• C++ Programming for Financial Engineering
    Highly recommended by thousands of MFE students. Covers essential C++ topics with applications to financial engineering. Learn more Join!
    Python for Finance with Intro to Data Science
    Gain practical understanding of Python to read, understand, and write professional Python code for your first day on the job. Learn more Join!
    An Intuition-Based Options Primer for FE
    Ideal for entry level positions interviews and graduate studies, specializing in options trading arbitrage and options valuation models. Learn more Join!

Simulation Heston Model

Joined
3/18/08
Messages
1
Points
11
Hello.
I need you help about the simulation of Heston Model.
I need to simulate this model using a Euler discetization scheme.
My problemes are :
Which values choose for the parameters. I found estimation on papers but i don't know if it's daily or annuals estimation ....
V0 = 0.1, k=2, theta = 0.01, sigma = 0.1, rho = 0.3, mu = ? 0.03 ?
I used a step of 0.02 = 1/50

I need to simulate daily assets values. I use the R code (also test in C++) below :

Heston<- function(mu, kapa, theta, sigma, rho)
{
v<-rep(0,100000)
x<-rep(0,100000)
v[1]<-0.01
x[1]<-5000
for (i in 2:100000)
{
normale1<-rnorm(1,0,1)
normale2<-rnorm(1,0,1)
normale3<-rho*normale1 + sqrt(1-rho*rho)*normale2
x<-x[i-1]*(1 + (mu/50) + sqrt(v[i-1]/50)*normale1)
v<-abs((kapa*theta/50) + (1-(kapa/50))*v[i-1] + sigma*sqrt(v[i-1]/50)*normale3)
}
x
}

Heston(0.0001,2,0.01,0.1,-0.3)

I also try to simulate returns (using ito lemme) but resultts were also very bad.

Fabien
 
for (i in 2:100000)
{
normale1<-rnorm(1,0,1)
normale2<-rnorm(1,0,1)
normale3<-rho*normale1 + sqrt(1-rho*rho)*normale2
x<-x[i-1]*(1 + (mu/50) + sqrt(v[i-1]/50)*normale1)
v<-abs((kapa*theta/50) + (1-(kapa/50))*v[i-1] + sigma*sqrt(v[i-1]/50)*normale3)
}

I also try to simulate returns (using ito lemme) but resultts were also very bad.

Fabien


Hallo Fabien,

why do you want to simulate an Heston-Price?
Do the closed-form-estimation!
But in answer of your question:
You need 2 "for-loops", one in cause of the time-discretization, and the other one to simulate Monte Carlo.
In your code you should get some results, that differs from time to time.


Bastian
 
It's very famous that Euler (and even Milstein) doesn't converge for Heston. You should try Andersen from BofA or Smith from Santander.
 
Hi,
I need to simulate an heston model too.. if i can have your help??
i will simulate the equations on the article "closed-form solution for options with stochastic volatility with applications to bond and currency options" by Heston.. page328/331
thanks a lot!!
 
Back
Top