claims made paper exam 5

[ed999]

is anyone else finding this paper a bit difficult? especially the section about calculating earned exposures.

gosh i hate this exam. so much to do in so little time... =(

[Wigmeister General]

What's difficult about calculating earned exposures?

[Shrek]

gosh i hate this exam. so much to do in so little time... =(
Yeah? Well, they only get easier :roll:

[Wigmeister General]

Shrek,

:rofl: x 100,000,000

[Shrek]

What can I say? I know you can't be everywhere at once, and thought I'd help out.

[aniasat]

DId anyone understand how they have calculated the last row of the exhibit for the third principle where they have a lag of $20 per year per lag?

Please explain....this is killing me since yesterday.

anita

[ed999]

Hey Aniasat,

What are you up to wrt to the ALL 10 Suggested syllabus? And are you still in the Part 5 Yahoo Study Group?

So far, I have most of the week 1-7 material done. By "done" I only mean tht I have READ the papers. And I still haven't read Krakowski's paper and HO Insurance to Value. I do however have read and done problems for Ratemaking, Schofield's Pure Prem to Rate, Risk Classification, ILF statistics by Lang, Deductible SN and most of Complement of Credibility. I have more than a shallow understanding of CGL Ratemaking, and Bouska's Expsure base paper.

As of now, I am re-doing WC ratemaking and have just read Claims Made Paper twice. After "week 7" I jumped to week 10 because I think Ratemaking should be combined with ratemaking. So my plan next is to read Risk Rating, Asset Share Pricing and Hurricane modeling and the more Mathematical Insurance to Value. I took a cursory look at the CPCU style papers on concepts of Insurance and don't know how I will get through them and memorize everything! I don't know how I can get through this and with the lack of CAS people here, I can't even get enough "i can identify with you" responses!

ALAS!

Anyway, look forward to hearing your side of the story.

[ed999]

oh and by the way, I will TRY to answer your question abou tthe claims made paper once I understand it! The second read was much better than the first. But there are still fundamental concepts which I dont' understand about Claims Made and Occurrence. Forgive my ignorance.

[Utanapishtim]

Aniasat: This is explained by the note beneath Figure 1...
NOTE: For convenience, we do not show lags greater than four years in the matrix. In practice, we group all losses with longer lags in the LAG 4 row.

So as we go from left to right by column in Figure 7, we add $20 to each row for inflation, then shift $20 from each row to the one beneath it for the change in report pattern. Since the fifth row contains the lag of 4+ years, the $20 being lagged further from it is retained in the same group, and it continues to add the $20 per year of inflation. Thus this row is increasing by $40/year, while the first row (zero year lag) remains constant because the changing report pattern to later years exactly offsets the inflation, for the first year.

Sound good? (:

[Wigmeister General]

DId anyone understand how they have calculated the last row of the exhibit for the third principle where they have a lag of $20 per year per lag?

Please explain....this is killing me since yesterday.

anita

Ask Marker & Mohl. They understand.

[Avi]

OK, I'll give it a shot. Excuse the formatting errors when pasting from excel.

This is what the tableau SHOULD look like, assuming a $20 per year increase in loss costs, with historical in Olive and Occurrence form Bolded.

200 220 240 260 280 300
200 220 240 260 280 300
200 220 240 260 280 300
200 220 240 260 280 300
200 220 240 260 280 300

Now, as their is a report lag of $20/yr per lag, that effectively means we take $20 from L_0,i and add it to L_4,i. But it is cumulative. So the first 220 becomes 200, and the bottom row becomes 240. Now going left-to-right, we expect $20 more of loss, so 240 becomes 260. But the 240 at the top loses $20 (and becomes 220) which flows through to the last lag for that year which becomes280. Now 280+20 = 300 due to regular loss inflation. Factor in another 20 from report lag, and we get 320. And so the pattern continues



200 200 220 240 260 280
200 220 240 240 280 300
200 220 240 260 280 300
200 220 240 260 280 300
200 240 280 320 360 400

Hope that helps.

[aniasat]

Thanks Avi,

My understanding of the whole table was right except that I didnt interpret the assumptions of the paper correctly.

This was very helpful.

Anita

[ReserveRage]

Avi's first row doesn't match with the paper (Utan's explanation).

I'm assuming the difference is that Avi projects loss trend first for all 5 years and forgot to adjust for settlement in year 1, then project.

Because we saw the 220 in L(0,1) change to 200 in report yr 1, basically telling us that claims are taking longer to report than anticipated, we'd expect 220 in L(0,2) not 240, with the +20 coming from inflation only. But because the claim reporting shift is in every report year, the expected 220, drops back down to 200. This method produces the constant 200 across row 0. Every year the bottom row gets this +20 from a report shift AND +20 from the expected inflation.

It's somewhat confusing because the paper doesn't specifically mention this.

[bokiz333]

I hope someone can help me out (mainly I hope some tells me that the paper is wrong but an explanation would be great in case its not)

Figure 8 on page 277 shows $400 under column 5 (in the example where they combine the two errors together). The Occurence premium total is listed as $1,530. When I add the diagonal I only come up with $1,480.

If I understand this paper correctly then the last column should actualy be $450 and the $400 is there in error. if that is the case that this paper makes perfect sense. If not then I just wasted a few hours and hope some here can please enlighten me. Thanks in advance.

[Commander Keen]

I hope someone can help me out (mainly I hope some tells me that the paper is wrong but an explanation would be great in case its not)

Figure 8 on page 277 shows $400 under column 5 (in the example where they combine the two errors together). The Occurence premium total is listed as $1,530. When I add the diagonal I only come up with $1,480.

If I understand this paper correctly then the last column should actualy be $450 and the $400 is there in error. if that is the case that this paper makes perfect sense. If not then I just wasted a few hours and hope some here can please enlighten me. Thanks in advance.

FYI...

Errata for
“Rating Claims-Made Insurance Policies”
By J. O. Marker and J. J. Mohl
Pricing Property and Casualty Insurance Products
Casualty Actuarial Society Discussion Paper Program, 1980

Page 277 –
L5,5 should read $450 instead of $400.

Page 282 –
Example 3 should read:
2/3 exposure to E i, j for i = 1, 2, 3, (not i = 0)
1/3 exposure to E i, j+1 for i = 1, 2, 3, (not i = 0)
8/9 exposure to E 0, j
1/9 exposure to E 0, j+1

Page 289 –
In the third of the three equations, the final subscript on "x hat" should be j.

Page 294 –
In the "additional report" equation near the top of the page, a
summation sign should appear in front of the L i,1.

:tup: