Actuarial Outpost
 
Go Back   Actuarial Outpost > Exams - Please Limit Discussion to Exam-Related Topics > CAS > CAS Exams > Exam 7 (new) - Estimation of Policy Liability & ERM
FlashChat Actuarial Discussion Preliminary Exams CAS/SOA Exams Cyberchat Around the World Suggestions

DW Simpson International Actuarial Jobs
Canada  Asia  Australia  Bermuda  Latin America  Europe


Reply
 
Thread Tools Search this Thread Display Modes
  #1451  
Old Yesterday, 04:31 PM
Abomb7894 Abomb7894 is offline
Member
CAS
 
Join Date: Jan 2014
Posts: 320
Default

Quote:
Originally Posted by DamSon View Post
Can someone explain why the number of testable correlations in Venter is n-3 choose 2? In a 4x4 triangle, n=4 so the number of testable correlations should be 0, but isn't there one set based on column 1 and 2 with 2 observations each?
Casual Fellow notes its not explained well, and says we want n-3 because we need at least 3 development factors and use choose 2 to ensure all column pairs, not just consecutive columns.

I can only assume you need at least 3 as T is distributed n-2, so you need some amount greater than that.
Reply With Quote
  #1452  
Old Yesterday, 07:28 PM
actuarialgenes's Avatar
actuarialgenes actuarialgenes is offline
Member
CAS
 
Join Date: Apr 2017
Location: New England
Studying for 7
Favorite beer: Wine
Posts: 140
Default

Quote:
Originally Posted by Abomb7894 View Post
The first list is from Marshall for Internal Systemic Risks.

The second risk is from Brehm talking about parameter risk I believe, but there's another list which is estimation, event, projection, and systematic. I think it's from two different chapters & is inconsistent.
Thanks!
Reply With Quote
  #1453  
Old Yesterday, 07:35 PM
actuarialgenes's Avatar
actuarialgenes actuarialgenes is offline
Member
CAS
 
Join Date: Apr 2017
Location: New England
Studying for 7
Favorite beer: Wine
Posts: 140
Default

Quote:
Originally Posted by DamSon View Post
Can someone explain why the number of testable correlations in Venter is n-3 choose 2? In a 4x4 triangle, n=4 so the number of testable correlations should be 0, but isn't there one set based on column 1 and 2 with 2 observations each?
I think it says somewhere in the text that you want to exclude columns with 2 (or fewer) elements. The reason (I think) is that the numerator in the T test statistic you calculate will be 0 (regardless of what you calculate for r).

So in an n x n triangle you have n-1 factor columns, and you cant use the last two (which have one and two elements), leaving you with (n-1)-2 = n-3 to choose 2 from.
Reply With Quote
  #1454  
Old Today, 01:23 AM
DamSon's Avatar
DamSon DamSon is offline
Member
CAS
 
Join Date: Sep 2018
Location: NYC
Studying for 7
Favorite beer: Alcoholic
Posts: 83
Default

Quote:
Originally Posted by Abomb7894 View Post
Casual Fellow notes its not explained well, and says we want n-3 because we need at least 3 development factors and use choose 2 to ensure all column pairs, not just consecutive columns.

I can only assume you need at least 3 as T is distributed n-2, so you need some amount greater than that.
Quote:
Originally Posted by actuarialgenes View Post
I think it says somewhere in the text that you want to exclude columns with 2 (or fewer) elements. The reason (I think) is that the numerator in the “T” test statistic you calculate will be 0 (regardless of what you calculate for “r”).

So in an n x n triangle you have n-1 factor columns, and you can’t use the last two (which have one and two elements), leaving you with (n-1)-2 = n-3 to choose 2 from.
Thanks everyone, what you guys said helps clear this up. As a follow up, are there any less common topics people feel like are likely to show up? In the middle of reviewing things like this right now, and I can't think of things past like table M, discounted reserves, and feldblum enhancement. Trying to go over problems I've done poorly on from the problem pack. Really need some new problems to work on now... tempted to do TIA exams 2-3 and just skip the BS questions

Last edited by DamSon; Today at 01:30 AM..
Reply With Quote
  #1455  
Old Today, 02:23 AM
redearedslider's Avatar
redearedslider redearedslider is offline
Member
CAS
 
Join Date: Oct 2015
Posts: 13,733
Default

Quote:
Originally Posted by DamSon View Post
Thanks everyone, what you guys said helps clear this up. As a follow up, are there any less common topics people feel like are likely to show up? In the middle of reviewing things like this right now, and I can't think of things past like table M, discounted reserves, and feldblum enhancement. Trying to go over problems I've done poorly on from the problem pack. Really need some new problems to work on now... tempted to do TIA exams 2-3 and just skip the BS questions
You did all the CF and RF problems? What did you think of the RF problem pack? Might have to skip it but I’ll skim if there’s good stuff that CF missed
__________________
Quote:
Originally Posted by Abraham Weishaus View Post
ASM does not have a discussion of stimulation, but considering how boring the manual is, maybe it would be a good idea.
Reply With Quote
  #1456  
Old Today, 02:57 AM
DamSon's Avatar
DamSon DamSon is offline
Member
CAS
 
Join Date: Sep 2018
Location: NYC
Studying for 7
Favorite beer: Alcoholic
Posts: 83
Default

Quote:
Originally Posted by redearedslider View Post
You did all the CF and RF problems? What did you think of the RF problem pack? Might have to skip it but I’ll skim if there’s good stuff that CF missed
Yeah, I've done all the calculation problems from both twice as well as all the past exam problems in the CF manual once. I also re-do questions I didn't do well on from both every few days, so there are some problems I've done around 6-7 times.

I like the problem pack a lot, there are a lot of good questions in there. CF problems take a lot of time to do since they're pretty comprehensive and test multiple areas at a time, but the RF problems test one topic at a time and are more in the format/wording of the exam imo. If you're looking to test your comprehension, you can't go wrong with either. However if you want more exam style prep questions, it's worth skimming through the problem pack on topics you don't feel as comfortable on.

For Brehm, I think CF is better than RF if you repeatedly do the essay questions, but I made flash cards for all of those so it takes less time cause writing out 40 shapland/brehm essay questions would just kill my hand. However, the RF questions on Brehm help to synthesize what you know in theory and apply it to specific scenarios he gives in many cases.
Reply With Quote
Reply

Tags
brehm is my daddy, cfrf needs a forum, cs was here, firsttag, hi cs, mack 69

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off


All times are GMT -4. The time now is 04:52 AM.


Powered by vBulletin®
Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
*PLEASE NOTE: Posts are not checked for accuracy, and do not
represent the views of the Actuarial Outpost or its sponsors.
Page generated in 0.27154 seconds with 10 queries