Uncertainty and Explainability in Deep Rough Volatility: A Neural Information-Theoretic Posterior Approach
Neural posterior inference calibrates rough Heston models and uses Hellinger-SHAP to explain implied-volatility information gain.
The paper presents simulation-based inference for rough Heston calibration, learning parameter posteriors from an implied-volatility surface with neural ratio estimation. Posterior samples pass through heteroscedastic neural surrogate pricers for path-dependent exotic options, combining parameter and surrogate uncertainty. Hellinger-SHAP attributes posterior information gain to maturity-moneyness regions rather than a single point estimate. In simulation, posterior-predictive intervals covered forward-start, barrier, and realized-variance claims more reliably than point plug-in prices.
- Neural ratio estimation learns rough Heston posteriors from an IV surface.
- Hellinger-SHAP locates maturity-moneyness regions that inform each parameter.
- Predictive intervals cover exotic claims better than point plug-in prices.
From the original article
1 min read at arxiv.orgDeep 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. We 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.
By Damiano Brigo · arxiv.org
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.