#include <qle/models/defaultlossmodel.hpp>

Inheritance diagram for DefaultLossModel:

Collaboration diagram for DefaultLossModel:

Protected Member Functions
	DefaultLossModel ()

Protected Attributes
RelinkableHandle< QuantExt::Basket >	basket_

Friends
class	QuantExt::Basket

Statistics
virtual Real	expectedTrancheLoss (const Date &d, Real recoveryRate=Null< Real >()) const

virtual Probability	probOverLoss (const Date &d, Real lossFraction) const

virtual Real	percentile (const Date &d, Real percentile) const
	Value at Risk given a default loss percentile. More...

virtual Real	expectedShortfall (const Date &d, Real percentile) const
	Expected shortfall given a default loss percentile. More...

virtual std::vector< Real >	splitVaRLevel (const Date &d, Real loss) const
	Associated VaR fraction to each counterparty. More...

virtual std::vector< Real >	splitESFLevel (const Date &d, Real loss) const
	Associated ESF fraction to each counterparty. More...

virtual std::map< Real, Probability >	lossDistribution (const Date &) const
	Full loss distribution. More...

virtual Real	densityTrancheLoss (const Date &d, Real lossFraction) const
	Probability density of a given loss fraction of the basket notional. More...

virtual std::vector< Probability >	probsBeingNthEvent (Size n, const Date &d) const

virtual Real	defaultCorrelation (const Date &d, Size iName, Size jName) const
	Pearsons' default probability correlation. More...

virtual Probability	probAtLeastNEvents (Size n, const Date &d) const

virtual Real	expectedRecovery (const Date &, Size iName, const DefaultProbKey &) const

virtual QuantLib::Real	correlation () const

void	setBasket (QuantExt::Basket *bskt)

virtual void	resetModel ()=0
	Concrete models do now any updates/inits they need on basket reset. More...

Detailed Description

Default loss model interface definition. Allows communication between the basket and specific algorithms. Intended to hold any kind of portfolio joint loss, latent models, top-down,....

An inconvenience of this design as opposed to the full arguments/results is that when pricing several derivatives instruments on the same basket not all the pricing engines would point to the same loss model; thus when pricing a set of such instruments there might be some switching on the basket loss models, which might require recalculations (of the basket) or not depending on the pricing order.

Definition at line 48 of file defaultlossmodel.hpp.

Constructor & Destructor Documentation

◆ DefaultLossModel()

DefaultLossModel ( )

protected

Definition at line 61 of file defaultlossmodel.hpp.

61{}

Member Function Documentation

◆ expectedTrancheLoss()

virtual Real expectedTrancheLoss	(	const Date &	d,
		Real	recoveryRate = `Null<Real>()`
	)		const

protectedvirtual

Reimplemented in GaussianLHPLossModel.

Definition at line 71 of file defaultlossmodel.hpp.

                                                                                            {
        QL_FAIL("expectedTrancheLoss Not implemented for this model.");
    }

◆ probOverLoss()

virtual Probability probOverLoss	(	const Date &	d,
		Real	lossFraction
	)		const

protectedvirtual

Probability of the tranche losing the same or more than the fractional amount given.

The passed lossFraction is a fraction of losses over the tranche notional (not the portfolio).

Reimplemented in GaussianLHPLossModel.

Definition at line 80 of file defaultlossmodel.hpp.

                                                                             {
        QL_FAIL("probOverLoss Not implemented for this model.");
    }

◆ percentile()

virtual Real percentile	(	const Date &	d,
		Real	percentile
	)		const

protectedvirtual

Value at Risk given a default loss percentile.

Reimplemented in HomogeneousPoolLossModel< copulaPolicy >, InhomogeneousPoolLossModel< copulaPolicy >, and GaussianLHPLossModel.

Definition at line 84 of file defaultlossmodel.hpp.

                                                                  {
        QL_FAIL("percentile Not implemented for this model.");
    }

◆ expectedShortfall()

virtual Real expectedShortfall	(	const Date &	d,
		Real	percentile
	)		const

protectedvirtual

Expected shortfall given a default loss percentile.

Definition at line 88 of file defaultlossmodel.hpp.

                                                                         {
        QL_FAIL("eSF Not implemented for this model.");
    }

◆ splitVaRLevel()

virtual std::vector< Real > splitVaRLevel	(	const Date &	d,
		Real	loss
	)		const

protectedvirtual

Associated VaR fraction to each counterparty.

Definition at line 92 of file defaultlossmodel.hpp.

                                                                          {
        QL_FAIL("splitVaRLevel Not implemented for this model.");
    }

◆ splitESFLevel()

virtual std::vector< Real > splitESFLevel	(	const Date &	d,
		Real	loss
	)		const

protectedvirtual

Associated ESF fraction to each counterparty.

Definition at line 96 of file defaultlossmodel.hpp.

                                                                          {
        QL_FAIL("splitESFLevel Not implemented for this model.");
    }

◆ lossDistribution()

virtual std::map< Real, Probability > lossDistribution ( const Date & ) const

protectedvirtual

Full loss distribution.

Definition at line 103 of file defaultlossmodel.hpp.

                                                                          {
        QL_FAIL("lossDistribution Not implemented for this model.");
    }

◆ densityTrancheLoss()

virtual Real densityTrancheLoss	(	const Date &	d,
		Real	lossFraction
	)		const

protectedvirtual

Probability density of a given loss fraction of the basket notional.

Definition at line 107 of file defaultlossmodel.hpp.

                                                                            {
        QL_FAIL("densityTrancheLoss Not implemented for this model.");
    }

◆ probsBeingNthEvent()

virtual std::vector< Probability > probsBeingNthEvent	(	Size	n,
		const Date &	d
	)		const

protectedvirtual

Probabilities for each of the (remaining) basket elements in the pool to have defaulted by time d and at the same time be the Nth defaulting name to default in the basket. This method is oriented to default order dependent portfolio pricing (e.g. NTDs) The probabilities ordering in the vector coincides with the pool order.

Definition at line 117 of file defaultlossmodel.hpp.

                                                                                   {
        QL_FAIL("probsBeingNthEvent Not implemented for this model.");
    }

◆ defaultCorrelation()

virtual Real defaultCorrelation	(	const Date &	d,
		Size	iName,
		Size	jName
	)		const

protectedvirtual

Pearsons' default probability correlation.

Reimplemented in ConstantLossModel< copulaPolicy >, and ExtendedConstantLossModel< copulaPolicy >.

Definition at line 121 of file defaultlossmodel.hpp.

                                                                                 {
        QL_FAIL("defaultCorrelation Not implemented for this model.");
    }

◆ probAtLeastNEvents()

virtual Probability probAtLeastNEvents	(	Size	n,
		const Date &	d
	)		const

protectedvirtual

Returns the probaility of having a given or larger number of defaults in the basket portfolio at a given time.

Reimplemented in ConstantLossModel< copulaPolicy >, and ExtendedConstantLossModel< copulaPolicy >.

Definition at line 127 of file defaultlossmodel.hpp.

                                                                        {
        QL_FAIL("probAtLeastNEvents Not implemented for this model.");
    }

◆ expectedRecovery()

virtual Real expectedRecovery	(	const Date &	,
		Size	iName,
		const DefaultProbKey &
	)		const

protectedvirtual

Expected RR for name conditinal to default by that date.

Reimplemented in GaussianLHPLossModel, ConstantLossModel< copulaPolicy >, and ExtendedConstantLossModel< copulaPolicy >.

Definition at line 132 of file defaultlossmodel.hpp.

                                                                                        {
        QL_FAIL("expected recovery Not implemented for this model.");
    }

◆ correlation()

virtual QuantLib::Real correlation ( ) const

protectedvirtual

Return single correlation number for one factor models. If not implemented or not applicable, returns a Null<Real>().

Reimplemented in PoolLossModel< CopulaPolicy >.

Definition at line 140 of file defaultlossmodel.hpp.

                                             {
        return QuantLib::Null<QuantLib::Real>();
    }

◆ setBasket()

void setBasket ( QuantExt::Basket * bskt )

private

Send a reference to the basket to allow the model to read the problem arguments (contained in the basket)

Definition at line 149 of file defaultlossmodel.hpp.

                                         {
        /* After this; if the model modifies its internal status/caches (if
        any) it should notify the  prior basket to recognise that basket is
        not in a calculated=true state. Since we dont know at this level if
        the model keeps caches it is the children responsibility. Typically
        this is done at the first call to calculate to the loss model, there
        it notifies the basket. The old basket is still registered with us
        until the basket takes in a new model....
        ..alternatively both old basket and model could be forced reset here
        */
        // RL: QuantExt version
        basket_.linkTo(QuantLib::ext::shared_ptr<QuantExt::Basket>(bskt, null_deleter()), false);
        resetModel(); // or rename to setBasketImpl(...)
    }

Here is the call graph for this function:

◆ resetModel()

virtual void resetModel ( )

privatepure virtual

Concrete models do now any updates/inits they need on basket reset.

Implemented in ConstantLossModel< copulaPolicy >, ExtendedConstantLossModel< copulaPolicy >, HomogeneousPoolLossModel< copulaPolicy >, InhomogeneousPoolLossModel< copulaPolicy >, GaussianLHPLossModel, and PoolLossModel< CopulaPolicy >.

Here is the caller graph for this function:

Friends And Related Function Documentation

◆ QuantExt::Basket

friend class QuantExt::Basket

friend

Definition at line 54 of file defaultlossmodel.hpp.

Member Data Documentation

◆ basket_

RelinkableHandle<QuantExt::Basket> basket_

mutableprotected

Definition at line 59 of file defaultlossmodel.hpp.

Protected Member Functions

Protected Attributes

Friends

Statistics

Detailed Description

Constructor & Destructor Documentation

◆ DefaultLossModel()

Member Function Documentation

◆ expectedTrancheLoss()

◆ probOverLoss()

◆ percentile()

◆ expectedShortfall()

◆ splitVaRLevel()

◆ splitESFLevel()

◆ lossDistribution()

◆ densityTrancheLoss()

◆ probsBeingNthEvent()

◆ defaultCorrelation()

◆ probAtLeastNEvents()

◆ expectedRecovery()

◆ correlation()

◆ setBasket()

◆ resetModel()

Friends And Related Function Documentation

◆ QuantExt::Basket

Member Data Documentation

◆ basket_