Do factor models capture both sentiment and limited attention?

We demonstrate that valuation uncertainty and information arrival are critical stock characteristics determining whether individual factors in leading factor models are influenced by sentiment or limited attention. Therefore, the ability of a factor model to explain cross-sectional stock returns depends on including two distinct types of factors: those that capture sentiment and those that tackle limited attention. Yet, many leading factor models include factors for sentiment but fall short in incorporating factors for limited attention. Our findings are important, guiding future research towards developing new factor models that more effectively capture both sentiment and limited attention compared to existing models. This includes uncovering powerful factors capable of simultaneously capturing sentiment and limited attention. © 2025 Published by Elsevier B.V.