Uncertainty and Explainability in Deep Rough Volatility: Neural Simulation-Based Inference for Rough Heston Calibration

Paper Overview

• Field: Machine Learning
• Authors: Damiano Brigo, Raphaël Huser, Dan Leonte
• Published: 2026-09-25
• arXiv: 2609.31570

Summary

Deep learning has substantially accelerated the calibration of complex stochastic-volatility models, but neural point calibration alone does not capture the uncertainty remaining after an implied-volatility (IV) surface has been observed. The authors develop a simulation-based inference framework for rough Heston (rHeston) calibration that learns the posterior distribution of the model parameters conditional on an IV surface.

Using neural ratio estimation, they obtain calibrated posterior samples that can be propagated through heteroscedastic neural surrogate pricers for path-dependent exotic options. The resulting posterior-predictive distributions combine residual parameter uncertainty with conditional surrogate uncertainty and yield uncertainty-aware price intervals.

The paper further introduces Hellinger-SHAP, an information-theoretic explainability method for posterior inference. Instead of attributing to single parameter point estimates, it applies local-to-background Kernel SHAP to a posterior information functional that measures shrinkage from prior to posterior. This identifies maturity-moneyness regions associated with posterior information gain for individual rHeston parameters.

Key Findings

• Posterior-predictive intervals provide calibrated or conservative coverage for forward-starting, barrier, and realized-variance contracts.
• Point plug-in prices can be severely unreliable for some contract types.
• Uncertainty quantification and explainability analysis together form a transparent framework for uncertainty-aware neural calibration and downstream exotic option pricing under a specified prior predictive model.

*Auto-collected on 2026-09-29*