Global Bootstrap
Global Bootstrap is a direct subtype of Bootstrap
aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa aaaaaaaa
TYPE INCLUSION RELATIONSHIPS
AVAILABLE FUNCTIONS
AVAILABLE CREATE FUNCTION KEYS
TYPICAL OBJECTS OF TYPE Global Bootstrap
This type represents data needed in the specification of a global bootstrapping algorithm.
Corresponds to the QuantLib template class GlobalBootstrap
An object of this type can be optionally used as input in the creation of yield curves when Build Method = Global BS is used.
Concretely, it can be then set as the value associated with the key Bootstrap Spec
It accomodates the addition of m extra time pegs (curve nodes) t'₁,t'₂,...,t'ₘ that introduce additional degrees of freedom with respect to the specified modelled quantity x (eg the zero rate) in the bootstrapping process.
Note that the exact type of modelled quantity is identified during a curve construction through the entry Modelled Qty
Concretely, without the global bootstrapping, the modelled quantity x is linked only to the n variables x(t₁),x(t₂),...,x(tₙ), where t₁,t₂,...,tₙ are the pillar times that correspond to the market instruments, typically their maturities.
The global bootstrapping lets the user specify m additional times t'₁,t'₂,...,t'ₘ, which are then linked to the corresponding variables x(t'₁),x(t'₂),...,x(t'ₘ)
These additional times t'₁,t'₂,...,t'ₘ should be distinct from the regular instrument pillar times t₁,t₂,...,tₙ
The effect is that the total number of variables that can be freely set by the error-minimization algorithm becomes n + m
These extra variables are associated with additional constraints that need to be fulfilled.
Technically, these additional constraints correspond to one or more new error terms that are added to the standard error terms of the differences between the quoted and implied market prices of the supplied instruments.
At each step of the bootstrapping process, an error function is constructed that equals the sum of the squares of all "alive" error terms.
The goal at each such step is the minimization of that error function.
The exact nature of the additional constraints (i.e. their error terms) is defined by a choice from the list Global Bootstrap Method