<t>Hihas anyone thought about which equity volatility should be used in the quanto adjustment (i.e. vol_equity x vol_fx x corr_Eq_fx) in the standard BS environment when the option is not at the money?Shall I use two different parameters (i.e. volatility taken on the strike of listed, vanilla produc...

<t>can anyone suggest any approach when choosing the optimal set of parameters (N, i.e. grid dimension, gamma, i.e. dumpening parameter, delta, i.e. gridsize for logstrikes) for implementing option pricing in the Heston model using fractional fft? Say I want to produce N=128 option prices with the o...

<t>Hi Sonyahthanks for your replyso you basically sugget the use of simulations only in order to properly select the multiplier of the strategy?Is it then correct to proceed with the pricing of the structured note in the way I specified in my previous post? (i.e. getting the theoretical bond floor v...

<t>Hi and thanks for your replysomething I don't get is why should I use Monte-Carlo when I don't really need to price any call on the CPPI, since the rebalancing strategy would allow me to guarantee 100 at the end (apart from crash risk). I mean, what should I do with the distribution of results of...

<t>Say I want to price a structured bond that has a fixed coupon structure, final capital redemption, and a final coupon based on the performance of an asset defined as a CPPI: i.e. a dynamic basket, made of a risky asset (such as a fund or funds portfolio) and a riskless asset (in this case made of...

Thank you for your reply Exoticathis sounds clear enough, and in line with my results

talking about standard errors of MonteCarlo, I was also wondering if the number of time steps actually affects this value or is it only influenced by the (square root of) number of paths in a set of simulations...

Thanks Antonio,do you have any reference for this? I'm quite curious about this dependence on parameters.

Say I want to simulate an Heston process using Monte-Carlo for a vanilla option.Is there an appropriate number of timesteps to be used in order to converge to the analytical result given a certain number of paths generated (i.e. a relationship between the two) ?Thank you

<t>Hi Athleticothanks for your kind reply!I tried following your instructions, but still cannot get a proper result from my MC. Here's the piece of the code that I use: mj = Exp(meanJ + 0.5 * deltaJ^2) - 1 r = (rate - yield - lambdaJ * mj - 0.5 * vol^2) * dt sd = vol* Sqr(dt) isJump = Rnd() < lambda...

<t>Thanks again rada, I tried changing also the jump term but I still get more or less the same results as before...this is interestingAthletico, I didn't include the vol diffusion part since it would be not affected by the jump process, assuming the two are independent. However, I modeled it the no...

Hi radathanks for your kind replyI actually rewrote the code correcting the drift to what you suggest but basically don't see any improvement in teh results.Maybe I should also rewrite the diffusion part in some way?

<t>HelloI'm trying to implement a discretization of Heston plus jumps model (or Bates model) in a Montecarlo framework.In order to check correctness of my code, I compare the results with those coming from FFT implementation for vanilla calls.Whilst I found the Heston with no jumps code implementati...

Thanks AVteverything is clear now, I tried your formula with these corrections for the yield and they're in line with other sources as well as with simulations.Thanks for your great help and work in this field, and for sharing it online here.M

<t> hi AVt,could you check if in your otherwise nice and helpful formula the dividend yield is correctly included? I'm facing some problems getting similar results between closed formula in your sheet and MC simulations when using a positive yield. Indeed I find results very close when, on the other...