# -*- coding: utf-8 -*-
"""
vollib.ref_python.black_scholes_merton.greeks.analytical
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
A library for option pricing, implied volatility, and
greek calculation. vollib is based on lets_be_rational,
a Python wrapper for LetsBeRational by Peter Jaeckel as
described below.
:copyright: © 2017 Gammon Capital LLC
:license: MIT, see LICENSE for more details.
vollib.ref_python is a pure python version of vollib without any dependence on LetsBeRational. It is provided purely as a reference implementation for sanity checking. It is not recommended for industrial use.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
"""
# -----------------------------------------------------------------------------
# IMPORTS
# Standard library imports
# Related third party imports
import numpy
from scipy.stats import norm
# Local application/library specific imports
from vollib.helpers import pdf
from vollib.ref_python.black_scholes_merton import d1, d2
N = norm.cdf
# -----------------------------------------------------------------------------
# FUNCTIONS - ANALYTICAL GREEKS
[docs]
def delta(flag, S, K, t, r, sigma, q):
"""Returns the Black-Scholes-Merton delta of an option.
:param flag: 'c' or 'p' for call or put.
:type flag: str
:param S: underlying asset price
:type S: float
:param K: strike price
:type K: float
:param t: time to expiration in years
:type t: float
:param r: annual risk-free interest rate
:type r: float
:param sigma: volatility
:type sigma: float
:param q: annualized continuous dividend yield
:type q: float
:returns: float
Example 17.1, page 355, Hull:
>>> S = 49
>>> K = 50
>>> r = .05
>>> t = 0.3846
>>> q = 0
>>> sigma = 0.2
>>> flag = 'c'
>>> delta_calc = delta(flag, S, K, t, r, sigma, q)
>>> # 0.521601633972
>>> delta_text_book = 0.522
>>> abs(delta_calc - delta_text_book) < .01
True
"""
D1 = d1(S, K, t, r, sigma, q)
if flag == 'p':
return -numpy.exp(-q * t) * N(-D1)
else:
return numpy.exp(-q * t) * N(D1)
[docs]
def theta(flag, S, K, t, r, sigma, q):
"""Returns the Black-Scholes-Merton theta of an option.
:param flag: 'c' or 'p' for call or put.
:type flag: str
:param S: underlying asset price
:type S: float
:param K: strike price
:type K: float
:param t: time to expiration in years
:type t: float
:param r: annual risk-free interest rate
:type r: float
:param sigma: volatility
:type sigma: float
:param q: annualized continuous dividend yield
:type q: float
:returns: float
The text book analytical formula does not divide by 365,
but in practice theta is defined as the change in price
for each day change in t, hence we divide by 365.
Example 17.2, page 359, Hull:
>>> S = 49
>>> K = 50
>>> r = .05
>>> t = 0.3846
>>> q = 0
>>> sigma = 0.2
>>> flag = 'c'
>>> annual_theta_calc = theta(flag, S, K, t, r, sigma, q) * 365
>>> # -4.30538996455
>>> annual_theta_text_book = -4.31
>>> abs(annual_theta_calc - annual_theta_text_book) < .01
True
Using the same inputs with a put.
>>> S = 49
>>> K = 50
>>> r = .05
>>> t = 0.3846
>>> sigma = 0.2
>>> flag = 'p'
>>> annual_theta_calc = theta(flag, S, K, t, r, sigma, q) * 365
>>> # -1.8530056722
>>> annual_theta_reference = -1.8530056722
>>> abs(annual_theta_calc - annual_theta_reference) < .000001
True
"""
D1 = d1(S, K, t, r, sigma, q)
D2 = d2(S, K, t, r, sigma, q)
first_term = (S * numpy.exp(-q * t) * pdf(D1) * sigma) / (2 * numpy.sqrt(t))
if flag == 'c':
second_term = -q * S * numpy.exp(-q * t) * N(D1)
third_term = r * K * numpy.exp(-r * t) * N(D2)
return - (first_term + second_term + third_term) / 365.0
else:
second_term = -q * S * numpy.exp(-q * t) * N(-D1)
third_term = r * K * numpy.exp(-r * t) * N(-D2)
return (-first_term + second_term + third_term) / 365.0
[docs]
def gamma(flag, S, K, t, r, sigma, q):
"""Returns the Black-Scholes-Merton gamma of an option.
