Paper: arXiv 2403.00746

Authors: Antonis Papapantoleon, Jasper Rou

Abstract

We develop a novel deep learning approach for pricing European options in diffusion models, that can efficiently handle high-dimensional problems resulting from Markovian approximations of rough volatility models. The option pricing partial differential equation is reformulated as an energy minimization problem, which is approximated in a time-stepping fashion by deep artificial neural networks. The proposed scheme respects the asymptotic behavior of option prices for large levels of moneyness, and adheres to a priori known bounds for option prices. The accuracy and efficiency of the proposed method is assessed in a series of numerical examples, with particular focus in the lifted Heston model.

Complexity vs Empirical Score

  • Math Complexity: 8.5/10
  • Empirical Rigor: 6.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper introduces a novel deep learning method for solving high-dimensional PDEs derived from rough volatility models, requiring advanced mathematics in gradient flows and neural network optimization. It includes numerical experiments (e.g., lifted Heston model) with error metrics and computational details, but lacks extensive real-world backtesting or code/data, placing it in the high-math, high-rigor quadrant.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Develop efficient pricing method for<br>European options in high-dimensional<br>rough diffusion models"] --> B["Methodology: Reformulate PDE as<br>Energy Minimization Problem"]
  B --> C["Time-stepping Approximation:<br>Deep Neural Networks"]
  C --> D["Computational Process:<br>Approximate value function via<br>sequential network training"]
  D --> E["Data/Inputs:<br>Lifted Heston Model &<br>Diffusion Models"]
  E --> F["Outcomes:<br>1. Asymptotic behavior respected<br>2. A priori bounds adhered to<br>3. Accurate & Efficient<br>4. High-dimensional capability"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes a novel deep learning time‑stepping deep gradient flow (TDGF) method for pricing European options in diffusion models.

Source sentence
We develop a novel deep learning approach for pricing European options in diffusion models, that can efficiently handle high-dimensional problems resulting from Markovian approximations of rough volatility models.

Assumes 5

  • The diffusion operator A is symmetric and positive semi‑definite. model
    Source sentence
    Assume that A is symmetric and positive semi-definite. Then, Uk solves (3.2) if and only if Uk minimizes …
  • The underlying asset price follows a Markovian diffusion process. model
    Source sentence
    Let S denote the price process of a financial asset that evolves according to a (Markovian) diffusion model
  • The time‑stepping scheme uses a backward differentiation formula and a fixed number of time steps N=100. other
    Source sentence
    We set the number of time steps equal to N = 100, use 2000 sampling stages in each time step
  • Neural network architecture: 3 layers with 50 neurons per layer, tanh and softplus activations. other
    Source sentence
    In the numerical experiments, we have used 3 layers and 50 neurons per layer. The activation functions we have selected are the hyperbolic tangent function, σ1(x) = tanh(x), and the softplus function, σ2(x) = log (ex + 1)
Show 1 more
  • Training uses the Adam optimizer with learning rate α = 3 × 10⁻⁴, (β1,β2) = (0.9,0.999) and zero weight decay. other
    Source sentence
    In the optimization stage, we use the Adam algorithm [41] with a learning rate α =

Data

  • Heston model parameter set synthetic unknown
    Source sentence
    We set the number of time steps equal to N = 100, use 2000 sampling stages in each time step, while in each sampling stage we take 600 samples per dimension

Costs

Not applicable The paper reports no trading or execution result.

Finds 3

  • The TDGF method achieves an overall pricing error on the order of 10⁻³, comparable to other deep learning methods but with significantly smaller training times. performance
    Source sentence
    The empirical results show that the proposed method is accurate, with a small overall error in the order of 10−3 and comparable with other deep learning methods, while the training times are significantly smaller compared to similar methods and do not grow significantly with the dimension of the underlying space.
  • TDGF training and evaluation times are faster and more consistent than the DGM method. performance
    Source sentence
    Overall, the TDGF method developed here performs well in the numerical experiments, with an error in the order of 10−3. Moreover, the training and computation (evaluation) times for the TDGF method are faster and more consistent compared to the popular DGM method.
  • When the dimension of variance processes and correlation increase in the lifted Heston model, the error increases due to the diffusion term becoming smaller. negative
    Source sentence
    However, when the dimension of the variance processes in the lifted Heston model increases and the correlation increases simultaneously, then the error also increases. This increase in the error is due to the diffusion term becoming smaller, which means that the explicit term in the decomposition (3.1) becomes the domi

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 21 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 3 papers in the archive (1 by independent authors)

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.

Cited References (39)

  1. Lifting the H eston model (2019)
  2. Markovian structure of the V olterra H eston model (2019)
  3. Multifactor approximation of rough volatility models (2019)
  4. Volatility models in practice: Rough, Path-dependent or Markovian? — → 2401.03345
  5. Implicit-explicit multistep methods for quasilinear parabolic equations (1999)
  6. Efficient option pricing in the rough H eston model using weak simulation schemes (2024)
  7. Markovian approximations of stochastic V olterra equations with the fractional kernel (2023)
  8. Weak Markovian approximations of rough H eston (2023)
and 31 more
  1. Pricing under rough volatility (2016)
  2. Short-time near-the-money skew in rough fractional volatility models (2019)
  3. Rough Volatility (2023)
  4. The pricing of options and corporate liabilities (1973)
  5. From elephant to goldfish (and back): memory in stochastic Volterra processes (2023)
  6. Generalized F eller processes and M arkovian lifts of stochastic V olterra processes: the affine case (2020)
  7. D elft B lue S upercomputer ( P hase 1) (2022)
  8. High-order compact finite difference scheme for option pricing in stochastic volatility models (2012)
  9. The deep R itz method: a deep learning-based numerical algorithm for solving variational problems (2018)
  10. The characteristic function of rough H eston models (2019)
  11. A novel pricing method for E uropean options based on F ourier-cosine series expansions (2009)
  12. The Volatility Surface: A Practitioner’s Guide (2012)
  13. Efficient Simulation of Affine Forward Variance Models (2022)
  14. Volatility is rough (2018)
  15. Discrete gradient flow approximations of high dimensional evolution partial differential equations via deep neural networks (2023)
  16. A deep implicit-explicit minimizing movement method for option pricing in jump-diffusion models — → 2401.06740
  17. Solving high-dimensional partial differential equations using deep learning (2018)
  18. A closed-form solution for options with stochastic volatility with applications to bond and currency options (1993)
  19. Sparse wavelet methods for option pricing under stochastic volatility (2005)
  20. Computational Methods for Quantitative Finance: Finite Element Methods for Derivative Pricing (2013)
  21. Deep learning volatility: a deep neural network perspective on pricing and calibration in (rough) volatility models (2021)
  22. Deep curve-dependent PDEs for affine rough volatility (2023)
  23. Random neural networks for rough volatility (2023)
  24. Mathematical Introduction to Deep Learning: Methods, Implementations, and Theory (2023)
  25. Adam: A method for stochastic optimization (2014)
  26. Deep N itsche method: Deep R itz method with essential boundary conditions (2019)
  27. The deep minimizing movement scheme (2023)
  28. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations (2019)
  29. DGM : A deep learning algorithm for solving partial differential equations (2018)
  30. Optimally weighted loss functions for solving PDE s with neural networks (2022)
  31. Markovian approximation of the rough B ergomi model for M onte C arlo option pricing (2021)

2 of 39 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.