tensorquant.pricers package

Submodules

tensorquant.pricers.black module

class tensorquant.pricers.black.BlackScholesPricer(dividend_model: str = 'continuous', use_implied_repo: bool = True)[source]

Bases: Pricer

calculate_price(product: VanillaOption, market_env: MarketEnvironment) → Tensor[source]

Price a VanillaOption using Black-Scholes (European) or Bjerksund-Stensland 1993 (American).

Delegates market-data extraction to _build_inputs(). Intermediate tf.Variable inputs are stored on self so that an external GradientTape (managed by Pricer.price()) can compute greeks.

Parameters:

  • product (VanillaOption) – The option to be priced.

  • market_env (MarketEnvironment) – The market environment providing access to market data (curves, spots, volatilities).

Returns:

The NPV of the option.

Return type:

tf.Tensor

tensorquant.pricers.black.bjerksund_stensland_calc(spot_price: Tensor, strike: Tensor, riskfree_rate: Tensor, volatility: Tensor, time_to_maturity: Tensor, dividend_yield: Tensor, option_type: OptionType) → Tensor[source]

American option price via the Bjerksund-Stensland (1993) approximation.

Handles calls directly and puts via put-call symmetry:

AmPut(S, K, r, b, σ, T) = AmCall(K, S, r-b, -b, σ, T)

Parameters:

  • spot_price – Spot (or spot-net for discrete-dividend model).

  • strike – Strike price.

  • riskfree_rate – Continuously compounded risk-free rate.

  • volatility – Implied volatility.

  • time_to_maturity – Time to expiry in year fractions.

  • dividend_yield – Continuous dividend yield q (0 for discrete model).

  • option_type – OptionType (Call or Put).

Returns:

American option price.

Return type:

tf.Tensor

tensorquant.pricers.black.blackscholes_calc(spot_price: Tensor, strike: Tensor, riskfree_rate: Tensor, volatility: Tensor, time_to_maturity: Tensor, dividend_yield: Tensor, option_type: OptionType) → Tensor[source]

European Black-Scholes price for a call or put.

Uses the unified φ formulation so that a single expression handles both option types without branching:

price = φ · [S·exp(-q·T)·N(φ·d1) - K·exp(-r·T)·N(φ·d2)]

where φ = +1 for calls and φ = -1 for puts.

Parameters:

  • spot_price – Spot (or dividend-adjusted spot for discrete model).

  • strike – Strike price.

  • riskfree_rate – Continuously compounded risk-free rate r.

  • volatility – Implied volatility σ.

  • time_to_maturity – Time to expiry in year fractions T.

  • dividend_yield – Continuous dividend yield q (repo included).

  • option_type – OptionType — Call (+1) or Put (-1).

Returns:

Option price.

Return type:

tf.Tensor

tensorquant.pricers.deposit module

class tensorquant.pricers.deposit.DepositPricer[source]

Bases: Pricer

calculate_price(product: Deposit, market_env: MarketEnvironment) → float[source]

Price a Deposit using discount curve.

Parameters:

  • product (Deposit) – The deposit to be priced.

  • market_env (MarketEnvironment) – The market environment providing access to market data (curves).

Returns:

The NPV of the deposit.

Return type:

float

Raises:

  • TypeError – If the product is not a Deposit.

  • ValueError – If the discount curve is not found.

tensorquant.pricers.factory module

class tensorquant.pricers.factory.PricerAssignment[source]

Bases: object

static create(product: Product)[source]

tensorquant.pricers.fixedflow module

Pricing di cash flow fissi

class tensorquant.pricers.fixedflow.FixedCouponDiscounting(coupon: FixedCoupon)[source]

Bases: object

calculate_price(discount_curve: RateCurve)[source]
class tensorquant.pricers.fixedflow.FixedLegDiscounting(leg: FixedRateLeg)[source]

Bases: object

calculate_price(discount_curve: RateCurve)[source]

tensorquant.pricers.floatingflow module

class tensorquant.pricers.floatingflow.FloatingCouponDiscounting(coupon: FloatingCoupon)[source]

