Abstract

Motivated by the implication of return extrapolation models that a joint consideration of past price changes and firm fundamentals could efficiently identify stock mispricing, we propose an integrated approach that combines fundamental and technical information. This integrated approach generates substantial economic gains, which are comparable to those of strategies double-sorted on characteristics related to high turnover and trading costs and state-of-the-art machine learning strategies in existing studies. The performance net of transaction costs is still attractive. Simple transaction cost mitigation approaches could further enhance the performance of the integrated approach by reducing portfolio turnover. Consistent with behavioural models, limits to arbitrage and information asymmetry play a significant role in explaining the super performance of this integrated approach.