Serving the Quantitative Finance Community

 
User avatar
v8625
Topic Author
Posts: 0
Joined: October 25th, 2005, 10:20 pm

VaR approximation - IR swaps

April 8th, 2009, 6:33 pm

Here is a new plain vanilla interest rate swap:pay 2.9%receive 3-mo USD LIBORon USD 10 million notionalfor 10 yearsBloomberg gives DV01 (net of both legs) USD -8,250and is showing market value of USD 286,000
 
User avatar
v8625
Topic Author
Posts: 0
Joined: October 25th, 2005, 10:20 pm

VaR approximation - IR swaps

April 8th, 2009, 6:36 pm

volatility for USSW10 Index HVT <GO> is 47 (per cent)
 
User avatar
v8625
Topic Author
Posts: 0
Joined: October 25th, 2005, 10:20 pm

VaR approximation - IR swaps

April 8th, 2009, 8:40 pm

H-m-m..., no "quick-and-dirty" way to approximate VaR? "Must" build 200-and-something curves and "must" do full valuation?
 
User avatar
jomni
Posts: 0
Joined: January 26th, 2005, 11:36 pm

VaR approximation - IR swaps

April 9th, 2009, 12:03 am

The problem with netting the legs before computing for VaR is that you assume the rates to go in the same direction at the same magnitude.This is not the case since rates used for both legs of your IRS are different despite being picked from the same curve.Taking basis risk and yield curve twists... correlation needs to be estimated for the rates.
 
User avatar
v8625
Topic Author
Posts: 0
Joined: October 25th, 2005, 10:20 pm

VaR approximation - IR swaps

April 14th, 2009, 5:16 pm

I would still like to try and find a way to get a sensible VaR estimate for a vanilla swap based solely on values provided by Bloomberg - without building/modeling anything. Any ideas in that sense would be highly appreciated.