Thank you for that screenshot! I don’t have TIA but would curious to how the TIA instructor would answer the points that piotr raised earlier. Maybe @SteveRoll could weigh in?
I think if Z is not limited by 1, we also allow the Neuhaus Reserve to be negative which doesn’t make sense on paid data.
If we consider an incremental paid triangle 2x2:
4 1
1
and make V1=1 and V2=1 then we get
m1=5/2
m2=1
ELR=7/2
p=5/7
q=2/7
Z(Neuhas)=5/2
R(ind)=2/5
R(coll)=1
R(Neuhaus)=-1/2
I’ve placed the question in the TIA discussion section. I’ve noticed her get back within a day before. Will report back.
2 Likes
@piotr @joejam
This was my understanding from when I used Rising Fellow in 2019-2020. I would probably make a note like “I recognize this assigns credibilities of >1 and <0 and would choose to use perhaps the Benktander method with credibility = p instead, but will complete the problem as written.”
2 Likes
With the new syllabus format being less than helpful, is anyone doing anything different compared to prior sittings to focus more on “Understand and Apply” and “Analyze and Evaluate”? I’m curious how these percent breakouts compare to prior exams. Seems like 0-10% “Remember” is light compared to prior exams I’ve taken.
1 Like
The way I understood it, and I could be very wrong, is that for the understand and apply section instead of asking to list the 3 Mack assumptions like they have in the past, they will ask what assumption is Venter correlation or Spearman testing.
For the analyze and evaluate section they would ask you to look at a graph of residuals and draw conclusions instead of just asking what increasing residuals means.
2 Likes
I expect a part of that exam to be similar to MASII (single/multiple choice questions requiring in-depth understanding of the concepts in the source material).
When using the least squares method in Brosius, should LRs be used if the premium is unstable or only if the book is shrinking or growing?
In conclusion the paper mentions adjustments for “significant exposure changes” and I understand it as premium instabilities.
If the intercept from least squares is different from 0, then using it in a low premium accident year seems to be out of proportion, whereas adjusting the triangle for exposures makes it commensurate.
I think that in Venter the triangles for additive methods (Cape Cod, Additive Chain Ladder) were adjusted for exposures, only the BF wasn’t, because it’s allowed to have different ultimate levels for different years.
1 Like
Improved Pearson testing environment, still no F4 though
With the new functionality, we still wouldn’t be expected to produce graphs right? Atleast for this sitting?
I don’t think so since they made no mention of that in any of their communications and did not give an example of creating a graph as a new question type. This is just my opinion.
- "Graphing/Drawing: Candidates will not be expected to do any graphing or drawing in their responses on the exams.
From here under Sample Exam.
2 Likes
Does anyone have a link to the updated or old blank Pearson Vue spreadsheet? I remember it was under “CAS sample testing” on Pearson / CAS website, but I just see a link to the new item type samples there.
EDIT: Question 12 in the new item types is just a blank spreadsheet. I also think that they updated the question for exam 7 - now it’s the least squares method calculation.
Did you see the typo in problem 11? They reference AY2020 on the left side of the screen but then in the spreadsheet it says 2021 for the year corresponding to the ultimate link ratio estimate that provided. I’m assuming they meant to say 2020, right? Like if you do the Least Squares method, you get that a < 0, so you have to use the link ratio method’s ultimate instead for 2020?
Edit - I saw the solutions they posted… they acknowledged the typo!
Where did they post the solutions?
It should be the 2nd link!
1 Like
When I look at the questions that have an excel area to solve I only see 3 or 4 rows. Does anyone else have that issue?
The testing window starts tomorrow, so best of luck on your exams everyone!
3 Likes
Good Luck, everyone! see you on the other side 
5 Likes