Serving the Quantitative Finance Community

 
User avatar
prodiptag
Topic Author
Posts: 0
Joined: September 12th, 2008, 4:41 pm

swaption prices from probability distribution

June 29th, 2012, 2:33 pm

the reverse of the more usual thing. I input a probability distribution to price a swaption - say a 1m10y payer of strike k. the way I price it is to break the payer swaption pay-off as an infinite sum of digital payers at strike k, k+deltak, k + 2*deltak ... and so on. The n-th such digital option has a pay-off = deltak*dv01(evaluated at k+n*deltak)*cumulative_prb(rate > k+n*deltak). The price of the payer is sum of all such digitals. If I just extract the probability distribution from market prices, and then feed it back through the above algo to get back the price, I do get close, but NOT close enough. Just wondering if any neumerical thing I am missing or there is some ceonceptual problem with the method. Thx vm for your help