<r>I was using the Fengler method before and found it a little disappointing (complex to implement, and some instabilities on the resulting vol surface).Rather than that, I switched to the method described in this paper: <URL url="http://www.zeliade.com/whitepapers/zwp-0005.pdfIt%27s">http://www.zel...

You should be prepared for some technical questions, you will most probably be asked some.Of course, it also depends on which team you are interviewing for. Is it for equity or ficc ?

I get 1/6 using

<r>I've found this using google, they also obtain 42 : <URL url="http://www.rpgcodex.net/phpBB/viewtopic.php?p=958326&sid=98493368af40e7e0260735fee8be7056They"><LINK_TEXT text="http://www.rpgcodex.net/phpBB/viewtopic ... be7056They">http://www.rpgcodex.net/phpBB/viewtopic.php?p=958326&sid=98...

<t>If we call s the state where the last throw is not a 6 and s' the state where the last throw is a 6.If you are in state s and throw the dice, you have probability 1/6 to have a 6 and reach state s' and probability 5/6 not to have a 6 and so to stay in state s. As you use one throw, you get e = 1 ...

<t>I found 42 as follows (not sure if it's correct though)Let us call e the expected number of throws knowing that the last number was not a 6.Let us call e' the expected number of throws knowing that the last number was a 6.Then e = 1 + 1/6 * e' + 5/6 * eAnd e' = 1/6 + 5/6 * (1+e)From there it's ea...

As MCarreira said, Gatheral's SVI is quite good for this and to ensure some arbitrage-free conditions. Here is a paper that details how to do SVI calibration in an efficient way.

<t>Hello Alinghi,You should only use your ZCB (or swap) quotes in order to obtain the F_t(0,t) + aF(0, t) part in your expression of . Indeed, for any possible value of a and sigma, P(0, t) will correspond to your ZCB quotes, this explains why your minimization procedure returns weird results. The f...

Thank you both for the insight.About the volatility cube, if others are interested, I found the following technical report.

<t>I have implemented a Markov functional Libor Forward/Swap model as described in Hunt and Kennedy's book "Financial derivatives in theory and practice". Calibration works very well on an implied volatility surface for digital caplets / swaptions.However I'm not sure about how to obtain this kind o...

<r>It is also possible to use interleaved Sobol sequences on each node. I.e. one node will generate sequence s_1, s_3, s_5... and the other one will generate s_2, s_4, s_6...An example of implementation is gived in "Quasirandom Number Generators for Parallel Monte Carlo Algorithms" B. C. Bromley 199...