Rebonato's work on cap swaption in BGM

elan · Post by **elan** » Tue Jun 29, 2004 1:01 pm

The calculations in this paper are quite ingeneous. One problem is that he ignores the existence of conventional day count fractions, and the result is more approximate than intended...

Nonius · Post by **Nonius** » Tue Jun 29, 2004 2:06 pm

I find day counts to be one of the stupidest and tedious things to worry about. Why can't everyone just conform to my rules of daycounting?

by the way, I thumbed through it, but didn't let it sink in. So, can one do plain vanilla pos-semi definite stuff to calibrate a la Stephen Boyd?

elan · Post by **elan** » Tue Jun 29, 2004 2:59 pm

I agree on day counts. Would also add business day conventions to the list of complaints. Basel III should get rid of all of these.

Sure one can. But this paper is about an intelligent way of deriving an approximate fast pricing formula for swaptions which doesn't require MC.

Nonius · Post by **Nonius** » Tue Jun 29, 2004 3:05 pm

I thought the Dufresne and Goldstein paper was mildly clever for the Affine models. Is it that sort of expansion trick?

One thing slightly annoying about Dufresne Goldstein is I don't see how they can price as fast as they say they do. the computations of the moments ain't that fast. lots of terms. but, it seems like what they do is just set some terms to zero, I think the 5 and 6th cumulant.

elan · Post by **elan** » Tue Jun 29, 2004 3:07 pm

I agree.