Exposing CallableFixedRateBond to QuantLibXL: how to use ShortRateModel instead of AffineModel

Posted by Lisa Ann on Sep 24, 2015; 10:04am
URL: http://quantlib.414.s1.nabble.com/Exposing-CallableFixedRateBond-to-QuantLibXL-how-to-use-ShortRateModel-instead-of-AffineModel-tp16923.html

I've followed this great tutorial and it worked; furthermore, I've exposed AmortizingFixedRateBond, AmortizingFloatingRateBond and AmortizingCmsRateBond, and they worked fine.

Now would like to expose a simple "wrapper" for CallableFixedRateBond, but I'm likely stuck with a design problem.

My idea is to extend QuantLibAddin\qlo\bonds.hpp and QuantLibAddin\qlo\bonds.cpp: declaration in the former should not require anything more than almost the same constructor used by FixedRateBond (plus const std::vector<QuantLib::Real>& callPrices, const std::vector<QuantLib::Real>& callDates and const boost::shared_ptr<QuantLib::ShortRateModel>& shortRateModel), implementation in the latter, on the contrary, is giving some issues.

Declaration: #include <qlo/shortratemodels.hpp> and #include <ql/models/shortrate/onefactormodel.hpp> should allow to use const boost::shared_ptr<QuantLib::ShortRateModel>& shortRateModel in constructor declaration... correct?

Implementation: here are some issues. #include <ql/experimental/callablebonds/callablebond.hpp>, #include <ql/experimental/callablebonds/treecallablebondengine.hpp> and #include <ql/instruments/callabilityschedule.hpp> are needed, of course.

The main issue is: C:\build_ql_1_6_0\QuantLibAddin\gensrc\metadata\types\types.xml has only AffineModel as libraryClass, whilst QuantLib::TreeCallableFixedRatebondEngine class wants a const boost::shared_ptr<ShortRarteModel>. Here we come to C:\build_ql_1_6_0\QuantLibAddin\gensrc\metadata\functions\bonds.xml: the only parameter different than other bonds' ones is the short rate model id. My guess:

        <Parameter name='ShortRateModel'>
            <type>QuantLib::AffineModel</type>
            <superType>libraryClass</superType>
            <tensorRank>scalar</tensorRank>
            <description>Short rate model ID.</description>
        </Parameter>

Nevertheless, I cannot use QuantLib::AffineModel type here because QuantLib::TreeCallableFixedRatebondEngine wants a boost::shared_ptr<ShortRarteModel>.

Any solution to use QuantLib::AffineModel as type but actually QuantLib::ShortRateModel as template for the boost::shared_ptr which I will pass to QuantLib::TreeCallableFixedRatebondEngine?