<t>"One point of having time dependency in SABR parameters may also be avoiding recalibration"I do not think that you'll avoid recalibration. In reality if 2 months pass (probably dependant on your market) you would have to recalibrate since the whole surface would have changed. Do you have a workin...

<t>Hi All,In the newest version (should be ready on Friday) I made some more comparisons for the heston model. The numbers look good. From a numerical perspective the SABR PDE is easier to solve than the Heston PDE since there are no drift terms in the SABR PDE.If we are lucky similar errors would o...

Do you have a draft copy for me? If not would you please give me the title of the paper?Thx for your help.R

<t>Hallo,We have relatively sparse set of implied volatilities and would like to "fill the gaps". The problem is that when we try to interpolate (using cubic spline) it seems to introduce arbitrage opportunities between the interpolated data. For example negative values for calendar spreads, call sp...

<t>Wazzup.I am using a local volatility model to price a lookback option, (Monte Carlo based). I am trying to find the "area" on the implied vol surface to which the lookback option is sensitive. For example say we have skews for the following maturities: 44day, 135day, 228day, 365day and 667day. Co...

<t>I busy searching for finite difference schemes that work well for the PDE's that arise in stochastic vol models. Like all applied math, it became clear to me that there is no quick fix. ADI sucks because it can not handle the cross derivative term, the transformations needed to removes this cross...

The fitting method in this paper by Daniel Duffy might be helpful.

<r>A finite difference method approach to value Americans in Heston can be found here, <URL url="http://www.cc.jyu.fi/~samikon/http://www.mit.jyu.fi/tene/papers/index.html"><LINK_TEXT text="http://www.cc.jyu.fi/~samikon/http://ww ... index.html">http://www.cc.jyu.fi/~samikon/http://www.mit.jyu.fi/te...

This might help:http://www.finmod.co.za/resources.htmlR

After 'fixing' my code I still get more or less the same results for (1) and (2):1) 5.341+0032)6.376e+003Let’s make sure we are working with the same parameters,where \kappa = rev_speed\theta = reference_vol\sigma0 (spot vol) = sigma\nu = volvolPS. Do you have Maltab??

There is some difficultly if vol is very small as well 1) 5.4225e+0032) 6.3838e+0033) 5.0302e+0034) 5.4637e+003These values should NOT be used as a benchmark.

Interesting side note on MC:Due to discretisation error MC can give negative values for processes that, in the continuous case, can not become negative. Take the lognormal (BS model) for example,dS = rS dt + sigma S dWThe explicit Euler discretisation,where ,clearly does not force .

Quote another trial for you:Still need the following parameters:T=?S=? (spot underlying)R

<t>I think that it is important to have a very efficient (or fast) way to value vanillas, for calibration purposes. At this stage of the battle it seems that closed form (or semi-closed form) is the best way to go. Maybe some time in the future we will be able to have a finite difference scheme that...

<r>Are you thinking about building a term structure for the parameters, like in the dynamic SABR model?According to me the Heston model is calibrated to all maturities and not just a single maturity whereas the SABR model is calibrated for a single tau only. If you want to price path dependants in t...