Hello,
I would like to use Quantlib to bootstrap yield curves in a multi-curve environment. I read Ballabio's guide "Implementing Quantlib" and browsed source code. I understand that IterativeBootstrap's algorithm, which relies on BootstrapHelper, tries to equate market quote with implied quote. I would like to use a slightly different approach: I would like to equate the implied value with 0. The result should be the same, but code could be simpler: for instance computing the swap value could be simpler that computing the swap par rate. (I see that Real SwapRateHelper::impliedQuote() contains the comment "weak implementation... to be improved") What is the easiest way to modify Quantlib code to perform this different algorithm? Perhaps I could write my own RateHelpers, where I redefine the quoteError method. But since IterativeBootstrap is quite general, perhaps there is a better method. Regards, Federico Cozzi ------------------------------------------------------------------------------ _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Hi Federico, the easiest way might be to modify the rate helpers so that quoteError() returns the implied value. If you want to contribute the modifications, you'd also have to keep the old behavior, so you should add a boolean parameter or an enumeration to select the new or the old calculation. This would allow you to reuse most of the current code. Ferdinando, you wanted to work in this direction, too. Any comments? (For instance, about extending this approach to other instruments such as deposits, FRAs or futures?) Luigi On Mon, Dec 21, 2015 at 11:02 AM Federico Cozzi <[hidden email]> wrote: Hello, -- <http://leanpub.com/implementingquantlib> ------------------------------------------------------------------------------ _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Free forum by Nabble | Edit this page |