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.

Pricing equity-linked life insurance contracts with multiple risk factors by neural networks

Equity-Linked Life Insurances on Maximum of Several Assets

A neural network model for solvency calculations in life insurance

Pricing equity-linked guaranteed minimum death benefits with surrender risk by complex Fourier series expansion method

INSURANCE VALUATION: A TWO-STEP GENERALISED REGRESSION APPROACH

Reinsurance with neural networks

Pricing participating longevity-linked life annuities: a Bayesian Model Ensemble approach

Pricing and hedging derivative securities with neural networks: Bayesian regularization, early stopping, and bagging

Model risk for European-style stock index options

Pricing And Risk Management Of Variable Annuities And Equity-Indexed Annuities

Machine Learning Methods for Pricing Financial Derivatives

Optimizing Neural Networks for Bermudan Option Pricing: Convergence Acceleration, Future Exposure Evaluation and Interpolation in Counterparty Credit Risk

Robust financial calibration: a Bayesian approach for neural SDEs

Valuing equity-linked death benefits with a threshold expense strategy

Assessing asset-liability risk with neural networks

Pricing Options and Computing Implied Volatilities using Neural Networks

Managing Weather Risk with a Neural Network-Based Index Insurance

Neural network regression for Bermudan option pricing

Option Pricing by Neural Stochastic Differential Equations: A Simulation-Optimization Approach

Hedging of unit-linked life insurance contracts with unobservable mortality hazard rate via local risk-minimization

Addressing the economic and demographic complexity via a neural network approach: risk measures for reverse mortgages