I'm using the CF manual and in the section on the Patrik paper there are both original questions as well as past exam questions which require you to allocate an aggregate reported loss number for a "Policies Written" reinsurance treaty to the appropriate accident years.
In CF's model solution, as well as in one of the model solutions in the examiners report (2012 exam, q 10), the reported losses are allocated to AY purely by the parallelogram method (ie, the share of loss allocated to an AY is equal to the AY's share of the total area of the parallelogram).
However, doesn't this overlook the fact that older AYs should be allocated a larger share of the loss given that they are more mature? Ie, even if 2 AYs have the same area within the parallelogram, wouldn't you want to assign more of the reported loss to the older year.
And as such, the loss shouldn't be allocated based on [Share of Area] but rather something like [Share of Area]*[% Reported].
Is this just a nuance which is assumed to be immaterial and is ignored for exam purposes?
In CF's model solution, as well as in one of the model solutions in the examiners report (2012 exam, q 10), the reported losses are allocated to AY purely by the parallelogram method (ie, the share of loss allocated to an AY is equal to the AY's share of the total area of the parallelogram).
However, doesn't this overlook the fact that older AYs should be allocated a larger share of the loss given that they are more mature? Ie, even if 2 AYs have the same area within the parallelogram, wouldn't you want to assign more of the reported loss to the older year.
And as such, the loss shouldn't be allocated based on [Share of Area] but rather something like [Share of Area]*[% Reported].
Is this just a nuance which is assumed to be immaterial and is ignored for exam purposes?
Parallelogram method applied to reported losses
0 commentaires:
Enregistrer un commentaire