kiyosi

外观

<kiyosi/pricing/implied.hpp> ​

cpp
#include <kiyosi/pricing/implied.hpp>

implied_volatility ​

cpp
template <typename Engine, typename Option>
Result<double> kiyosi::implied_volatility(const Engine &engine, const Option &option, const PricingContext &context, double observed_price, ImpliedVolatilitySettings settings={})

Uses safeguarded Brent interpolation in volatility; the caller's bounds must bracket the quote. For CRR trees, the interval is intersected with the tree's admissible volatility range. 模板参数

  • Engine — Pricing engine providing price(option, context).
  • Option — Instrument accepted by the engine.

参数

  • engine — Pricing engine used for each trial volatility.
  • option — Instrument to value.
  • context — Market state whose volatility is replaced for each trial.
  • observed_price — Finite market price to match.
  • settings — Positive bounds and convergence controls. Built-in Monte Carlo trials share one seed per solve when the engine has no explicit seed; tolerance applies to that sampled price curve and does not bound sampling error. CPU European Monte Carlo trials reuse per-solve samples with storage linear in path count. CUDA accumulator and autocallable trials reuse a workspace local to each solve. A root is not guaranteed unique; known states with volatility-independent remaining cashflows, including fixed touch payments, are rejected. Built-in autocallable Monte Carlo engines reject fixed cashflows after input validation, before executing a simulation backend.

返回值: Implied volatility, or a validation, bracketing, pricing, or convergence error.

implied_volatility< AnalyticBarrierEngine, BarrierOption > ​

cpp
template <>
Result<double> kiyosi::implied_volatility<AnalyticBarrierEngine, BarrierOption>(const AnalyticBarrierEngine &engine, const BarrierOption &option, const PricingContext &context, double observed_price, ImpliedVolatilitySettings settings)

Scheduled analytic barriers validate their calendar once per solve.

implied_volatility< AnalyticBinaryBarrierEngine, BinaryBarrierOption > ​

cpp
template <>
Result<double> kiyosi::implied_volatility<AnalyticBinaryBarrierEngine, BinaryBarrierOption>(const AnalyticBinaryBarrierEngine &engine, const BinaryBarrierOption &option, const PricingContext &context, double observed_price, ImpliedVolatilitySettings settings)

implied_volatility< AnalyticBinaryBarrierEngine, TouchOption > ​

cpp
template <>
Result<double> kiyosi::implied_volatility<AnalyticBinaryBarrierEngine, TouchOption>(const AnalyticBinaryBarrierEngine &engine, const TouchOption &option, const PricingContext &context, double observed_price, ImpliedVolatilitySettings settings)

implied_volatility< MonteCarloVanillaEngine, EuropeanOption > ​

cpp
template <>
Result<double> kiyosi::implied_volatility<MonteCarloVanillaEngine, EuropeanOption>(const MonteCarloVanillaEngine &engine, const EuropeanOption &option, const PricingContext &context, double observed_price, ImpliedVolatilitySettings settings)

implied_volatility< MonteCarloVanillaEngine, AmericanOption > ​

cpp
template <>
Result<double> kiyosi::implied_volatility<MonteCarloVanillaEngine, AmericanOption>(const MonteCarloVanillaEngine &engine, const AmericanOption &option, const PricingContext &context, double observed_price, ImpliedVolatilitySettings settings)

Reuses request-local exercise buffers across trials; growth factors are recomputed.

implied_volatility ​

cpp
template <AutocallableProduct Note>
Result<double> kiyosi::implied_volatility(const FiniteDifferenceAutocallableEngine &engine, const Note &option, const PricingContext &context, double observed_price, ImpliedVolatilitySettings settings={})

Finite-difference autocallables validate immutable contracts and calendars once per solve.

implied_volatility< MonteCarloAccumulatorEngine, Accumulator > ​

cpp
template <>
Result<double> kiyosi::implied_volatility<MonteCarloAccumulatorEngine, Accumulator>(const MonteCarloAccumulatorEngine &engine, const Accumulator &option, const PricingContext &context, double observed_price, ImpliedVolatilitySettings settings)

Reuses CUDA buffers within one solve without sharing storage between requests.

implied_volatility ​

cpp
template <AutocallableProduct Note>
Result<double> kiyosi::implied_volatility(const MonteCarloAutocallableEngine &engine, const Note &option, const PricingContext &context, double observed_price, ImpliedVolatilitySettings settings={})

Reuses CUDA buffers within one solve without sharing storage between requests.

implied_coupon ​

cpp
template <typename Engine, typename Option>
Result<double> kiyosi::implied_coupon(const Engine &engine, const Option &option, const PricingContext &context, double observed_price, CouponQuoteConvention convention, ImpliedCouponSettings settings={})

Solves a Snowball engine's price curve in its knock-out coupon. 参数

  • engine — Pricing engine used for each trial coupon.
  • option — Snowball-family instrument to value.
  • context — Market state used for every trial.
  • observed_price — Finite market price to match.
  • convention — Whether the maturity coupon shifts with the quoted coupon.
  • settings — Finite coupon bounds and convergence controls. Built-in Monte Carlo trials share one seed per solve when the engine has no explicit seed; tolerance applies to that sampled price curve and does not bound sampling error. Built-in Monte Carlo engines share paths between coupon bounds. Built-in Monte Carlo and FD engines try interpolation, verify its price residual, then bisect if needed. States with no quoted coupon exposure on the engine's sampled price curve are rejected.

返回值: Implied coupon, or a validation, bracketing, pricing, or convergence error.

implied_coupon ​

cpp
template <typename Engine, typename Option>
Result<double> kiyosi::implied_coupon(const Engine &engine, const Option &option, const PricingContext &context, double observed_price, ImpliedCouponSettings settings={})

Solves the engine's price curve in an unambiguous product coupon, such as a Phoenix coupon. Built-in Monte Carlo trials share one seed per solve when the engine has no explicit seed; tolerance applies to that sampled price curve and does not bound sampling error. Built-in Monte Carlo engines share paths between coupon bounds. Built-in Monte Carlo and FD engines try interpolation, verify its price residual, then bisect if needed. States with no quoted coupon exposure on the engine's sampled price curve are rejected. 返回值: Implied coupon, or a validation, bracketing, pricing, or convergence error.

implied_coupon ​

cpp
template <AutocallableProduct Note>
Result<double> kiyosi::implied_coupon(const FiniteDifferenceAutocallableEngine &engine, const Note &option, const PricingContext &context, double observed_price, CouponQuoteConvention convention, ImpliedCouponSettings settings={})

Reuse grids, buffers and compatible factors within one solve; requests share no storage.

implied_coupon ​

cpp
Result<double> kiyosi::implied_coupon(const FiniteDifferenceAutocallableEngine &engine, const PhoenixOption &option, const PricingContext &context, double observed_price, ImpliedCouponSettings settings={})