Class AmericanPathPricer

Namespace

NanoQL.PricingEngines

Assembly

NanoQL.dll

The American early-exercise path pricer: the payoff of the underlying at a time step, the state the underlying scaled by the strike, and a polynomial basis of the state with the payoff as one more function.

public sealed class AmericanPathPricer : IEarlyExercisePathPricer<MonteCarloPath, double>

Inheritance

object

AmericanPathPricer

Implements

IEarlyExercisePathPricer<MonteCarloPath, double>

Inherited Members

object.GetType()

object.ToString()

object.Equals(object)

object.Equals(object, object)

object.ReferenceEquals(object, object)

object.GetHashCode()

Remarks

Immutable. The state is the underlying divided by the strike where the payoff is striked, for the regression's numerical stability. The Legendre and first-kind Chebyshev bases are refused, as QuantLib refuses them; the second-kind Chebyshev basis is NaN above 1, so a call's regression, on its in-the-money paths, is refused by the singular value decomposition.

Constructors

AmericanPathPricer(Payoff, int, LsmBasisSystemPolynomialType)

Initializes a new instance of the AmericanPathPricer class.

public AmericanPathPricer(Payoff payoff, int polynomialOrder, LsmBasisSystemPolynomialType polynomialType)

Parameters

payoff Payoff

The payoff.

polynomialOrder int

The basis' highest order.

polynomialType LsmBasisSystemPolynomialType

The basis' family: monomial, Laguerre, Hermite, hyperbolic or second-kind Chebyshev.

Exceptions

ArgumentNullException

payoff is null.

ArgumentOutOfRangeException

polynomialOrder is negative.

ArgumentException

The family is Legendre or first-kind Chebyshev ("insufficient polynomial type").

NanoQLException

The family is unknown.

Methods

BasisSystem()

The basis: the polynomials of the state, then the payoff of the state.

public ImmutableArray<Func<double, double>> BasisSystem()

Returns

ImmutableArray<Func<double, double>>

The basis functions.

Evaluate(MonteCarloPath, int)

The payoff of the underlying at a time step.

public double Evaluate(MonteCarloPath path, int t)

Parameters

path MonteCarloPath

The path.

t int

The time step's index.

Returns

double

The exercise value.

Exceptions

ArgumentOutOfRangeException

t isn't a point of the path, where QuantLib reads past its end.

State(MonteCarloPath, int)

The scaled underlying at a time step.

public double State(MonteCarloPath path, int t)

Parameters

path MonteCarloPath

The path.

t int

The time step's index.

Returns

double

The state.

Exceptions

ArgumentOutOfRangeException

t isn't a point of the path, where QuantLib reads past its end.