Deep ReLU network expression rates for option prices in high-dimensional, exponential Lévy models

DOI10.1007/S00780-021-00462-7MaRDI QIDQ2238770zbMATH OpenOpenAlexWikidataFDO

Authors Lukas Gonon, Christoph Schwab

Publication date 2 November 2021

Published in Finance and Stochastics (Search for Journal in Brave)

Copyright license Creative Commons Attribution 4.0 International

Full work available at URL https://arxiv.org/abs/2101.11897

zbMATH Keywords

Barron spacecurse of dimensionalitydeep neural networkexpression rateLévy processoption pricingRademacher complexity

Mathematics Subject Classification ID

Integro-partial differential equations (45K05) Processes with independent increments; Lévy processes (60G51) Artificial neural networks and deep learning (68T07) Derivative securities (option pricing, hedging, etc.) (91G20) Financial applications of other theories (91G80)

Abstract: We study the expression rates of deep neural networks (DNNs for short) for option prices written on baskets of $d$ risky assets, whose log-returns are modelled by a multivariate L'evy process with general correlation structure of jumps. We establish sufficient conditions on the characteristic triplet of the L'evy process $X$ that ensure $varepsilon$ error of DNN expressed option prices with DNNs of size that grows polynomially with respect to $mathcalO(varepsilo{n}^{-1})$, and with constants implied in $mathcalO(cdot)$ which grow polynomially with respect $d$, thereby overcoming the curse of dimensionality and justifying the use of DNNs in financial modelling of large baskets in markets with jumps. In addition, we exploit parabolic smoothing of Kolmogorov partial integrodifferential equations for certain multivariate L'evy processes to present alternative architectures of ReLU DNNs that provide $varepsilon$ expression error in DNN size $mathcalO(|log(varepsilon){|}^a)$ with exponent $asimd$, however, with constants implied in $mathcalO(cdot)$ growing exponentially with respect to $d$. Under stronger, dimension-uniform non-degeneracy conditions on the L'evy symbol, we obtain algebraic expression rates of option prices in exponential L'evy models which are free from the curse of dimensionality. In this case the ReLU DNN expression rates of prices depend on certain sparsity conditions on the characteristic L'evy triplet. We indicate several consequences and possible extensions of the present results.

Review text

The authors study the expression rates of deep neural networks for option prices written on baskets of \(d\) risky assets whose log-returns are modelled by a multivariate Lévy process with general correlation structure of jumps. They establish sufficient conditions on the characteristic triplet of the Lévy process that ensure \(\varepsilon\)-error of option prices with deep neural networks of size that grows polynomially with respect to \(\mathcal{O}(\varepsilon^{-1})\), and with constants implied in \(\mathcal{O}(\cdot)\) which grow polynomially in \(d\), thereby overcoming the curse of dimensionality and justifying the use of deep neural networks in financial modelling of large baskets in markets with jumps.

Recommendations

• DNN expression rate analysis of high-dimensional PDEs: application to option pricing
• FFT-network for bivariate Lévy option pricing
• The deep parametric PDE method and applications to option pricing
• Deep neural network framework based on backward stochastic differential equations for pricing and hedging American options in high dimensions
• Pricing of high-dimensional American options by neural networks

Cites work

• 10.1162/153244303321897690
• A Feynman-Kac-type formula for Lévy processes with discontinuous killing rates
• A Finite Difference Scheme for Option Pricing in Jump Diffusion and Exponential Lévy Models
• A neural network-based policy iteration algorithm with global \(H^2\)-superlinear convergence for stochastic games on domains
• Analysis of the generalization error: empirical risk minimization over deep artificial neural networks overcomes the curse of dimensionality in the numerical approximation of Black-Scholes partial differential equations
• ANISOTROPIC STABLE LEVY COPULA PROCESSES — ANALYTICAL AND NUMERICAL ASPECTS
• Backward stochastic differential equations and Feynman-Kac formula for Lévy processes, with applications in finance
• Backward stochastic differential equations and integral-partial differential equations
• Barrier options and touch-and-out options under regular Lévy processes of exponential type
• Characterization of dependence of multidimensional Lévy processes using Lévy copulas
• Computational methods for quantitative finance. Finite element methods for derivative pricing
• Concentration inequalities. A nonasymptotic theory of independence
• Convergence rates of best \(N\)-term Galerkin approximations for a class of elliptic SPDEs
• Deep hedging
• Deep learning in high dimension: neural network expression rates for generalized polynomial chaos expansions in UQ
• Deep optimal stopping
• Deep ReLU networks and high-order finite element methods
• DNN expression rate analysis of high-dimensional PDEs: application to option pricing
• Exponential ReLU DNN expression of holomorphic maps in high dimension
• Fast deterministic pricing of options on Lévy driven assets
• Financial Modelling with Jump Processes
• Free boundary regularity in the parabolic fractional obstacle problem
• scientific article; zbMATH DE number 107823 (Why is no real title available?)
• scientific article; zbMATH DE number 481056 (Why is no real title available?)
• scientific article; zbMATH DE number 918811 (Why is no real title available?)
• scientific article; zbMATH DE number 1402217 (Why is no real title available?)
• scientific article; zbMATH DE number 3223856 (Why is no real title available?)
• Integro-differential equations for option prices in exponential Lévy models
• Lévy matters III. Lévy-type processes: construction, approximation and sample path properties
• Lévy Processes and Stochastic Calculus
• Mathematical finance
• Minimal \(f^q\)-Martingale measures for exponential Lévy processes
• Minimal entropy preserves the Lévy property: how and why
• Numerical methods for Lévy processes
• Optimal approximation of piecewise smooth functions using deep ReLU neural networks
• Probability in Banach spaces. Isoperimetry and processes
• Rectified deep neural networks overcome the curse of dimensionality for nonsmooth value functions in zero-sum games of nonlinear stiff systems
• Second-order elliptic integro-differential equations: viscosity solutions' theory revisited
• The critical price for the American put in an exponential Lévy model
• Uniform error estimates for artificial neural network approximations for heat equations
• Universal approximation bounds for superpositions of a sigmoidal function
• Variational solutions of the pricing PIDEs for European options in Lévy models

Cited in

(24)

• DNN expression rate analysis of high-dimensional PDEs: application to option pricing
• The deep parametric PDE method and applications to option pricing
• Solving non-linear Kolmogorov equations in large dimensions by using deep learning: a numerical comparison of discretization schemes
• Neural network stochastic differential equation models with applications to financial data forecasting
• Deep ReLU neural networks overcome the curse of dimensionality for partial integrodifferential equations
• Deep ReLU neural network approximation in Bochner spaces and applications to parametric PDEs
• The Kolmogorov infinite dimensional equation in a Hilbert space via deep learning methods
• Collocation approximation by deep neural ReLU networks for parametric and stochastic PDEs with lognormal inputs
• Pricing options on flow forwards by neural networks in a Hilbert space
• Applications of artificial neural networks to simulating Lévy processes
• Deep neural network expressivity for optimal stopping problems
• Approximation rates for deep calibration of (rough) stochastic volatility models
• Rectified deep neural networks overcome the curse of dimensionality when approximating solutions of McKean-Vlasov stochastic differential equations
• Detecting asset price bubbles using deep learning
• Strong convergence of the Euler scheme for a delayed jump-diffusion CIR model with Markovian switching
• Multilevel Picard approximations overcome the curse of dimensionality when approximating semilinear heat equations with gradient-dependent nonlinearities in L^p-sense
• Deep ReLU neural networks overcome the curse of dimensionality when approximating semilinear partial integro-differential equations
• Multiagent relative investment games in a jump diffusion market with deep reinforcement learning algorithm
• Multilevel Picard approximations overcome the curse of dimensionality in the numerical approximation of general semilinear PDEs with gradient-dependent nonlinearities
• Deep learning numerical methods for high-dimensional quasilinear PIDEs and coupled FBSDEs with jumps
• Full error analysis of the random deep splitting method for nonlinear parabolic PDEs and PIDEs
• Multilevel Picard algorithm for general semilinear parabolic PDEs with gradient-dependent nonlinearities
• Multilevel Picard approximation algorithm for semilinear partial integro-differential equations and its complexity analysis
• Designing universal causal deep learning models: the case of infinite-dimensional dynamical systems from stochastic analysis

This page was built for publication: Deep ReLU network expression rates for option prices in high-dimensional, exponential Lévy models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2238770)