Bases: object

amount(term_structure) → float[source]
calc_forward(ref_start, ref_end, term_structure)[source]
calculate_price(disc_curve: RateCurve, est_curve: RateCurve)[source]
floating_rate(start_date: date, end_date: date, term_structure: RateCurve)[source]
class tensorquant.pricers.floatingflow.FloatingLegDiscounting(leg: FloatingRateLeg)[source]

Bases: object

calculate_price(disc_curve, est_curve)[source]
class tensorquant.pricers.floatingflow.OisCouponDiscounting(coupon: FloatingCoupon)[source]

Bases: object

amount(term_structure: RateCurve) → float[source]
calculate_price(term_structure: RateCurve)[source]
floating_rate(start_date: date, end_date: date, term_structure: RateCurve)[source]
class tensorquant.pricers.floatingflow.OisLegDiscounting(leg: FloatingRateLeg)[source]

Bases: object

calculate_price(term_structure)[source]

tensorquant.pricers.fradiscounting module

class tensorquant.pricers.fradiscounting.FraPricer[source]

Bases: Pricer

calculate_price(product: Fra, market_env: MarketEnvironment) → float[source]

Price a Forward Rate Agreement (FRA).

Parameters:

  • product (Fra) – The FRA to be priced.

  • market_env (MarketEnvironment) – The market environment providing access to market data (curves).

Returns:

The NPV of the FRA.

Return type:

float

Raises:

  • TypeError – If the product is not a Fra.

  • ValueError – If the discount or forward curve is not found.

tensorquant.pricers.pricer module

class tensorquant.pricers.pricer.Pricer[source]

Bases: ABC

Abstract base class for pricing financial products.

This abstract class defines the interface for pricing financial products. Concrete implementations must provide a method for calculating the price of a product based on the trade date and market curves.

calculate_price()[source]

Abstract method to be implemented by subclasses to calculate the price of a product.

price()[source]

Calculates the price of a product and optionally returns the gradient if autodiff is enabled.

abstract calculate_price(product, market_env: MarketEnvironment)[source]

Abstract method to calculate the price of a financial product.

Parameters:

  • product (Product) – The financial product to be priced.

  • market_env (MarketEnvironment) – The market environment providing access to market data (curves, spots, volatilities).

Returns:

The calculated price of the product.

Return type:

float

Notes

This method must be implemented by any subclass of Pricer.

price(product: Product, market_env: MarketEnvironment, autodiff: bool = False)[source]

Calculates the price of a financial product, with optional automatic differentiation.

Parameters:

  • product (Product) – The financial product to be priced.

  • market_env (MarketEnvironment) – The market environment providing access to market data (curves, spots, volatilities).

  • autodiff (bool, optional) – Whether to compute gradients using TensorFlow’s autodiff. Defaults to False.

property tape

tensorquant.pricers.swapdiscounting module

class tensorquant.pricers.swapdiscounting.OisPricer[source]

Bases: Pricer

calculate_price(product: Ois, market_env: MarketEnvironment)[source]

Price an Overnight Indexed Swap (OIS).

Parameters:

  • product (Ois) – The OIS to be priced.

  • market_env (MarketEnvironment) – The market environment providing access to market data (curves).

Returns:

The NPV of the OIS.

Return type:

float

Raises:

  • TypeError – If the product is not an Ois.

  • ValueError – If the discount curve is not found.

class tensorquant.pricers.swapdiscounting.SwapPricer[source]

Bases: Pricer

calculate_price(product: Swap, market_env: MarketEnvironment)[source]

Price a plain vanilla interest-rate swap.

Parameters:

  • product (Swap) – The swap to be priced.

  • market_env (MarketEnvironment) – The market environment providing access to market data (curves).

Returns:

The NPV of the swap.

Return type:

float

Raises:

  • TypeError – If the product is not a Swap.

  • ValueError – If the discount or estimation curve is not found.

Module contents