Hi folks,I am new to this forum although i have been following it for quite some time. I need some help regarding the problem i am working on.The problem is on pricing of a CMS spread Range Accrual note, whose characteristics are as follows(i) The payoff depends upon the the spread (5yr CMS - 2yr CMS) > 0. Coupon accrues on the days on which the above condition is satisfied(ii) Maturity = 5 years and annual coupon payment of 10% on 10million USD notionalCan somebody suggests how to go about pricing the above note?The methodology i had thought of is:(i) Modelling both the 5yr CMS rate and 2 yr CMS rate daily until maturity using HW 2 factor spot rate model and Trinomial Tree method(ii) Compare the rates generated daily check whether the spread (5yr CMS - 2yr CMS) > 0(iii) Caculate the discounted values of all the annual payoffs to obtain the necessary risk premiumI also wanted to know if its better to use 2 factor HW model or if any other model would make more sense in this case and also if Trinomial tree method is a good option than Monte carlo simulation. In Trinomial tree method i feel it requires generating two seperate trees simultaneously with some correlation (between 5 yr and 2 yr CMS rates) between them which i am not clear how to go about it and since delta-T is 1 day in this case it becomes a computational problem as well. Can somebody suggests a solution here?Thanks

Hi dhillybabu, may I ask u how do u treat correlation in HW?Anyway, daily checking of optionalities is always time consuming...

Hi,1. are you sure, that a 2 factor model will do ? Once we investigated the (callable floating) range accrual of 5Y and concluded, that the prices stabilize with 5 factor BGM. 2. Regarding the correlation : => Why don't you use an implicit method ?=> The differential operator can be decomposed to /partial_{\xi \xi} ^{2} + /partial_{\nu \nu} ^{2} with $\xi = x+y$ and $nu = x-y$So, I guess, that if you do not have a boundary exhibiting cartesian symmetry, you can use this trick.Or, the ADI by Creig & Sneyd is able to handle mixed derivatives.

> Or, the ADI by Creig & Sneyd is able to handle mixed derivativesWhy do you advise this method? Just curious.

Looks like the USD 2-10 spread is finally going to invert, I wonder how are those people holding callable spread range accrual conditional upon a positive curve is feeling now!?