:param flag: 'c' or 'p' for call or put.
:type flag: str
:param S: underlying asset price
:type S: float
:param K: strike price
:type K: float
:param t: time to expiration in years
:type t: float
:param r: annual risk-free interest rate
:type r: float
:param sigma: volatility
:type sigma: float
:param q: annualized continuous dividend yield
:type q: float
:returns: float
Example 17.4, page 364, Hull:
>>> S = 49
>>> K = 50
>>> r = .05
>>> t = 0.3846
>>> q = 0
>>> sigma = 0.2
>>> flag = 'c'
>>> gamma_calc = gamma(flag, S, K, t, r, sigma, q)
>>> # 0.0655453772525
>>> gamma_text_book = 0.066
>>> abs(gamma_calc - gamma_text_book) < .001
True
"""
D1 = d1(S, K, t, r, sigma, q)
numerator = numpy.exp(-q * t) * pdf(D1)
denominator = S * sigma * numpy.sqrt(t)
return numerator / denominator
[docs]
def vega(flag, S, K, t, r, sigma, q):
"""Returns the Black-Scholes-Merton vega of an option.
:param flag: 'c' or 'p' for call or put.
:type flag: str
:param S: underlying asset price
:type S: float
:param K: strike price
:type K: float
:param t: time to expiration in years
:type t: float
:param r: annual risk-free interest rate
:type r: float
:param sigma: volatility
:type sigma: float
:param q: annualized continuous dividend yield
:type q: float
:returns: float
The text book analytical formula does not multiply by .01,
but in practice vega is defined as the change in price
for each 1 percent change in IV, hence we multiply by 0.01.
Example 17.6, page 367, Hull:
>>> S = 49
>>> K = 50
>>> r = .05
>>> t = 0.3846
>>> q = 0
>>> sigma = 0.2
>>> flag = 'c'
>>> vega_calc = vega(flag, S, K, t, r, sigma, q)
>>> # 0.121052427542
>>> vega_text_book = 0.121
>>> abs(vega_calc - vega_text_book) < .01
True
"""
D1 = d1(S, K, t, r, sigma, q)
return S * numpy.exp(-q * t) * pdf(D1) * numpy.sqrt(t) * 0.01
[docs]
def rho(flag, S, K, t, r, sigma, q):
"""Returns the Black-Scholes-Merton rho of an option.
:param flag: 'c' or 'p' for call or put.
:type flag: str
:param S: underlying asset price
:type S: float
:param K: strike price
:type K: float
:param t: time to expiration in years
:type t: float
:param r: annual risk-free interest rate
:type r: float
:param sigma: volatility
:type sigma: float
:param q: annualized continuous dividend yield
:type q: float
:returns: float
The text book analytical formula does not multiply by .01,
but in practice rho is defined as the change in price
for each 1 percent change in r, hence we multiply by 0.01.
Example 17.7, page 368, Hull:
>>> S = 49
>>> K = 50
>>> r = .05
>>> t = 0.3846
>>> q = 0
>>> sigma = 0.2
>>> flag = 'c'
>>> rho_calc = rho(flag, S, K, t, r, sigma, q)
>>> # 0.089065740988
>>> rho_text_book = 0.0891
>>> abs(rho_calc - rho_text_book) < .0001
True
"""
D2 = d2(S, K, t, r, sigma, q)
if flag == 'c':
return t * K * numpy.exp(-r * t) * N(D2) * .01
else:
return -t * K * numpy.exp(-r * t) * N(-D2) * .01
if __name__ == "__main__":
from vollib.helpers.doctest_helper import run_doctest
run_doctest()