• We are pleased to announce that the winner of our Feedback Prize Draw for the Winter 2024-25 session and winning £150 of gift vouchers is Zhao Liang Tay. Congratulations to Zhao Liang. If you fancy winning £150 worth of gift vouchers (from a major UK store) for the Summer 2025 exam sitting for just a few minutes of your time throughout the session, please see our website at https://www.acted.co.uk/further-info.html?pat=feedback#feedback-prize for more information on how you can make sure your name is included in the draw at the end of the session.
  • Please be advised that the SP1, SP5 and SP7 X1 deadline is the 14th July and not the 17th June as first stated. Please accept out apologies for any confusion caused.

Exam Style Question, CH4 - Reinsurance

V

veeko

Member
The exam style question in Chapter 4 part (ii) asks us to calculate the likelihood of the data and gives information that 10 claims were received, 4 of which involved the reinsurer.

In the solution, it considers the 6 claims which are <M and do not involve the reinsurer. ie. they calculate Pr(X<M)^(n-m). However, in the core reading on page 24 of Chapter 4, to calculate the likelihood, they consider the claims that do involve the reinsurer, ie. they calculate Pr(X>M)^m.

Is my reasoning correct that one of these has to be wrong or am I confused somewhere along the line.

Thanks.
 
If a) I understand what you are asking and b) my memory of that material is sufficiently good (I got an exemption from Uni),

I would have thought that it would have involved a product of the 2 items you have there along with a 10 C 4 factor.
ie P(X<=m)^(n-x)*P(X>m)^x* (n C x), observed value of x is 6, n=10.

Of course if you (or the question) mean by "calculate the likelihood", calculate a function of the likelihood in order to maximise it (rather than the absolute number = likelihood), then you could always knock off constants like the n C x.
(I originally had 10 C 4 but thought that it should be n C x, which then made it non-constant. However I think in maximising likelihoods the x is known and then we can treat it all as constants, The variable being the parameters of the distribution)

Who knows (and I'm really guessing here), if you play with it long enough you could even find a constant to knock off that allows you to just include the P(X>m)^x

Hope that helps some. Does that make any sense at all? The more I think about it myself the worse it seems to me.
 
Last edited by a moderator:
The exam style question in Chapter 4 part (ii) asks us to calculate the likelihood of the data and gives information that 10 claims were received, 4 of which involved the reinsurer.

In the solution, it considers the 6 claims which are <M and do not involve the reinsurer. ie. they calculate Pr(X<M)^(n-m). However, in the core reading on page 24 of Chapter 4, to calculate the likelihood, they consider the claims that do involve the reinsurer, ie. they calculate Pr(X>M)^m.

Is my reasoning correct that one of these has to be wrong or am I confused somewhere along the line.

Apologies for the delay Veeko. I had referred this to my colleague - and then have just found out that they hadn't replied.

The CR question has information about the insurers claims and not about the part paid by the reinsurer. That's why they use the PDF for the insurer's part and just the probability for the reinsurer's part.

The exam style question at the end has information about the reinsurer's claims but not the insurer's claims. Hence it is the other way round.

In summary, we use the PDF for the bit we know about and the P(X>M) or P(X<M) for the bit we don't have the full information for - the "censored" bit - where we only have the number of claims but not the amounts.

Hope this clears this up.
 
Back
Top