Thanks for your answer Thomas, made me laugh a lot )

<t>I would say the best would be to find an opportunity in a diversified hedge fund, with desks doing global macro/long short/risk arb/vol arb/credit/FX, etc... and with, first, much freedom, being able to chose between direct trading and research from time to time, being considered directly as a tr...

"linear in case of pure diffusion" ?Can you elaborate ?

Hi,I'm wondering how you would prove (or at least explain) that the term-structure of ATM vol is in square root of maturity.And same question with the skew (slope of vol around the money).My interest is to understand what assumptions are behind this !Any point on that ?Thanks

I doubt it too, I imagine it was just an image, 'cause I'm wondering if there are 1M investment bankers in the whole world actually ...The last figure I heard was 40,000 traders ...However, I agree the world would be running at least as well without them as with them !

<t>Probably, yes...Actually, different forms of solutions can be derived, and while one seems to work properly, another poses a pb of discontinuity with the complex log, I don't know if you're aware of that ...For this purpose, I invite you to check this article from Lord and Khal : "Why the Rotatio...

<t>ASA is not a perfect global optimizer, it can miss the global optimum...The only way to do a global optimization is to go through the whole space of parameters actually...Wether you use one value or another for your kappa, you can find it approximately stationary : just recalibrate with a local o...

<t>Here are some points about calibration :- You should first try and calibrate to a single smile (but to many points, like 16) : for a classical smooth shape, you are supposed to be able to fit perfectly the smile with the Heston model, and even, there will be different sets of solutions possible !...

<t>V1 = instantaneous varV2 = longrun varlambda = mean-reversioneta = vol of volrho = correlI didn't try the FFT, I just went for numerical integration procedures ... at first, I tested my pricer with simpson's rule which turns out to be as slow to converge as simple to implement.To obtain a decent ...

<r>You can test your results againt Alan pricer at <URL url="http://www.optioncity.netElse">www.optioncity.netElse</URL>, here are some results I used to test my code :spot 100time r V1 V2 lambda eta rhoparameters: 1.0,0,0.04,0.04,2.0,0.39,-0.64 Strike, HestonPrice, HestonVola 60, 40.226886649835, 0...

Your question seems to be on the wrong forum but above all, it's almost impossible to understand, lolLooks like your junky portrait is well-chosen !

thanks for your answers!

<t>Intuitively, the slope of the smile is a measure of the skewness of the implied distribution, and the convexity of the smile a measure of its kurtosis.Are there quantitative relations or rules of thumb on that ?I was wondering, for example, what level of (negative) convexity of the smile implies ...

I didn't know NORMSINV function, I'm using Box Muller method to do so .. does NORMSINV work well ?