WRAP: Adversarial Training Tackles Deep Hedging in Nonstationary Markets via DRO
chaumian · x · 2026-10-07
A new arXiv paper introduces WRAP (Wasserstein-Reweighting Adversarial Perturbation), a drift-aware adversarial training framework for deep hedging in nonstationary markets.
- The problem: deep hedging learns trading policies from historical or simulated trajectories, which may not represent future market conditions under nonstationarity.
- Method: derived from a two-budget distributionally robust optimization (DRO) formulation anchored to a weighted empirical reference distribution; an adversary can reweight trajectories under a φ-divergence constraint and perturb paths under an optimal-transport constraint.
- Theory: a joint first-order expansion decomposes the leading-order loss increase into a reweighting term (dispersion of hedging losses) and a transport term (sensitivity to path perturbations), replacing the distributional inner supremum with a tractable finite-dimensional attack.
- Experiments on stationary and nonstationary Heston dynamics and a generalized affine diffusion (GAD) show complementary gains.
More from Research
- NVIDIA's VeriFine Co-Evolves Policy and Judge to Scale Self-Improvement in Embodied Reasoning — nvidia · 2026-10-07
- Single Image to Full 3D Scene: Adaptive Chunking Extends Object Generators to Outdoor Rome — Jiraphon Yenphraphai · 2026-10-07
- EmbodiedSmith: Recursive Self-Improvement Flywheel Scales Embodied Training Data in Simulation — Yikai Qin · 2026-10-07
- Video World Models Flunk Physics: Best Model Scores 57.76/100 on New 40-Task Benchmark — Mingju Gao · 2026-10-07
- Researcher Hails New Theoretical Breakthrough in Gradient Descent and Matrix Multiplication — cloneofsimo · 2026-10-07
- OpenAI claims 372 unsolved problems cracked, averaging about 3 hours each — i_dg23 · 2026-10-07