Neural Options Lab: Deep Learning for Options Pricing

∂

Neural Options Lab Deep Pricing | Deep Hedging | XAI

Connecting...

Live Market

Contract

Spot S₀ 100

Strike K 100

Maturity T 1.00y

Model Inputs

Volatility σ 25%

Risk-free rate r 4.0%

Simulation

Moneyness S/K 1.00

Outside trained domain [0.5, 2.0] (Surrogate is extrapolating)

Discrete arithmetic-average Asian, 50 monitoring dates. MC uses antithetic sampling + geometric-Asian control variate. Greeks are exact autograd derivatives of the network.

Neural Net -

-

price + 5 Greeks in one autograd pass

Speedup -

-

Monte Carlo -

-

Δ Delta-

Γ Gamma-

ν Vega-

Θ Theta-

ρ Rho-

Monte Carlo Convergence

MC estimate ±95% CI vs path count (against the instant neural price)

Latency

wall-clock per pricing task (log scale)

Error Distribution: Single vs. Ensemble

signed pricing error vs high-precision Monte Carlo references

Feature Attributions

Integrated Gradients vs an ATM minimal-option baseline

Price Surface

Price / Strike over Moneyness and Maturity

Monte Carlo Convergence

Latency

Error Distribution: Single vs. Ensemble

Feature Attributions

Price Surface

Real-Time Neural Pricing

Hedging Simulator

Policy Behaviour

Risk Report