Survey

[Sew Lining]

Anyone want to argue about the quality of the exam before taking the survey?



QUALITY OF THIS EXAM

SCALE OF 1 TO 5
1 Coverage of Syllabus Material 1 = INADEQUATE
2 Exam Clarity 1 = NOT CLEAR
3 Exam Length 1 = TOO SHORT
4 Exam Difficulty 1 = EASY
5 Overall Quality of Exam 1 = LOW


FREE FORM
6 Regarding the quality of this exam (not the syllabus readings), what specific comment or suggestions do you have? (Comments on possible errors in specific questions on the exam should be sent to the CAS Office as soon as possible in a separate letter.)


SELECT ONE
7 Please select one of the following:
Could attempt all questions, review some answers, and still had time left over.
Could attempt all questions as well as had time to review some answers.
Could attempt all questions but not enough time to review any answers.
Not enough time to attempt all questions.
Not enough time to answer questions that I felt I had the knowledge to answer.



QUALITY OF THE SYLLABUS
8 What readings were particularly valuable?
9 What readings were of questionable quality/value?
10 Do you have any other comments on specific readings or the syllabus in general?


QUALITY OF THE EXAM SITE
11 *At which site did you take the exam? Please provide specific comments you may have about the exam site?


GENERAL COMMENTS
12 Do you have any other comments or suggestions about this exam?

[Sew Lining]

There was one question asking about the predictability of excess losses.

I think this question is problematic because two papers contradict each other:

Teng, pg 426 item 3.a.
...excess loss is harder to estimate than total loss.

Venter, pg 5
This makes excess losses less heavy-tailed and thus more predictable than the total losses.

I gave both arguments and cited which paper they came from, but I don't see how the CAS can mark one answer wrong and one correct. IMO the problem with this question is that from my experience, excess losses are harder to estimate due to the variability of higher layers. I think most actuaries would agree. However, the only paper (to my recollection) that makes an argument one way or the other was venter (he says they're easier), so I assume that's what they were looking for.

Anyone find other possible problem questions?

[Redhead]

Brutal. That's the only word to describe that exam. I didn't have enough time to finish, and that was even skipping the ones I didn't know. Far, far too long for a 4 hr test imho. Brutal.

[tommie frazier]

i felt well prepared, and I found it reasonable. I had time. not plenty, but enough. got through it all and had about 15 minutes left. put papers in order, scribble da few formulas on the things I skipped earlier, and fixed one error I found when stacking the papers (as opposed to actually reviewing the work in detail).

I note sew's issue above, as I wrote the answer that went with the other part of that one (I think it was the LDD pricing issue of it being harder to predict XS) which within that paer is clear. but the other paper, which they had us write out in this exam, also noted that it is less easier in some regard. I do find that one to be something they might have issues since the answers are not required to be constrained by what this author said.

[ramanujan]

I thought the exam was reasonable as well. I had time to attempt every question though I left part b or c of a few questions here and there. I tried to write down something on each question even if It was BS because I knew I would not have time to come back.

I thought the same thing as Sew when I read the predictability question but since the question was worded as "for the purpose of experience rating", I assumed this question came from Venter's article and wrote what he said. To confuse matters, Q34 asked why LDD is riskier? one of the reasons is exactly opposite of what Venter says but I wrote what LDD article said. So basically I said two opposite things in these two answers because I think this was safe and this is what they wanted but I agree that both arguements should be fine.