Hello, I'll be grateful if the tutors could help clarify the following queries on the solution in the study material. Case Study 2 part (iii): New administration system I agree that the inflation-adjusted chain ladder (IACL) method is unaffected by the change. In the average claim size method (ACPC), the solution says: “the avg claim size factors will have to be increased if the level of prudence is to be maintained.” Is it the case that we think those claims marked as “settled” are not truly settled, i.e. have potential to reopen? Otherwise I don’t see a reason to hold extra reserves than what the method suggests... Then the solutions goes on: “to do this, the company could obtain new “as if” claims triangles, as if the new admin system had always been in place”. My understanding of the ACPC method is: use triangles to project a) the number of all claims, and b) the total claims, by AY. These are projected to ultimate and after adjusting for inflation etc., the ratio of b) over a) gives the ACPC. This process is unaffected by the change in the admin system (marking claims as settled), in the same way that the IACL is unaffected by the change. Please clarify? In the approximate method to estimating the ACPC, it appears the method produces an (increased) ACPC that would produce the same reserves as held last year after considering the reduced number of OS claims as a result of the change. So we want the level of reserves held to be unaffected by the change, hence “maintaining the level of prudence” earlier. Would this be correct? Many thanks in advance!
Not necessarily, that's not really the point. It doesn't matter whether the reserves will have to be used to pay open claims or reopened claims. The point is, if we use the same multipliers as before but apply these to a lower number or claims, the ultimate claims estimate will be lower. If the company was using a standard average cost per claim method, yes. But that's not what the question tells us, it gives us a description of "a sort-of-weird-sounding-ACPC-method". This is typical of the jump from ST7/8 level to SA3 level; you have to consider less standard scenarios. The examiners made several comments about this in the April examiners' report. For example, they said: "... take the time to thoroughly think about what is actually going on in question situations proposed rather than simply considering [...] standard techniques ..." and: "Many candidates still struggle to tailor their answers to the specifics of the question (potentially not even reading the question thoroughly enough to identify the specifics)." Yes. If we accept that the previous method gave an appropriate level of reserves, then the new method will only be valid if it gives a similar level of reserves.
Many thanks Katherine! I've let your feedback to "sink in" over the days before posting... 1) I understand the claims reserve will be lower. My thinking is: why is this undesirable? The method is: No. of open claims X ACPCfactor - sum of paid on those claims. If the factor represents the ultimate claim amount for a claim, and we assume there is some prudence built in the factor, then yes a drop in no of open claims and the sum paid will be observed leading to a lower reserve (due to prudence in factor). But if we believe those claims that are marked as settled are truly settled, then we need not worry, need we? Why should we increase the ACPC factor? 2) If this is indeed a non-conventional ACPC method, how can the examiners expect us to know the process of how it is carried out, without giving more details? I say this because there is reference to as-if triangles with the new admin system, which I can only possibly comment upon if I understand the process. And no amount of rereading the question will help! It feels anyone can set questions that discuss briefly about non conventional methods and expect candidates to comment on aspects without fully understanding the mechanics. I think it takes more work to set fair questions. Anyway, appreciate your help in trying to shed some light on this. Other fellow students: if anyone can spot flaws and correct my thinking would be most welcome and appreciated!
1) Don't overcomplicate this. By applying the original GUF to a lower figure, you'll get a lower reserve. If you assume that the original method gave a reasonable reserve, it follows that the new reserve must be too low. 2) It's perfectly reasonable to expect a qualified actuary to able to decide on the suitability of any method. So that's what the examiners want to test you on. A good exam technique is to: read the details discuss any ambiguity in the way the examiners have worded the question discuss the effect on the reserves discuss the limitations of the method.
Many thanks Katherine. Appreciate your taking the time to explain. Yes, I'd agree. I was just attempting to understand the full picture around the mechanics, but that can as you say "overcomplicate".
