Loss level versus time interpolated scalar copula type parametric
correlation term structure. Represents the correlation for the credit loss
level of a given portfolio at a given loss level and time.
todo
The relation to a given basket is to be made explicit for bespoke
models to be implemented.
todo
Consider moving to a matrix data structure. A matrix might make some
computations heavy, template specialization on the dimension might be an
alternative to having two classes, one for scalars and another for matrices.
todo
Rethink all the data structure with a basket where current losses are
not zero.
todo
In principle the 2D interpolator is left optional since there are
arbitrage issues on the interpolator type to be used. However one has to be
careful when using non local interpolators like CubicSplines which have an
effect on the past (calibrated) coupons of previous tenors.
Matrix based Base Correlation Term Structure
Loss level versus time interpolated scalar copula type parametric correlation term structure. Represents the correlation for the credit loss level of a given portfolio at a given loss level and time.
The relation to a given basket is to be made explicit for bespoke models to be implemented.
Consider moving to a matrix data structure. A matrix might make some computations heavy, template specialization on the dimension might be an alternative to having two classes, one for scalars and another for matrices.
Rethink all the data structure with a basket where current losses are not zero.
In principle the 2D interpolator is left optional since there are arbitrage issues on the interpolator type to be used. However one has to be careful when using non local interpolators like CubicSplines which have an effect on the past (calibrated) coupons of previous tenors.