american-options 0.1.4

Added new payoffs

American Options Library

The Library that is using various methods to price the variety of options of american type.

Installation

pip install american_options

Get started

How to obtain option price with this library:

from american_options import Option, Underlying



#Set up your parameters

n_sims = 10000   

T = 1

sigma1 = 0.4

sigma2 = 0.8

r = 0.06

N = 255

spot1 = 100

spot2 = 120

K1 = 100

K2 = 110

First you need to create assets you want to write options on. You can do this in following way:

# Instantiate an Underlying object

asset1 = Underlying(spot_price=100, r=0.07)

asset2 = Underlying(spot_price=120, r=0.07)

Next step is to calibrate assets for the use of different models.

# Calibrate created assets

asset1.calibrate_GBM(sigma=0.4, values_per_year_GBM=N)

asset1.calibrate_JD(sigma=0.4, jump_intensity=1/3)

asset2.calibrate_GBM(sigma=0.4, values_per_year_GBM=N)

asset2.calibrate_JD(sigma=0.2, jump_intensity=1/2)

Now create options, for that we need payoff functions as well. You can import some of them from payoffs module:

from american_options.payoffs import *



option1 = Option(underlyings=asset1, payoff_func=lambda trajectory: payoff_creator_1d(trajectory, Call_Payoff, K=100, barrier=True, barrier_level=140), T=1)

option2 = Option(underlyings=(asset1, asset2), payoff_func=lambda trajectories: double_max_put(trajectories, 100), T=1)

And finally obtaining a price!

# Pricing with state space partitioning method

option1.ssp(n_sims=10000, mode='JD')

option2.ssp(n_sims=10000, mode='GBM')



# pPricing with Longstaff-Schwartz method

option1.LS(n_sims=10000, mode='JD')

option2.ssp(n_sims=10000, mode='GBM')