• 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!

Another Calibration Understanding Problem in Heston FX Model

Joined
2/9/15
Messages
15
Points
13
Hey,

I've got an understanding problem with the calibration procedure. I want to price with the Heston Model some FX options and I dont understand it. I've looked up a lot of papers regarding the Heston Model and my question is:
A lot of papers use a loss-function to calibrate the model to market prices. The formula looks kind this way

min sqr( ∑ w ( CModel - CMarket)^2)

w=weights Cmodel=Model implied prices CMarket=Market obeserved Prices

The point i dont get is, if i want to calibrate the model to market prices where the hell do i get the model implied prices from??? I mean I need to calibrate the model to calculate model prices or am I wrong?
Or do I need to calculate with the Vola Surface some Black Scholes prices and put them here in? This cant be right.

I hope anyone can help me out :)

regards
boulala
 
Thats all? I guess some initial parameters, put them into heston and calculate some calls?
And this values I use to put them into the calibration scheme?
After that I can calculate the calls with my estimated parameters?
 
if u trust ur optimization algorithm (particle swarm), then yes
otherwise u have to start somewhere close

also if u trust ur optimization and believe the params will somehow sorta converge to a region, u can obtain a time series of market prices and do the following:
  1. make up some initial value first then run optimization on the market prices at t_0
  2. use the results as initial params to get params for t_1
  3. continue the process till t_n and hopefully now the params are (kinda) stable and u can simply recycle yesterday's params as initial input to compute today's params
 
Thanks Into Darkness. I will try and hopefully get a good result.
 
Back
Top