PDA

View Full Version : What were the questions again?

Bobby
05-05-2011, 07:30 PM
Below are the problems that I remember (not in any particular order).

If you can add any problems or details to problems, please do and I'll edit the OP. I think there were 37 problems in total. I added some points-per-problem estimates, but not confident on these.

1. ISO PAM problem (1.5 points)
2. Two competing companies. One used classification analysis and other not. Calculate profit and explain how one company can gain market share. (3 points)
3. Essay question that listed law changes / book changes and you estimated the effect on reserve estimates (3.25 points)
4. Another essay question on reserve estimates: given true IBNR and estimates of various methods, explain differences (2.5 points)
5. Additive ALAE problem (2.5 points)
6. S&S recoveries question (2 points)
7. WC Experience Mod Quesiton, calculate mod, standard premium, and retro premium (2.75 points)
8. Given Base Rate, revised relativities, calculate EP for a calendar year and On-level factors (3.25 points)
9. Berquist-Sherman Reserve Problem - estimate ultimate claims (3 points)
10. Credibility-Weighted Classification Relativities (4 points)
11. Operating Expense Ratio Problem (1.25 points)
12. Claim Development / B-F IBNR (2.5 points)
13. Cape Cod Problem (2 points)
14. Is Paid Development / Reported Development Speeding Up (1.75 points)
15. Calculate exposures for CY, PY (2 points)
16. Calculate losses for CY, PY (2 points)
17. Make Reported Triangle (2 points)
18. Asset Share Pricing Problem (3 points)
19. Car years vs miles driven exposure base (2 points)
20. Development and Trend Overlap Fallacy (1 point)
21. Calculate large commercial insured rate discount based on expenses being reduced as premium increases
22. Big Rate Change Problem (6.75 points)
23. Reinsurance XS of loss vs QS problem (1.5 points)
24. Premium Trend Problem (2.5 points)
25. Conger ULAE Problem (1.5 points)
26. Expected Claims Method Problem
27. Two Purposes of Risk Classification (2 points)
28. Loss Elimination Ratio from deductible to larger deductible (1.5 points)
29. Discuss GLM output problem (1 point)
30. Excess loss trend (1 point)
31. Tail Severity Problem (2 points)

Inconceivable
05-05-2011, 07:33 PM
This isn't what you want to hear probably but isn't the exam going to be out pretty soon?

Bobby
05-05-2011, 07:34 PM
This isn't what you want to hear probably but isn't the exam going to be out pretty soon?

Yeah, hoping. But just couldn't wait....

Number Theory
05-05-2011, 07:39 PM
do they post the exam?

earthsunmoonsky
05-05-2011, 07:43 PM
I got my exam (5b) in the mail today. I'm too lazy to type them all up, but if you want to know what a particular question was I'll type it up for you.

Bobby
05-05-2011, 07:45 PM
Wow that was quick. Any way you could just write a very short description and the point value for each problem? Something like "ISO PAM Problem (1.5 points)" would suffice.

Vorian Atreides
05-05-2011, 08:03 PM
Exam will be posted late next week or early the following week.

actuarialista
05-05-2011, 08:09 PM
Wow that was quick. Any way you could just write a very short description and the point value for each problem? Something like "ISO PAM Problem (1.5 points)" would suffice.

I just got my exam back and quickly typed this up, so there are probably mistakes, but here goes:

1. ISO PAM (1.5)

2. Auto exposure (1.5)

3. Car years, written, earned, etc. (1.25)

4. Auto with classes and base rate increase, EP and on-level (1.5)

5. Two-step prem. trend (2.25)

6. HO incurred losses, CY, AY, etc. (2.5)

7. Overlap fallacy (1)

8. OER (1.25)

9. Rate indication (6.75)

10. LR vs PP method (1.5)

11. Companies A and B, new rating variable (2.75)

12. Two purposes of risk classification (1)

13. GLM (1)

14. LER (1.5)

15. Indic. relativities, classes have different trends (3)

16. WC prem discount (1.75)

17. Excess trend (1)

18. Base rate w/ min. prem. (1.25)

19. WC experience rating (2.5)

20. Asset share (3.5)

21. Combining two lines (2)

22. Construct cumulative AY reported loss triangle (2.25)

23. Given triangles, is claim closure speeding up? (1)

24. Reported devel. and tail (3.5)

25. ECT (2.5)

26. Estimate tail (2.5)

27. BF and CC (3)

28. Frequ-sev. (1.75)

29. Case o/s devel. (1.25)

30. Given avg case resv and cumul. pd, are devel. methods suitable? (2.5)

31. Berquist-Sherman (3.75)

32. S&S (3)

33. Quota share vs. XOL (1.5)

34. ALAE devel. (1.75)

35. Conger-Nolibos (2)

36. Impact of various changes on devel. methods (3)

37. Books A, B, C, D, IBNR by various methods, discuss possible changes (1.75)

Bobby
05-05-2011, 08:17 PM
I just got my exam back and quickly typed this up, so there are probably mistakes, but here goes:

Thank you!!!

Number Theory
05-05-2011, 08:20 PM
I dont remember number 10... I hope I didnt skip one and then misnumber the remaining problems cause that would suck!!

earthsunmoonsky
05-05-2011, 08:21 PM
What'd you guys put for 30 -- Given avg case resv and cumul. pd, are

a) rptd devel. methods suitable?
b) pd devel. methods suitable?
c) expected claims method suitable?

Bobby
05-05-2011, 08:24 PM
In that problem, wasn't the trend for the last avg paid and case a lot different than historical? If so, I put:

rptd dev not suitable because latest avg case was crazy.
pd dev not suitable because latest avg paid was crazy too.
expected claims suitable if correct lr is picked.

earthsunmoonsky
05-05-2011, 08:28 PM
I think it was only the avg case that was crazy... I calculated age-to-age factors for each, and for case O/S the last valuation was very high.

For paid the age to age was the same all the way down.

So I put, no for a, and yes for b. I don't remember c.

Should we have calculated something other than age to age?? -- I thought that's all they tested on this exam ;)

Number Theory
05-05-2011, 08:31 PM
esms-that's what I put as well.

Bobby
05-05-2011, 08:31 PM
Oh, maybe it was just case avg that was odd. in that case i said paid would be alright.

But I calculated the trend downwards (at each maturity), i.e. the 2008 avg case @ 12 mos, the 2009 avg case @ 12 mos. I didn't look at trend sideways (e.g. age-to-age).

booyah81
05-05-2011, 10:25 PM
The LR vs PP question was just when is one more appropriate. I think one was commercial complex exposures, one was two new rating variables, might've been a third one.

CASualty
05-06-2011, 12:34 AM
The LR vs PP question was just when is one more appropriate. I think one was commercial complex exposures, one was two new rating variables, might've been a third one.

Complex Exposures should be the LR method (since PP requires a clearly defined exposure in order to calculate current average premium).

Two new rating variables . . . I put LR Method because it isn't distorted and accounts for correlation between rating variables better. It was really a guess though.

I think the third might have been something like a new LOB? If so, I put PP Method because LR Method requires an existing rate.

FourKicks
05-06-2011, 12:47 AM
Two new rating variables . . . I put LR Method because it isn't distorted and accounts for correlation between rating variables better. It was really a guess though.

for this one i put PP method, my reasoning being that new rating variables will entail rate changes, which means if you use LR method you'll have to adjust premium to put it on-level which could be computationally intensive.

i don't believe exposure correlation is a concern when calculating overall indicated rates, only when classifying risks.

CASualty
05-06-2011, 01:39 AM
for this one i put PP method, my reasoning being that new rating variables will entail rate changes, which means if you use LR method you'll have to adjust premium to put it on-level which could be computationally intensive.

i don't believe exposure correlation is a concern when calculating overall indicated rates, only when classifying risks.

I like your answer . . . that seems correct. I was just guessing, and my "correlation" comment applies to classification, not overall indications. Oh well . . . I wasn't expecting a point there anyway :)

booyah81
05-06-2011, 08:25 AM
for this one i put PP method, my reasoning being that new rating variables will entail rate changes, which means if you use LR method you'll have to adjust premium to put it on-level which could be computationally intensive.

i don't believe exposure correlation is a concern when calculating overall indicated rates, only when classifying risks.

That's what I put.

JasonScandopolous
05-06-2011, 08:42 AM
That's what I put.

ditto. From the W&M readings, this is almost certainly what they wanted as the answer.

Bobby
05-06-2011, 08:47 AM
for this one i put PP method, my reasoning being that new rating variables will entail rate changes, which means if you use LR method you'll have to adjust premium to put it on-level which could be computationally intensive.

i don't believe exposure correlation is a concern when calculating overall indicated rates, only when classifying risks.

That's what I put.

ditto. From the W&M readings, this is almost certainly what they wanted as the answer.

Fourthed.