggests that our model offers an effective benchmark model for empirical asset pricing in the Chinese stock market.