Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 2610.16732

GROMM: Generative Risk-neutral Option Market Models

Magnus Wiese·Hans Buehler·Blanka Horvath
View original source ↗Read source document ↗Export metadata ↓

Abstract

We present a statically and dynamically arbitrage-free European option market model with smooth interpolation between option prices, based on the SANOS framework, and prove that its dynamics admit a unique strong solution. In spirit, it is a stochastic model for local volatility when the latter is used as a representation of a European option price surface, not as a pricing model. We illustrate the strength of our approach on S&P Index Option data from January 2022 to August 2025. We show that the model's initial state fits within bid/ask for more than 95% of relevant market option prices. We then train the model dynamics under $\mathbb{P}$ and $\mathbb{Q}$ and show that the model is indeed free of arbitrage while fitting the dynamics of the market well. We also propose a range of BHW-tests to test for absence of arbitrage.

Explore related subjects

q-fin.MFmath.PRq-fin.CPq-fin.PRq-fin.ST

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Magnus Wiese, Hans Buehler, Blanka Horvath. 2026-10-12. GROMM: Generative Risk-neutral Option Market Models. https://arxiv.org/abs/2610.16732

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Interior smoothness and zero-variance geometry for finite-maturity Heston American puts

We study finite-maturity American puts in the Heston model without the Feller condition. At positive variance, the exercise boundary is smooth and strictly decreasing in remaining maturity and variance. At zero variance, the endpoint is locally Lipschitz in remaining maturity, the boundary moves with bounded speed in maturity uniformly down to zero variance, and it extends continuously to the axis. Each boundary slice at positive maturity has a regular $C^1$ closure tangent to the zero-variance axis, its stock threshold has variance derivative tending to minus infinity, and both partial derivatives of the inverse boundary graph vanish along the slice. The continuation premium at the axis contact stock is of order variance, and as a consequence smooth fit fails in the variance direction at the axis. At expiry the whole boundary, including its axis edge, converges to the strike. We give quantitative boundary gap bounds, an exact limiting weighted boundary Hessian, and vertical identities retaining the maturity term. Exact boundary gap equivalents remain open.

q-fin.MF↗
arXiv2026

Regime Detection in Latent Factor Models -- A Bayesian Decision-Theoretic Approach

Regime-switching multi-factor models are widely used in quantitative finance, but filtering does not by itself determine when an agent should act on inferred regime changes. We formulate online regime detection with latent factors and hidden regimes as a Bayesian decision problem balancing false alarms, detection delay, and misclassification. Since posterior regime probabilities are generally non-Markovian, the problem is represented as a measure-valued optimal stopping problem and approximated using particle filtering with Monte Carlo dynamic programming. A single-change-point Heston model provides a benchmark. With observed variance, the problem reduces to a two-dimensional Markov free-boundary problem, while under latent variance decisions are based on the filtered posterior measure. Numerical experiments in Heston and regime-switching Vasicek models show that the latent-information procedure can outperform feasible plug-in rules.

q-fin.MF↗
arXiv2026

Portfolio Decarbonization under Incomplete Carbon and Transition Information

We study dynamic portfolio decarbonization under incomplete information about firms' carbon intensities and the underlying transition state. The investor maximizes expected utility of terminal wealth net of a penalty on perceived cumulative financed emissions. Asset returns and carbon intensities depend on a stochastic transition factor, while the investor may need to infer the current transition state, its long-run mean and firms' carbon intensities from market observations and periodic noisy disclosures. The optimal strategy combines a Merton-type component with climate-dependent expected-return adjustments, a direct carbon intensity penalty and intertemporal hedging demands associated with stochastic carbon intensities and learning. In a numerical application, incomplete information has limited effects on welfare and target satisfaction when prior beliefs are well specified. By contrast, biased beliefs generate welfare losses in either direction but asymmetric environmental consequences: optimistic and pessimistic beliefs induce opposite portfolio tilts and therefore affect the probability of exceeding the cumulative emissions target differently.

q-fin.MF↗