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


Upload your resume securely at https://www.dwsimpson.com
to be contacted when new jobs meet your skills and objectives.


Reply
 
Thread Tools Search this Thread Display Modes
  #1001  
Old 06-16-2020, 11:40 AM
trueblade's Avatar
trueblade trueblade is offline
Member
CAS
 
Join Date: Jul 2012
Location: Castlevania
Studying for Freedom
Posts: 1,296
Default

Quote:
Originally Posted by jumpyshrimp View Post
do you find it harder than 8 & 9? I'm thinking about doubling up 7 & 8 but not sure how hard 8 is. I'd love to learn 9 but the exam date is too close.
Some other considerations:

1) How many people on your team are doubling up exams? If everyone else is doubling up and you are the only who is not doing so, would you be handling all their work? If that's the case, maybe you should be doubling up as well?

2) Study days/hours: How many study days are you getting if you were to double up? Does your company even allow students to write two exams?

3) Job security: Is there a possibility that your company will cut people in the near future? Maybe you should focus on work (and not doubling up) for now to reduce the chance of being laid off?

I raised these points because where I work, pretty much everyone is planning to double up. I'm not even sure who's going to do all these work in fall LOL
__________________


ACAS 7 8 9

Last edited by trueblade; 06-16-2020 at 11:44 AM..
Reply With Quote
  #1002  
Old 06-16-2020, 12:00 PM
SkolChicago SkolChicago is offline
Member
CAS
 
Join Date: Apr 2015
Posts: 1,278
Default

Shapland, regarding negative incremental values for entire diagonal. Is the r* in formula 4.5 (pg 21) just formula 4.4? I don’t see r* anywhere else.
__________________
ACAS 7 8 9
Reply With Quote
  #1003  
Old 06-16-2020, 03:11 PM
Death by Exam Death by Exam is online now
CAS
 
Join Date: Feb 2019
Posts: 17
Default

Quote:
Originally Posted by SkolChicago View Post
Shapland, regarding negative incremental values for entire diagonal. Is the r* in formula 4.5 (pg 21) just formula 4.4? I donít see r* anywhere else.
I believe r* is still just your sampled residual as normal.
__________________
P, FM, IFM, MAS-I, MAS-II, Exam 5
Reply With Quote
  #1004  
Old 06-16-2020, 04:46 PM
SkolChicago SkolChicago is offline
Member
CAS
 
Join Date: Apr 2015
Posts: 1,278
Default

For Clark, why doesn't Cape Cod scale up the unpaid % by the (inverse of the) growth function at the truncation point? It seems that if you just take the difference between the growth functions at truncation and at the average age, your reserves would be understated.

I suppose the answer may just be "because the source says so".
__________________
ACAS 7 8 9
Reply With Quote
  #1005  
Old 06-16-2020, 07:24 PM
AbedNadir's Avatar
AbedNadir AbedNadir is offline
Member
CAS SOA
 
Join Date: Mar 2014
Studying for FCAS
Posts: 2,881
Default

Quote:
Originally Posted by SkolChicago View Post
For Clark, why doesn't Cape Cod scale up the unpaid % by the (inverse of the) growth function at the truncation point? It seems that if you just take the difference between the growth functions at truncation and at the average age, your reserves would be understated.

I suppose the answer may just be "because the source says so".
it's an adjustment done because you may believe that the full growth function is unreasonable to extend out so far. from that point of view, if you don't do it like that the reserves would be overstated. if you take the opposite point of view that your model is the truth, then yea what you said would be true

did I understand you correctly or was I off here?
__________________
Reply With Quote
  #1006  
Old 06-16-2020, 11:13 PM
SkolChicago SkolChicago is offline
Member
CAS
 
Join Date: Apr 2015
Posts: 1,278
Default

Quote:
Originally Posted by AbedNadir View Post
it's an adjustment done because you may believe that the full growth function is unreasonable to extend out so far. from that point of view, if you don't do it like that the reserves would be overstated. if you take the opposite point of view that your model is the truth, then yea what you said would be true

did I understand you correctly or was I off here?
So here’s a contrasting example between the two methods: let’s say that the payment pattern pays 1/3 of ultimate in each the first three years. Then let’s say that the problem asks you to truncate after two years (an unreasonable example, I know).

Under the LDF method, you would adjust the % earned below the truncation point to assume 100% paid at that point. So in this (LDF method) example, we would adjust and now have 50% paid in year 1 and 50% in year 2.

