Serving the Quantitative Finance Community

 
User avatar
fab10ab
Topic Author
Posts: 0
Joined: February 13th, 2006, 1:37 pm

Summit cap implied volatility problem

July 29th, 2009, 2:31 pm

We are using Summit to calculate implied volatility surfaces for interest rate derivatives (mostly caps and floors). Summit reads in a matrix of flat volatilities from Bloomberg, and calculates an implied volatility matrix from these flat vols. It then interpolates this matrix to get forward volatilities to price caplets at each reset date. On some days the matrix of forward volatilities is completely wrong, and consequently, so is the cap valuation. Has someone else ever experienced this problem? Thanks.
 
User avatar
Arn
Posts: 0
Joined: January 5th, 2004, 9:33 am

Summit cap implied volatility problem

May 7th, 2013, 12:20 pm

hi fab10ab,I haven't experienced the problem, but there are several things to be noted.The method is quite simplistic but has the merit to be easy to understand (which is good), albeit for a market user (not a market maker of course) it should suffice in most cases. You can override it with your own with the API.Being simple, the method it is easy to test back: so you reprice the original caps with the output caplet vols at market...then it is working. (you can open a cap trade input screen and select to price it with a market vol that you input manually and this way you can check easily)Sometimes with stale data, the market vols will be slightly inconsistent (too steep for instance) and there might not be a solution that allows for those prices in the first place.It should be re-run a few times during the day especially if the yield curve moves, because for pricing the caps you need the yield curve and you need it also to calibrate, and if it changed in the meanwhile then you will not re-price the caps at market.