August 24th, 2009, 8:29 am
QuoteOriginally posted by: mathmarcQuoteOriginally posted by: MartinghoulQuoteOriginally posted by: pcgWill there not always be a markt consensus on what curve to discount on ? Because if you are using a diff curve , there will be a problem if one is trying to terminate a swap and the settlement amounts will be different based on your discounting being different .No, there won't necessarily be a consensus. That's precisely the point of the discussion in the two threads I have mentioned...I fully agree, there won't necessarily be a consensus. The only moment were there is (should be) a consensus is for ATM standard vanilla swap. In that case the value is 0. For all other cases, it depends. This is not only theory, it is also what one notices in practice. To notice that you don't have to try to terminate a swap. If you try to enter into an asset swap (for example to asset swap a bond) and the bond price is away from par, you will see noticeable differences between banks quotes. Some banks use Libor for the non-par part some use very different curves (OIS, Limean, Libis, ...)Which is why par-par asset swaps are sorta anathema to me now... For all their drawbacks, it's yield-yield only for yours truly.