The number of years for inflation in the IFOA answer used from the beginning of the year of cover - loss year (therefore it is, for example in the first line 1.05^(2020-2010) = 1.63). When I practice, I use 2021-2010, assuming claim is paid in 2020 policy on average at beginning of 2021. So my answer differs. Tell me if I am wrong? If I do it wrong that way, how many points I make, assuming all other steps the same with the IFOA?
As long as you make a sensible assumption, clearly state it and then carry out a calculation which is consistent with that assumption, you should still get full credit. The policy is written in 2020, so if we assume that claims occur uniformly over the policy period, the average claim occurs mid-2020. By assuming the average claim is paid at the beginning of 2021 (=end of 2020), you are effectively assuming a 6 month reporting and settlement delay, so as long as you state that you should be ok. However, you would also need to adjust your starting point accordingly though, unless of course you are assuming that future reporting and settlement delays differ to those in the past.
May I clarify, and tell me where am I wong: - The reinsurance policy written at BEG 2020 will cover for all insurance policy throughout 2020. - Policy are uniformly written in 2020, therefore mid point is 1/7/2020 - The claim delay is 6 months, therefore claim mid point is 1/1/2021 That's why I use 1/1/2021 Based on what you said, maybe you only care for the reinsurance policy and not the insurance policy?
Excess of loss reinsurance is usually written on a losses occurring basis. So the contract in the question will most likely cover losses arising between 1 Jan 2020 and 31 Dec 2020. So the average accident date will be mid-2020.
So a reinsurance treaty covering in 2010 will cover for all loss occurring in 2010, regardless of when the policy is written from the insurer's side right? Therefore the average accident date will be mid 2010? What if the insurer know in advance that they will have a claim in 2010, then buy the reinsurance to "fraud"?
Yes if the treaty is written on a losses occurring basis then that is exactly how it works. Have another read of Section 5.3 of Chapter 5. This last bit sounds very odd to me, how would an insurer know in advance that there would definitely be a claim in the future? They may expect to receive an unknown number of claims of unknown amounts, but if you know for sure that an event is going to happen then it breaches one of the criteria for the risk to be (re)insurable as per section 1.1 of Chapter 2.
