Sensitivity Analysis in QuantlibXL

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Sensitivity Analysis in QuantlibXL

Ben Watson

Hi There,

 

I am looking for examples of how to create a delta sensitivity of a swap to a particular curve. I see the example in BucketAnalysis.xlsx and it provides a sensitivity to a discounting curve that has been created using a single qlSimpleQuote. However in my case I have a vector of instruments that have been created when I constructed the curve. When I pass this  vector to qlBucketAnalysisDelta2 I get the error below which looks like some casting error.

“qlBucketAnalysisDelta2 - operToVectorImpl: error converting parameter 'SimpleQuote' to type 'class QuantLib::Handle<class QuantLib::Quote>' : Unable to coerce value from type 'class boost::shared_ptr<class ObjectHandler::Object>' to type 'class QuantLib::”

 

It is also not very clear in the documentation what the second parameter inventively called Parameters is for. The documentation requests a parameters vector. I was thinking that could be a set of instruments.

 

Am I looking at this problem correctly – what I am after is a delta ladder that show the risk of a swap or portfolio of swaps to a discount curve. An Excel example would be ideal, but happy with whatever advice I can get.

 

Down the track I would also like to look at gamma and model to do OIS discount where we project from a dual bootstrapped Libor curve and discount on an OIS curve. The delta risk in this case would be to both the project and discounting curves.

 

Regards

 

Ben


------------------------------------------------------------------------------

_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users
Reply | Threaded
Open this post in threaded view
|

Re: Sensitivity Analysis in QuantlibXL

Sparviero
I think qlLegBPS will works fine. Anyway you can shift IR curves by 1bps and compute the difference between the NPVs.

Cheers,

Iacopo