Distributional Adversarial Attacks and Training in Deep Hedging

2025-08-20

AI Summary

This paper addresses the limited robustness of classical deep hedging strategies under minor input distribution shifts. To this end, we propose a distributionally robust training framework specifically designed to withstand distribution-level adversarial perturbations. Our key innovation lies in generalizing pointwise adversarial attacks to the distributional level by formulating a computationally tractable adversarial optimization objective over a Wasserstein ball. The method integrates deep hedging, distributionally robust optimization, and stochastic process modeling, employing adversarial training to explicitly enhance model resilience against model misspecification and market uncertainty. Empirical evaluations demonstrate that the proposed approach significantly outperforms conventional deep hedging models in out-of-sample performance, distributional shift resistance, and generalization stability. By bridging distributional robustness with financial hedging, our framework delivers a more reliable and practically viable deep learning solution for real-world financial decision-making.

Technology Category

• Machine Learning: Adversarial Learning & Robustness
• Computer Vision: Adversarial Attacks & Robustness
• Search and Optimization: Adversarial Search

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

In this paper, we study the robustness of classical deep hedging strategies under distributional shifts by leveraging the concept of adversarial attacks. We first demonstrate that standard deep hedging models are highly vulnerable to small perturbations in the input distribution, resulting in significant performance degradation. Motivated by this, we propose an adversarial training framework tailored to increase the robustness of deep hedging strategies. Our approach extends pointwise adversarial attacks to the distributional setting and introduces a computationally tractable reformulation of the adversarial optimization problem over a Wasserstein ball. This enables the efficient training of hedging strategies that are resilient to distributional perturbations. Through extensive numerical experiments, we show that adversarially trained deep hedging strategies consistently outperform their classical counterparts in terms of out-of-sample performance and resilience to model misspecification. Our findings establish a practical and effective framework for robust deep hedging under realistic market uncertainties.

Problem

• Robustness of deep hedging under distributional shifts
• Vulnerability to input distribution perturbations
• Adversarial training for resilient hedging strategies

Innovation

• Adversarial training for deep hedging robustness
• Distributional attacks over Wasserstein ball perturbations
• Computationally tractable reformulation for resilient strategies