11
$\begingroup$

What is the formula for the vanilla option (Call/Put) price in the Heston model?

I only found the bi-variate system of stochastic differential equations of Heston model but no expression for the option prices.

$\endgroup$
0

2 Answers 2

8
$\begingroup$

In the Heston Model we have \begin{align} C(t\,,{{S}_{t}},{{v}_{t}},K,T)={{S}_{t}}{{P}_{1}}-K\,{{e}^{-r\tau }}{{P}_{2}} \end{align} where, for $j=1,2$

\begin{align} & {{P}_{j}}({{x}_{t}}\,,\,{{v}_{t}}\,;\,\,{{x}_{T}},\ln K)=\frac{1}{2}+\frac{1}{\pi }\int\limits_{0}^{\infty }{\operatorname{Re}\left( \frac{{{e}^{-i\phi \ln K}}{{f}_{j}}(\phi ;t,x,v)}{i\phi } \right)}\,d\phi \\ & {{f}_{j}}(\phi \,;{{v}_{t}},{{x}_{t}})=\exp [{{C}_{j}}(\tau ,\phi )+{{D}_{j}}(\tau ,\phi ){{v}_{t}}+i\phi {{x}_{t}}] \\ \end{align}

and

\begin{align} & {{C}_{j}}(\tau ,\phi )=(r-q)i\phi \,\tau +\frac{a}{{{\sigma }^{2}}}{{\left( ({{b}_{j}}-\rho \sigma i\phi +{{d}_{j}})\,\tau -2\ln \frac{1-{{g}_{j}}{{e}^{{{d}_{j}}\tau }}}{1-{{g}_{j}}} \right)}_{_{_{_{{}}}}}} \\ & {{D}_{j}}(\tau ,\phi )=\frac{{{b}_{j}}-\rho \sigma i\phi +{{d}_{j}}}{{{\sigma }^{2}}}\left( \frac{1-{{e}^{{{d}_{j}}\tau }}}{1-{{g}_{j}}{{e}^{{{d}_{j}}\tau }}} \right) \\ \end{align}

where \begin{align} & {{g}_{j}}=\frac{{{b}_{j}}-\rho \sigma i\phi +{{d}_{j}}}{{{b}_{j}}-\rho \sigma i\phi -{{d}_{j}}} \\ & {{d}_{j}}=\sqrt{{{({{b}_{j}}-\rho \sigma i\phi )}^{2}}-{{\sigma }^{2}}(2i{{u}_{j}}\phi -{{\phi }^{2}})} \\ & {{u}_{1}}=\frac{1}{2}\,,\,{{u}_{2}}=-\frac{1}{2}\,,\,a=\kappa \theta \,,\,{{b}_{1}}=\kappa +\lambda -\rho \sigma \,,\,{{b}_{2}}=\kappa +\lambda \,,\ {{i}^{2}}=-1 \\ \end{align} Other representations:

  1. Carr-Madan (1999)
  2. Lewis (2000)
  3. Attari (2004)
  4. Gatheral (2006)
  5. Albercher (2007)
$\endgroup$
4
  • 1
    $\begingroup$ For numerical estimation, the "Re" part of the integral would by approximated by FFT? $\endgroup$
    – emcor
    Jul 5, 2015 at 11:22
  • 1
    $\begingroup$ @emcor Yes,the fast Fourier transform was applied by Carr and Madan (1999) to speed up the computation of option prices $\endgroup$
    – user16651
    Jul 5, 2015 at 12:36
  • 1
    $\begingroup$ Thx I posted a follow-up question on the Bates SVJ model here. $\endgroup$
    – emcor
    Aug 5, 2015 at 12:47
  • $\begingroup$ What is $x$ in these equations? $\endgroup$
    – EpicAdv
    Dec 20, 2019 at 13:45
1
$\begingroup$

The Heston model is given by the following equations in a risk-neutral measure: $$ dS_t = (r - q)S_t dt + \sqrt{V_t} S_t dW_t, \\ dV_t = \kappa(\theta - V_t)dt + \sigma_V \sqrt V_t dZ_t,\\ dS_t dW_t = \rho dt. $$

The call formula from (Lipton, 2002) is as follows:

$$ C(K, T) = S_0 e^{-qT} - \frac{Ke^{-rT}}{\pi} \int\limits_0^\infty \mathfrak{Re}\left[ e^{(iu+\frac 1 2)k} \phi_T(u - \tfrac i 2)\right] \frac{du}{u^2 + \tfrac 1 4}, \\ k = \ln(S_0 / K) + (r-q)T, $$ where $\phi_T$ is the characteristic function. For Heston model: $$ \phi_T(u) = \exp[A(u) + B(u)V_0],\\ A(u) = \frac{\kappa \theta}{\sigma_V^2} \left[(\beta-d)T - 2\ln \left(\frac{ge^{-dT} - 1}{g-1} \right) \right],\\ B(u) = \frac{\beta-d}{\sigma_V^2}\left( \frac{1-e^{-dT}}{1-ge^{-dT}} \right), \\ d = \sqrt{\beta^2 - 4\hat \alpha \gamma},\\ g = \frac{\beta-d}{\beta+d},\\ \hat\alpha = -\frac 1 2 u(u+i), \quad \beta = \kappa - iu\sigma_v \rho, \quad \gamma = \frac 1 2 \sigma_V^2 $$ The code in Python is below:

import numpy as np
import scipy.integrate

r = 0.02
q = 0.01
s = 100.0
v = 0.04
kappa = 1.0
theta = 0.04
sigma_v = 0.2
rho = -0.7
k = 100.0
tau = 1.0

def phi(u, tau):
    alpha_hat = -0.5 * u * (u + 1j)
    beta = kappa - 1j * u * sigma_v * rho
    gamma = 0.5 * sigma_v ** 2
    d = np.sqrt(beta**2 - 4 * alpha_hat * gamma)
    g = (beta - d) / (beta + d)
    h = np.exp(-d*tau)
    A_ = (beta-d)*tau - 2*np.log((g*h-1) / (g-1))
    A = kappa * theta / (sigma_v**2) * A_
    B = (beta - d) / (sigma_v**2) * (1 - h) / (1 - g*h)
    return np.exp(A + B * v)

def integral(k, tau):
    integrand = (lambda u: 
        np.real(np.exp((1j*u + 0.5)*k)*phi(u - 0.5j, tau))/(u**2 + 0.25))

    i, err = scipy.integrate.quad_vec(integrand, 0, np.inf)
    return i

def call(k, tau):
    a = np.log(s/k) + (r-q)*tau
    i = integral(a, tau)        
    return s * np.exp(-q*tau) - k * np.exp(-r*tau)/np.pi*i
    
call(k, tau)
  • (Lipton, 2002): The Vol Smile Problem, Risk
$\endgroup$
1
  • $\begingroup$ Thanks, but can you point out the differences to the formula in the first answer here? $\endgroup$
    – emcor
    Feb 28 at 18:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.