Why Market-Cap Neutralization May Not Remove Small-Cap Model Bias
Summary
This brief Q&A explains why neutralizing individual factors for market capitalization does not guarantee that a trained stock-selection model will have no size exposure. After neutralization, linear groupings of the factor may contain both large- and small-cap stocks, yet a nonlinear model can still learn patterns that favor smaller companies.
The document’s central lesson is to distinguish factor-level neutrality from the behavior of the complete model and its selected portfolio. It offers this conceptual explanation but provides no example, diagnostic procedure, dataset, or empirical results. Researchers should therefore treat the claim as a warning about possible nonlinear exposure and inspect the model’s actual selections and portfolio characteristics rather than assuming factor preprocessing has eliminated size bias.
Key ideas
- Neutralizing a factor for market capitalization does not ensure the final model portfolio is size-neutral.
- Linear factor groups may mix large- and small-cap stocks while a nonlinear model still learns size-related patterns.
- Model and portfolio exposures should be checked after training instead of inferred from preprocessing alone.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.