PDA

View Full Version : Question 5


MGN
07-07-2009, 06:26 PM
This is fun.

So on this one, candidate 236 got .5 out of 2 and I think that makes sense. I'm guessing the .5 was for part b because none of the others are well explained or really make sense.

882, on the other hand, got 1 point, and I think this could be argued.

For part a, his answer is very close to the sample answer and doesn't seem to miss anything. +.5

Part b, it seems like he should get at least .25. He says "when there is a major growth in policy selling", which despite the awkward wording is close to the sample answer, though he does not say anything about the resultant shift in average accident date, which is important. +.25

Part c is tougher, he says "When there is a major change in economic environment because sometimes losses tend to emerge on a calendar year basis after, for example, news emerges about asbestos danger." The sample answer says something about changes in the social and legal climate causing severity to correlate more with report date. He doesn't quite say that it can cause severity to correlate with report date, but has the general idea right. I'd say +.25 here too.

For part d, he says "if there is a major change in reporting patterns...claim counts will distort projections." This is not what the sample answers provide, but it seems correct to me. If there's been a change in the claim reporting pattern but the exact nature of it is unknown, using earned exposures seems much wiser than using claim counts. +.5 if I'm grading it.

So I don't know, it's possible that the last question got 0 pts and the middle two got .25 each. If that's the case, I'd say he could have a case for an appeal. Of course, if he got his scoresheet and saw "50-74%", three months of waiting to see the sample solution would likely dull his memory of it well past the point where that's possible.

On the other hand, it could be that he got .5 for the last one but 0 for the middle two, in which case the rubrick must have not been assigning partial credit, period, because he seems to have the right idea for those but the rationale is poor in both cases.

Thoughts?

intelmic
07-08-2009, 11:59 AM
I agree with you.

Candidate 882 had a fair answer for part d) that isn't like the sample solution.

Part b): He knew about the book growing. Is it enough to get a partial credit. Is a partial credit even possible for this question... I would think yes as 0 to 0.5 is quite a step.

Part c): I think his answer matches enough the solution. It's just restated in his own words... In my opinion, "losses tend to emerge on a calendar year basis..." equals "closely correlate with the date of report of the claim".

I wish we had the sub-points grid for both candidates.

think888
07-16-2009, 06:08 PM
Agree. Harsh grading there. Maybe on (c) they didn't like his use of the words "calendar year basis".

In a perfect world, this release of the candidate papers needs to be accompanied by release of grader notes. Maybe next time!

Daniel Roth
07-16-2009, 06:29 PM
Reminder that part of this process will be to identify a few questions where we will go back and get the actual grading rubric. If you think this should be among those chosen, make a note.

I have not looked at the rubric for this question and do not know how/why the graders accepted specific items, or not, but my personal view on 882 part d. is that it did not receive credit because it is not correct.

The answer suggests that when there is a change in reporting pattern, a more accurate estimate of unpaid claim amounts will be derived from earned exposures (which presumably remained stable) rather than claim counts. The answer may sound plausible on the surface, but if there is a change in reporting pattern, the actuary needs to make adjustments in his estimation methods to reflect this shift, not find a way to ignore it. For example, if reporting has slowed, then IBNR needs to be higher than the data would otherwise suggest because, due to the shift, relatively more incurred claims are not yet reported. If a more stable method based on earned exposures is used, the IBNR will be understated.

PhildeAssociate
07-16-2009, 06:52 PM
Reminder that part of this process will be to identify a few questions where we will go back and get the actual grading rubric. If you think this should be among those chosen, make a note.

I have not looked at the rubric for this question and do not know how/why the graders accepted specific items, or not, but my personal view on 882 part d. is that it did not receive credit because it is not correct.

The answer suggests that when there is a change in reporting pattern, a more accurate estimate of unpaid claim amounts will be derived from earned exposures (which presumably remained stable) rather than claim counts. The answer may sound plausible on the surface, but if there is a change in reporting pattern, the actuary needs to make adjustments in his estimation methods to reflect this shift, not find a way to ignore it. For example, if reporting has slowed, then IBNR needs to be higher than the data would otherwise suggest because, due to the shift, relatively more incurred claims are not yet reported. If a more stable method based on earned exposures is used, the IBNR will be understated.


Dan,

Do they make you retake all the exams before you can become exam chair?

Regards,

PhildeAssociate

Daniel Roth
07-17-2009, 11:08 AM
Dan,

Do they make you retake all the exams before you can become exam chair?

Regards,

PhildeAssociate
Thankfully, no. Are you suggesting that they should have because my quoted reasoning is suspect?

PhildeAssociate
07-17-2009, 02:13 PM
Thankfully, no. Are you suggesting that they should have because my quoted reasoning is suspect?

No, I was actually semi-impressed that you would take the time to host an exam 6 seminar in this forum section.

Random forum posters are just looking for any reason to discredit the exam process, yet you write nice long posts teaching them about unpaid claim amounts and earned exposures.

Might as well start a thread in this section called - "Dan Roth teaches Exam 6". :tup: