Real-life question for you Schedule P experts

[johnny storm]

Any help is appreciated. Maybe the exam committee will ask it. If you want to make sure you know the material, you should try this question (can I say anything else to egg you on?).




From

http://www.actuarialoutpost.com/actu...d.php?t=104802


Thanks, I'll post this in the Part 7 section as well.

I got most of what I wanted from the Feldblum paper, and thanks for pointing out that it was updated. I still have a couple questions, if you are familiar enough to answer them. I will post some numbers with my question, it might help.

I have a "prior" line in a 2005 AS, Part 2B, that show under the 2003 2004 and 2005 columns
4,232,029 4,380,663 4,506,820

1) I understand this to mean that these are the estimated ultimate losses for all accidents occurring prior to 1/1/1996 which are still open? It can't mean all accidents prior to that, but it's not clear to me exactly what accidents are intended to be included.
2) Assuming I'm right, how does a company calculate this? They didn't set up a projection for individual claims, so it seems like a completely bogus number to me. It seems like you just add in reserves ( = part 2 – part 3) for the ninth year each year, and subtract out the payments, and then consider that to be the new number.
3) This would negate any re-estimation of the ultimates, right? So, if a company has big changes in their estimated ultimate losses for, say, asbestos claims, then that would never be captured here? I don’t see how, given the instructions included in Feldblum’s note, which say that you take all of the numbers from the prior year’s AS? Or, does the far right column, which comes from Part 1 of the AS, allow you to bring in the re-estimated numbers?

The most important issue to me (this is all I really care about, but I feel I have to understand the other questions to understand this:
4)How do I measure adverse or favorable development? There is of course a one-year and two-year development column – can I just use those? Seems like I can’t, unless I misunderstand something in my questions above, because it’s not really measuring development in the usual way. (the numbers in those columns are the numbers you would expect based on the numbers above - 126,157 and 274,791)
What about prior years’ adverse development (all years)? It flows up, so it seems like if I’m wrong and the one-year and two-year numbers are right, then I can similarly do that for older years, and if they’re wrong, then I’m SOL.

[KidCA]

Originally Posted by johnny storm

I have a "prior" line in a 2005 AS, Part 2B, that show under the 2003 2004 and 2005 columns
4,232,029 4,380,663 4,506,820

1) I understand this to mean that these are the estimated ultimate losses for all accidents occurring prior to 1/1/1996 which are still open? It can't mean all accidents prior to that, but it's not clear to me exactly what accidents are intended to be included.

It's only fitting I give you a start. For the prior year's row in part two, you need the prior year's (PY) A.S. Determine the PY reserves for the priors row and 1st AY (Part 2 - Part 3). Add these to the current AS's Part 3. This is the current year's Prior Years on Part 2. The most recent year Part 2 is from Part 1.

I think the IBNR is getting whittled by adding on paid amounts during the most recent CY. However, per the PY part 2, you are adding in new IBNR for the 1st AY after the prior year's row.

Originally Posted by johnny storm

How do I measure adverse or favorable development? There is of course a one-year and two-year development column – can I just use those? Seems like I can’t, unless I misunderstand something in my questions above, because it’s not really measuring development in the usual way.

Part 2 has ultimate figure development. If there is consistent upward development (for any/all years), that could signal a problem. You can also incorporate Part 6 EP to look at loss ratios and see how they are developing left and right and up and down.

You can also use part 5 claim information with parts 2-4 to look at settlement rates and/or reserve strengthening... it's up to you what you want to look at. Schedule P just throws out possible tools to test different adequacies.