July 28th, 2005, 5:52 pm
Hi, I haven't looked at your attachment, but here isone approach that should work for you.Define the function f(t,S,theta) = delta = dC/dS, wheretheta is whatever you are calling your extra affine state variable(s).If you take the S-derivative of your original PDE, you will see thatf() satisfies an equation of the same form, except thatthere is no discounting term, and the r in front of the first partialhas become r + sig^2. Anyway, repeat your solution methodwith this (new affine) equation, which now has the payoff f(T,S,theta) = 1(S > K),and you should get what you need.regards, p.s. finally looked at the attachment. You 'could' do it that way, too,by changing integration variables: z -> z + i in your next-to-last equation. This is effectively a contour move, which is legal because there are nopoles crossed under the move.
Last edited by
Alan on July 27th, 2005, 10:00 pm, edited 1 time in total.