This is question for Mark. I am trying to calculate deltas for a structured product whose payoff is given by the followingif gearingC * CMS_C(t) - gearingD * CMS_D(t) >0 then coupon(t+1) is: floor< gearingA * CMS_A(t) - gearingB * CMS_B(t) <capif gearingC * CMS_C(t) - gearingD * CMS_D(t) <0 then coupon(t+1) is: nominal fixed couponfor t= 1:30 yrs and it is a callable product . Here the reference is one CMS spread and coupon is another CMS spreadThe pricing of the structure is simple and straightforward but I want to figure out how I can apply a partial proxy method to calculate deltas and whether it is possible to use partial proxy here at all. Any help will be appreciated.regards,Ahsan

Hi, Ahsan, may I know what model you calibrated to price this structure? Do you use stochastic volatility Libor Market Model since there are four CMS smile curves and correlation structure among them?

No, I am not using a stochastic volatility model just a deterministic vol model with skew. I am working on industrial use of stoch vol models but I still have issues with them.

The purpose of partial proxy is to eliminate discontinuties in the pay-off. This suggests that you want to fix gearingC * CMS_C(t) - gearingD * CMS_D(t)you could do this by using a 2 or more factor model and fixing CMS_C and CMS_D using linearization. Or you could fixgearingC * CMS_C(t) - gearingD * CMS_D(t) itself using linearization.

Thank you Mark. I may come back with more questions once I have worked out the details.regardsAhsan