• 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!

Calibration on heston Stochastic Volatility Model

Joined
9/18/14
Messages
4
Points
11
Does anyone know how and where to get option real market data to calibrate heston model?

I study this paper:
http://math.nyu.edu/~atm262/fall06/compmethods/a1/
nimalinmoodley.pdf

it only tells me that Vanilla calls on Anglo American shares, listed on LSE, was used as market data. I presumed LSE means London Stock Exchange, then i checked on its website http://www.londonstockexchange.com/ but couldn't find any option market data.

Moreover, if anyone knows papers about Heston's Calibration with step by step and detail procedure.. please, tell me. I really need it for my research.

Thanks A Lot,
Wish you all the best :)
 
Good option data is hard to come by. Ask your school.
 
ummm.. is that really that hard?
i really need it for simulation, but every paper that i have read didn't tell me how and where they get their market data..
 
Ask your school to see if they have access to any option databases. Yes it is much harder to come by than data for underlying, there's simply much more of it. If your school doesn't have it, you might need to make a connection with somebody at a financial institution.
 
Think Bloomberg, Wharton Research Data Service etc. No, it ain't free.

Incidentally, I think Moodley is an excellent introduction to calibration (and it includes fully functional Matlab code)!
 
Last edited:
Think Bloomberg, Wharton Research Data Service etc. No, it ain't free.

Incidentally, I think Moodley is an excellent introduction to calibration (and it includes fully functional Matlab code)!

Yup, i believe so. Moodley use Solver, lsqnonlin and Adaptive Simulated Annealing. They are good for beginners.

But, recently i found paper: papers.ssrn.com/sol3/papers.cfm?abstract_id=1566975&download=yes by Manfred Gilli. It use heuristic method such as Differential Evolution. This method have many advantages above all, so i decided to focus on it.
 
Yup, i believe so. Moodley use Solver, lsqnonlin and Adaptive Simulated Annealing. They are good for beginners.

But, recently i found paper: papers.ssrn.com/sol3/papers.cfm?abstract_id=1566975&download=yes by Manfred Gilli. It use heuristic method such as Differential Evolution. This method have many advantages above all, so i decided to focus on it.

Sadly, Fabrice Rouah passed away some months ago. RIP.
 
I found the Moodley thesis to be a good introduction to the Heston model and calibration. I wasn't able to reproduce the Calibration results on page 29 with the sample option data he provides on page 43. When I loaded that p.43 data into the calibration functions he provided in Matlab, I calculated different parameters- probably to be expected since Moodley mentions that the data used in calibration was only similar to that on page 43.

Since the option data wasn't available, what I did, though a bit crude, was download option data from google finance for a variety of expiries and strikes. For example for GOOG options, I visited:

https://www.google.com/finance/option_chain?q=NASDAQ:GOOG&ei=HFz0VOH6IMeCwgOA64CYDA

I then worked the option chain data into a text file format that could be read by Moodley's functions and ran the calibration on that data (requires you to add in columns representing interest rate and expiry). I was also able to confirm that I got the similar results pricing the options on a finite difference mesh (instead of the typical closed-form pricing method Moodley uses, via FFT), and then calibrating with ASA or a Nelder Mead non-linear optimizer.

The @kiramipt book is a great reference for the Heston model, and I believe Safari even allows you to view it free for a limited time:

https://www.safaribooksonline.com/library/view/the-heston-model/9781118548257/

If any of my finite difference code or optimization would help, I would be happy to share it.
 
Hello rmg, perhaps you should try the code from Dr Rouah book, it's another methodology: loss functions instead of nonlinear least squares...
 
Back
Top