I've been told this is an obscure topic, but it came up on the Spring 2014 exam, so I decided to spend the time deciphering what the GAAP book has to say about how to calculate mortality margins. I finally figured it out by trying to do it myself, and then verifying that the GAAP book does it the same way (see attached spreadsheet).
In the Spring 2014 exam problem, there is a non-zero expected mortality margin in 2014 when calculated at the end of 2013. The GAAP book develops the mortality margin from the perspective that pricing mortality is used to calculate the annual payment, which implies that expected mortality margins equal zero. However, it is still possible to have a non-zero expected mortality margin if there is a mortality assumption update after the product is priced.
That's just one of many confusing aspects of the GAAP book's treatment of this subject. Hopefully, my spreadsheet sheds some light on all of this. But beware of spending too much time on this. It would be strange if the SOA asked another such question in the Fall sitting, wouldn't it?
Incidentally, it looks at first glance like the model solution to question 5c calculates the account value released on death differently from my spreadsheet. That's just because the expected account balance in the model solution is conditioned on the policy remaining inforce. My spreadsheet is actually consistent with the model solution.
:spnner:
In the Spring 2014 exam problem, there is a non-zero expected mortality margin in 2014 when calculated at the end of 2013. The GAAP book develops the mortality margin from the perspective that pricing mortality is used to calculate the annual payment, which implies that expected mortality margins equal zero. However, it is still possible to have a non-zero expected mortality margin if there is a mortality assumption update after the product is priced.
That's just one of many confusing aspects of the GAAP book's treatment of this subject. Hopefully, my spreadsheet sheds some light on all of this. But beware of spending too much time on this. It would be strange if the SOA asked another such question in the Fall sitting, wouldn't it?
Incidentally, it looks at first glance like the model solution to question 5c calculates the account value released on death differently from my spreadsheet. That's just because the expected account balance in the model solution is conditioned on the policy remaining inforce. My spreadsheet is actually consistent with the model solution.
:spnner:
Variable Payout Annuities having Life Contingencies
0 commentaires:
Enregistrer un commentaire