pfhedge.nn¶

pfhedge.nn provides torch.nn.Module that are useful for Deep Hedging.

See PyTorch Documentation for general usage of torch.nn.Module.

Hedger Module¶

Module to hedge and price derivatives.

Black-Scholes Layers¶

`nn.BlackScholes`	Creates Black-Scholes formula module from a derivative.
`nn.BSAmericanBinaryOption`	Black-Scholes formula for an American binary option.
`nn.BSEuropeanOption`	Black-Scholes formula for a European option.
`nn.BSEuropeanBinaryOption`	Black-Scholes formula for a European binary option.
`nn.BSLookbackOption`	Black-Scholes formula for a lookback option with a fixed strike.

Whalley-Wilmott Layers¶

nn.WhalleyWilmott

Creates a module for Whalley-Wilmott's hedging strategy.

Nonlinear Activations¶

`nn.Clamp`	Clamp all elements in `input` into the range \([\min, \max]\).
`nn.LeakyClamp`	Leakily clamp all elements in `input` into the range \([\min, \max]\).

Loss Functions¶

`nn.HedgeLoss`	Base class for hedging criteria.
`nn.EntropicLoss`	Creates a criterion that measures the expected exponential utility.
`nn.EntropicRiskMeasure`	Creates a criterion that measures the entropic risk measure.
`nn.ExpectedShortfall`	Creates a criterion that measures the expected shortfall.
`nn.QuadraticCVaR`	Creates a criterion that measures the QuadraticCVaR.
`nn.IsoelasticLoss`	Creates a criterion that measures the expected isoelastic utility.

Multi Layer Perceptron¶

nn.MultiLayerPerceptron

Creates a multilayer perceptron.

Other Modules¶

`nn.Naked`	Returns a tensor filled with the scalar value zero.
`nn.SVIVariance`	Returns total variance in the SVI model.