Causal Factor Investing: Can Factor Investing Become Scientific?

Journal articles in factor investing literature make associational claims, deny the causal content of factor models, and do not propose experiments for falsifying causal mechanisms. This prevents factor investing from advancing beyond its current phenomenological stage. Solutions are proposed to transform factor investing into a scientific discipline.

Format: Paperback / softback
Length: 75 pages
Publication date: 09 November 2023
Publisher: Cambridge University Press

The vast majority of journal articles in the field of factor investing assert associational claims, disregarding the underlying causal structure of factor models. Authors fail to identify the specific causal graph that corresponds to the observed phenomenon, instead relying on correlations to justify their chosen model specification. Moreover, they fail to propose experimental designs to test the validity of causal mechanisms. Without a robust causal theory, the findings derived from these studies are highly suspect, as they are prone to backtest overfitting and incorrect specification choices. This element distinguishes between type-A and type-B spurious claims and elucidates how both types hinder the advancement of factor investing beyond its current phenomenological stage. It conducts an in-depth analysis of the current state of causal confusion within the factor investing literature and presents potential solutions aimed at transforming factor investing into a rigorous and scientific discipline. This title is also accessible as Open Access on Cambridge Core.

The vast majority of journal articles in the field of factor investing assert associational claims, disregarding the underlying causal structure of factor models. Authors fail to identify the specific causal graph that corresponds to the observed phenomenon, instead relying on correlations to justify their chosen model specification. Moreover, they fail to propose experimental designs to test the validity of causal mechanisms. Without a robust causal theory, the findings derived from these studies are highly suspect, as they are prone to backtest overfitting and incorrect specification choices. This element distinguishes between type-A and type-B spurious claims and elucidates how both types hinder the advancement of factor investing beyond its current phenomenological stage. It conducts an in-depth analysis of the current state of causal confusion within the factor investing literature and presents potential solutions aimed at transforming factor investing into a rigorous and scientific discipline. This title is also accessible as Open Access on Cambridge Core.

Weight: 154g
Dimension: 152 x 228 x 8 (mm)
ISBN-13: 9781009397292