However, based on the problems I’ve done, the process of adjusting the patterns for truncation is different. I would’ve thought that the reserve at 12 months would be OLEP*ELR*50%, where the 50% is the unreported claims adjusted for truncation. Instead, the problems show the reserve as OLEP*ELR*(2/3 - 1/3). Here the 2/3 is the growth function at truncation, the 1/3 is the growth function at current eval.

Do you see how I’m perceiving these as inconsistent?
__________________
ACAS 7 8 9
Reply With Quote
  #1007  
Old 06-17-2020, 03:01 PM
Rev_'s Avatar
Rev_ Rev_ is offline
Member
CAS
 
Join Date: Jun 2017
College: Seattle Pacific
Posts: 69
Default

Has anyone used the Bootstrap model excel file? I essentially have the problem where the paid+unpaid isn't even close to the ultimate if the paid development technique results in a higher ultimate than the incurred technique (using the incurred CL).
Reply With Quote
  #1008  
Old 06-17-2020, 07:39 PM
Candidate X's Avatar
Candidate X Candidate X is offline
Member
CAS
 
Join Date: Oct 2015
Posts: 328
Default

Quote:
Originally Posted by SkolChicago View Post
So here’s a contrasting example between the two methods: let’s say that the payment pattern pays 1/3 of ultimate in each the first three years. Then let’s say that the problem asks you to truncate after two years (an unreasonable example, I know).

Under the LDF method, you would adjust the % earned below the truncation point to assume 100% paid at that point. So in this (LDF method) example, we would adjust and now have 50% paid in year 1 and 50% in year 2.

However, based on the problems I’ve done, the process of adjusting the patterns for truncation is different. I would’ve thought that the reserve at 12 months would be OLEP*ELR*50%, where the 50% is the unreported claims adjusted for truncation. Instead, the problems show the reserve as OLEP*ELR*(2/3 - 1/3). Here the 2/3 is the growth function at truncation, the 1/3 is the growth function at current eval.

Do you see how I’m perceiving these as inconsistent?
I think I see what you mean and I do think that Clark's method of truncation is inconsistent between the LDF and CC methods. For the CC to be consistent mathematically, I believe you'd just need to add a step where you divide the growth function by the %reported at the truncation point (equivalent to dividing each LDF by the LDF at truncation). In your case you'd divide the growth function outputs at each age by 2/3, so you'd get OLEP*ELR*(2/3 - 1/3) / (2/3) = OLEP*ELR*(1 - 1/2)... exactly as you thought it should be.

BUT, unfortunately, Clark didn't do it that way, so it doesn't matter for the purposes if this exam
Reply With Quote
  #1009  
Old 06-17-2020, 08:31 PM
Candidate X's Avatar
Candidate X Candidate X is offline
Member
CAS
 
Join Date: Oct 2015
Posts: 328
Default

Quote:
Originally Posted by Rev_ View Post
Has anyone used the Bootstrap model excel file? I essentially have the problem where the paid+unpaid isn't even close to the ultimate if the paid development technique results in a higher ultimate than the incurred technique (using the incurred CL).
What data are you plugging through? It's not clear what you mean since paid+E(unpaid) = E(ultimate)

I believe the bootstrap "Mean Unpaid" indications should be close to the chain ladder estimated unpaid if enough iterations are run.
Reply With Quote
  #1010  
Old 06-18-2020, 10:08 PM
trueblade's Avatar
trueblade trueblade is offline
Member
CAS
 
Join Date: Jul 2012
Location: Castlevania
Studying for Freedom
Posts: 1,296
Default

Quote:
Originally Posted by SkolChicago View Post
For Clark, why doesn't Cape Cod scale up the unpaid % by the (inverse of the) growth function at the truncation point? It seems that if you just take the difference between the growth functions at truncation and at the average age, your reserves would be understated.

I suppose the answer may just be "because the source says so".
Didn't go through all the replies, but here's how I interpret it (the idea is from TIA)

CC method:

R = Prem * ELR * [G(x trunc) - G(x)] =>
R = Prem * ELR * G(x trunc)* [1 - G(x)/G(x trunc)] (consistent with the LDF method)

Think of the bolded part as the ultimate loss adjusted by the growth function at T
__________________


ACAS 7 8 9

Last edited by trueblade; 06-18-2020 at 10:24 PM..
Reply With Quote
Reply

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 05:45 PM.


Powered by vBulletin®
Copyright ©2000 - 2020, 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 0.20285 seconds with 12 queries