Abstract:Stochastic differential equation (SDE) models are the foundation for pricing and hedging financial derivatives. The drift and volatility functions in SDE models are typically chosen to be algebraic functions with a small number (less than 5) parameters which can be calibrated to market data. A more flexible approach is to use neural networks to model the drift and volatility functions, which provides more degrees-of-freedom to match observed market data. Training of models requires optimizing over an SDE, which is computationally challenging. For European options, we develop a fast stochastic gradient descent (SGD) algorithm for training the neural network-SDE model. Our SGD algorithm uses two independent SDE paths to obtain an unbiased estimate of the direction of steepest descent. For American options, we optimize over the corresponding Kolmogorov partial differential equation (PDE). The neural network appears as coefficient functions in the PDE. Models are trained on large datasets (many contracts), requiring either large simulations (many Monte Carlo samples for the stock price paths) or large numbers of PDEs (a PDE must be solved for each contract). Numerical results are presented for real market data including S&P 500 index options, S&P 100 index options, and single-stock American options. The neural-network-based SDE models are compared against the Black-Scholes model, the Dupire's local volatility model, and the Heston model. Models are evaluated in terms of how accurate they are at pricing out-of-sample financial derivatives, which is a core task in derivative pricing at financial institutions.

Simulation of Lévy-driven models and its application in finance

Simulation of Levy-Driven Models and Its Application in Finance

New Methods of Simulating Lévy Processes

Simulation of an Integro-Differential Equation and Application in Estimation of Ruin Probability with Mixed Fractional Brownian Motion

Fast and General Simulation of Lévy-driven OU processes for Energy Derivatives

An introduction to Lévy processes with applications in finance

Monte Carlo Simulation in Financial Engineering

Empirical Likelihood Methods Based on Characteristic Functions with Applications to Lévy Processes

Monte Carlo Simulation for Trading Under a Lévy-Driven Mean-Reverting Framework

Parameter Estimation for Geometric Lévy Processes with Constant Volatility

Probability-conservative simulation for Lévy financial model by a mixed finite element method

Monte Carlo Methods in Financial Engineering

Simulation of Implied Volatility Surfaces via Tangent Levy Models

Credit Derivatives Pricing Based on Lévy Field Driven Term Structure

Machine Learning Methods for Pricing Financial Derivatives

On Stochastic Partial Differential Equations and their applications to Derivative Pricing through a conditional Feynman-Kac formula

Lévy Processes For Finance: An Introduction In R

Structural credit risk model driven by Lévy process under knight uncertainty

Machine learning for quantitative finance: fast derivative pricing, hedging and fitting

Neural network stochastic differential equation models with applications to financial data forecasting

Exact simulation pricing with Gamma processes and their extensions