wilmott.com

Is anyone familiar with this method of calibrating a vol smile to market prices? See www.damianobrigo.it for more info.I am wondering how to implement it is practice. I have been told that if I model option prices as (for N=2 case): w.BS(v1, f1) + (1-w).BS(v2,f2), where BS is Black Scholes option pricing formula, v1 and v2 are two process vols which are independent of strike. f1,f2 are two process forward prices. Market option prices are given by BS(v(s),f) where v(s) is vol which as we all know depends on strike s. Then minimise sumsq diffs between model and market prices.Are there any additional constraints here? e.g. do I have to constrain v1, v2 or f1,f2 ???Any info much appreciated.THanks,Jon

Yeah,I think you have to put some constrains on it, at least a poistive v1,v2 is necessary. Edward

You also might want 0<w<1 to ensure positivity.K.

Assuming you want your implicit risk-neutral density to be risk-neutral,w*f1 + (1-w)*f2 =f,with f the forward price for when the options expire.SV

Hi svetlana,have u ever tried to put on the constrant u mentioned?I had tried,but am a little confused about it. Since F is usually a number with decimals,say F=exp(-0.02*1). How can we find a w that exactly fit the constrant? Please kindly advise.

Hi "dinner"I have sent you a private message. Methods and results for several risk-neutral density estimation methods are in Chapter 16 of Taylor's Asset Price Dynamics, Volatility, and Prediction, available from the Wilmott bookshop and elsewhere.SV