Causal Network Representations in Factor Investing

47 Pages Posted: 24 Jan 2024

See all articles by Clint Howard

Clint Howard

Robeco Quantitative Investments

Harald Lohre

Robeco Quantitative Investments; Lancaster University Management School

Sebastiaan Mudde

Erasmus University Rotterdam (EUR)

Date Written: December 30, 2023


Given the mounting criticism of correlation-based models in the factor investing literature, we use causal discovery algorithms to revisit three salient investment applications through the lens of novel causal network representations. First, we investigate peer group neutralization for long–short equity factors. While causality-based peer groups often surpass correlation-based or industry counterparts in Sharpe ratio performance, industry groups remain crucial for volatility reduction. Second, we derive a long–short low centrality equity factor from the causal network where peripheral stocks emerge as potential hedges against value, and central stocks tend to be larger, value companies. Lastly, a causal network density indicator demonstrates return predictability, with a diminishing density predicting negative equity market returns. Causal networks thus provide novel insights into factor investing, yet their implementation calls for cautious pioneering as they bring about computational complexity and render interpretability more challenging.

Keywords: causal discovery, factor investing, asset pricing, financial networks, market timing

JEL Classification: C32, C38, G11, G12

Suggested Citation

Howard, Clint and Lohre, Harald and Mudde, Sebastiaan, Causal Network Representations in Factor Investing (December 30, 2023). Available at SSRN: or

Clint Howard (Contact Author)

Robeco Quantitative Investments ( email )

Weena 850
Rotterdam, 3014 DA

Harald Lohre

Robeco Quantitative Investments ( email )

Weena 850
Rotterdam, 3011 AG

Lancaster University Management School

Lancaster LA1 4YX
United Kingdom


Sebastiaan Mudde

Erasmus University Rotterdam (EUR)


Do you have negative results from your research you’d like to share?

Paper statistics

Abstract Views
PlumX Metrics