<t>The simplest solution is to have an "if statement" in your program or spreadsheet, so that when the model outputs a negative number you replace it with zero or some really small number (0.01%). ----------------------------------------------------------------------------Ahh.. I did that you see, b...

OK. So I made the volatility smaller, and that solved the problem.But out of curiosity still, if the volatility was large, how do I make it non-negative?

Any idea of how to do this? In terms of computational algorithm?I must be doing something wrong, because my path keeps going negative, but CIR don't go negative....

<t>How can I simulate the path of a CIR process?I need to simulate the whole path,dr=k(h-r)dt+s*sqrt(r)dW(t)do i just simulate withr(t)=r(t-1)+k(h-r(t-1))*dt+s*sqrt(r(t-1))*I*sqrt(dt)where I is a N(0,1)..basically, i am doing that at the moment, but the volatility bit always drags r(t) into negative...

<t>Finally, I have a set of results to work with!Just like to say thanks to both of you kr and Aaron,for taking the time to answer idiotic questions and hopeless queries...I'd read the paper so many times before hand I thought I couldn't have missed it...Thanks for putting me on the right track anyw...

<t>I've GOT IT!! I've GOT IT!It was in the paper all along!!pg.19"The par coupon rate on a given sinking-fund bond is the scheduled coupon rate c with the property that the initial market value is eqaul to its principal/initial face value F(0)"That's it!Thank you kr!! and Aaron...you've been great h...

<t>I am not sure what you mean luizvus.A deterministic one as in one that changes according to just time?I don't need any underlying volatility factor to drive it like a brownian motion if thats what you mean.I just need it to be the distribution to exhibit a U(0,1) PDF if I draw infinite samples fr...

<t>KR have kindly pointed out that all I have to do is find the NPV of a tranche, based on the prioritisation scheme and an instance of the underlying assets performance. Then given the default details, the coupon spread is merely the amount needed to pay the tranche back up to par.Any ideas on how ...

<t>>-based on the cashflows to the tranches, you calculate their NPV by discounting; since the tranches may default, you'll need a nonzero >coupon to get the tranche prices up to par.Hmm, I think I have worked this out and sort out my confusion, just to confirm it:--NPV of cashflows(based on an inst...

<t>Thank you kr for taking time to actually read the paper on my behalf, it is very much appreciated..very kind of you.I will now take a bit of time to digest it, although it seems fairly straight forward now that you say it.Basically, the non-zero coupon is just the amount needed to pay up the tran...

<t>Thanks kr. I did read somewhere that the spread produced were way too low, and in fact the correlation of default risk was also too low. But since I only have three months to reproduce something similiar to Duffie & Garleanu's effort from not even knowing the difference of structural and redu...

Any suggestions?How about pricing of any securitisation notes?Is it a matter of working out the cash flow and then discounting it?But how will I find the par coupon spread for this?

Thanks guys for your replies. I am using Microsoft Visual C++perhaps not the best choice really... and there are so much problems with name space and such that I can't use the Mersenne twister I found on google. Nvm.. thanks anyway.

<t>I'm not sure what you mean by simulating them separately and adding them together. If probability of jump depends on level of the diffusion process, how can you simulate the jump process on its own?As for correlation, do you mean correlation between different bonds or correlation of diffusion pro...

<t>I am trying to write a function which draws a from a U(0,1) uniform distribution. I can use srand() and rand() to do this fairly sufficiently if I were to make just one sample.But as it is, I will be making repeated calls to the function for random numbers. I have planted the seed for srand() wit...