Forward Correlations from Caps/Swaptions

Tapiwa · March 20th, 2003, 9:59 am

Apologies for any repetition of this issue but I have read back through some topics and still can't quite grasp the concept. Is there any way to infer correlations between Euribor contracts, (i.e the red and green Euribor contracts) by looking at the relative volatilities of caps/floors and swaptions. For example, if I was looking at 1Y*2Y cap/floor vols and 1Y*1Y swaption vols, could I make any inferences about the red and green Euribor contracts without wanting to directly model the actual number.

FDAXHunter · March 20th, 2003, 10:33 am

yes, there is.You simply need to look at the caplet vol vs the swaptions vol, and you can bootstrap a implied forward correlation.It usually displays a structure where the maximum correlation implied is around the red months (around .9) and then falling off after that (which makes sense).I'm not sure about looking at the direct numbers and getting a rule-of-thumb number. Let me think about it, there might be a rule of thumb in there.Hope this helps.

Tapiwa · March 20th, 2003, 10:41 am

Thanks a lot, just knowing there is a way is a great help.If it is not too much trouble, any expansion you could do on the practical method of doing this would be great.

tubul · March 21st, 2003, 6:06 am

Hi Tapiwa,There is indeed a method. U can use the "string mkt model" (implemented just like a multifactor HJM) for doing so. I am attaching a recent paper that does precisely what u are looking for.Regards,Tubul

Tapiwa · March 21st, 2003, 8:42 am

aha,I'll go away and have a look at this method. Thanks very much.