<t>Hi,I'm looking for an external economic scenario generator, which can produce real world scenarios and if possible also risk neutral scenarios (with similar dynamics). The tool has to be user-friendly, but complex enough to capture realistic behaviour of all asset classes. I have come across the ...

<r>Hi,I have written a paper about analytical valuation of average CMS-rate options, with very good results. I have done this in the context of insurance companies, who have sold such options to their policyholder.But I'm wondering whether these options are also regularly traded on the financial mar...

Sorry, in the 2th sentence I mean "below" instead of "above"...

<t>You will probably recognise the situation that there is a mistake in your model, you know it is probably a stupid or obvious mistake, but you can't find it. Well, that is exactly where I am now. There is a mistake in my implementation of the UPM-model of Brigo & Mercurio. The simulated option...

It is in Appendix B in the book of Brigo & Mercurio.

But there are good approximations, see the paper of Lord.

Or you can look at the book of Brigo & Mercurio, where this issue is well described.

You can simulate the option in a T20 forward measure. Then for each payoff for t < T20, you bring it to T20 by dividing the payoff by the simulated P(t,T20)'s.The resulting payoffs can be discounted by P(0,T20).

Why not use the standard closed form solution mentioned in Hull?

Why don't you use the method of schrager & pelsser? This gives an analytical formula for the swaptions in n-factor affine models, so no numerical solving is necessary and the results are still good.

<t>I think that if you are simulating in the risk neutral measure, then the 3M libor and other swap rates are not martingales under that measure, so the average of the simulations will never be equal to the rates from the yield curve. It is better to check the P(0,T)'s from the yield curve with the ...

There is no exact analytical formula, but with a relatively simple numerical exercise you can do the method of jamshidian, that is given in most books about interest rate models.Furthermore, a good approximation is given by Schrager & Pelsser.

Thanks guys.

<t>Hi,If you come into the area of valuations in incomplete markets (such as insurance products) and you want to do valuations with realistic assumptions, you will end up with PDE's that are 5-dimensional or more. The computer doesn't even begin with solving this with the finite difference method be...

<t>The correlation matrix has to be positive semidefinite, or else the procedure won't work.That means that sometimes with very high correlations or correlations of different classes that don't match (for example correlation {a,b} = 1, {a,c} = 1, but {b,c} = 0, as an extreme example) you will get er...