Taming the Factor Zoo: A Comprehensive Review of Methodologies for Screening Factor Redundancy in Asset Pricing Models

Empirical asset pricing has been changing and improving rapidly, moving from the single-factor CAPM to multiple-factor models. But on the other hand, this has also brought the so-called “Factor Zoo,” that is, hundreds of potential pricing factors have been found, which leads to a big problem of dimension and redundancy. The method used in this paper is a systematic literature review and theoretical analysis, the main difficulty being to identify and remove redundant factors. The Fama-French three-factor model to the q-factor model. The paper also discusses existing screening methods for redundancies, such as spanning regression, and new techniques like machine learning. The findings of the study are that traditional spanning tests are still important, but not enough for high-dimensional data. Advanced statistics like regularization and dimension reduction give us better ways to tell apart the real pricing reasons and extra noise that doesn’t matter. This paper concludes that the future of research in asset pricing is to develop screening methods based on economics. This is necessary to resolve the redundancy problem and restore simplicity to factor models.