Attribution Methods in Asset Pricing: Do They Account for Risk?

TL;DR

This paper evaluates whether attribution methods in machine learning accurately reflect financial risks in asset pricing, proposing domain-specific axioms and analyzing the effectiveness of popular methods like Shapley value and Integrated Gradients.

Contribution

It introduces domain-specific axioms for attribution in asset pricing and assesses the suitability of existing methods to reflect financial risks accurately.

Findings

Shapley value and Integrated Gradients preserve most axioms but not all.

Attribution methods can reflect risks if applied appropriately.

Certain axioms are incompatible with some attribution methods.

Abstract

Over the past few decades, machine learning models have been extremely successful. As a result of axiomatic attribution methods, feature contributions have been explained more clearly and rigorously. There are, however, few studies that have examined domain knowledge in conjunction with the axioms. In this study, we examine asset pricing in finance, a field closely related to risk management. Consequently, when applying machine learning models, we must ensure that the attribution methods reflect the underlying risks accurately. In this work, we present and study several axioms derived from asset pricing domain knowledge. It is shown that while Shapley value and Integrated Gradients preserve most axioms, neither can satisfy all axioms. Using extensive analytical and empirical examples, we demonstrate how attribution methods can reflect risks and when they should not be used.