Pricing.PV.PricingParameters

Classes

PricingParameters(n: int, l: int, m: int, ...) Class that contains the numerical parameters to be passed to the american pricer
class Pricing.PV.PricingParameters.PricingParameters(n: int, l: int, m: int, p: int, first_guess_method: <module 'Pricing.PV.FirstGuessMethod' from '/home/docs/checkouts/readthedocs.org/user_builds/americanpricer/checkouts/latest/Lib/Pricing/PV/FirstGuessMethod.py'>, iteration_type: <module 'Pricing.PV.IterationType' from '/home/docs/checkouts/readthedocs.org/user_builds/americanpricer/checkouts/latest/Lib/Pricing/PV/IterationType.py'>)[source]

Class that contains the numerical parameters to be passed to the american pricer

Parameters:

n : int

Number of points in the time-discretization of the boundary.

l : int

Number of points for Gauss-Legendre quadrature for 1D integration inside an iteration of fixed-point algorithm.

m : int

Number of iterations in the fixed point algorithm.

p : int

Number of points for Gauss-Legendre quadrature for 1D integration for the final price when the boundary is computed.

first_guess_method : FirstGuessMethod

The algorithm used to choose a first guess of the exercice boundary.

Values can be :member:`~PV.Pricing.FirstGuessMethod.Trivial` or :member:`~PV.Pricing.FirstGuessMethod.Smart`.

  • Trivial uses a flat-firest guess corresponding to K * Max(1,r/q)
  • Smart uses a solver in order to have a good first guess (slows down the pricing time, so use if you want

a more precise result)

iteration_type : IterationType

The iterative algorithm that is used in order to solve the fixed point problem \(f(B)=B\)

Can be “Richardson”, “Partial_Newton_Jacobi” or “Newton_Jacobi” to solve the fixed point problem f(B) = B

  • Richardson uses a naive fixed-point iteration
  • Newton_Jacobi mixes Richardson iterations with the Jacobi-Newton way to solve \(g(B) = 0\) where

\(g(B) = f(B) - B\) * Partial_Newton_Jacobi is like Newton_Jacobi but the derivative expression is approximated by putting at 0 smaller terms (faster than Newton_Jacobi)

Attributes

first_guess_method
iteration_type
l
m
n
p