Hi Does anyone have insight into an appropriate range for kappa (rate of mean reversion) in the heston model - i dont think i've ever seen a paper with an example of kappa less than 1 - are values such as 0.21 realistic??Thanks

Hard to say, IMO, given only that. What's the security, what's the option data (qualitatively), and what are the other parameter estimates?regards,

I've finished my master thesis about calibration of the heston model and so on. A value of \kappa after the calibration is realistic, depending on which error-measure is used, i mean the APE,AAE, ARPE or MSE. According to my results i found, that a a value of \kappa must be near or greater than 2 for the first 3 measures (APE,AAE,ARPE). BUT the result which came out for the MSE was woundering. I have: \kappa=0.1837. I have test the goodness of the fit of the model prices to the markt prices with a set of parameter,which i got from the MSE-calibration,of course including this \kappa=0.1837, and according to the Figure i will say, that this result is absolutly acceptabel. Thus do not doubt in ur results first, when u haven't test its goodness.

TO shamimafshani and Alan: i'm looking for the S&P 500 index option prices,say with a amount of about 50-80 on the next trading day after the 18.04.2002. Since the S&P 500 index options are traded on each the 3rd. friday in a trading month, is the next trading day on the 16.05.2002?? looking forward...and Best Regards!

Hi AlanI realize its a vague question - i'm trying to prove a result which i think i can show will hold for rho<0 and kappa>sqrt(0.5*exp(-pi))=0.147 so wanted to get an idea of whether the restriction is heavy (for equity in general). Maybe i should rephrase by asking if there is any relationship between the rate of mean reversion (kappa) and the vol of the vol parameter? Should kappa be greater than the vol of vol maybe?Hi zizou027Thanks for the info - busy working on my MSc - Focusing on Forward Starting options in the affine framework (Heston+jumps) - seems getting data for forward starting options isnt possible - not sure about your question

As a practical matter, for the spx, value of kappa this low would be unusual in my opinion.I would suspect a poor data set (say improperly calibrating to too few maturities) knowing nothing else but that number.Mathematically, the model allows any kappa > 0, assuming you want a mean-reverting vol.

Thanks Alan was hoping for that kind of answer cause i dont have any practical experience