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


Not looking for a job? Tell us about your ideal job,
and we'll only contact you when it opens up.
https://www.dwsimpson.com/register


Reply
 
Thread Tools Display Modes
  #2101  
Old 07-14-2017, 12:59 PM
FCAS_to_be FCAS_to_be is offline
Member
CAS
 
Join Date: Oct 2015
Posts: 321
Default

Quote:
Originally Posted by ExamsRock View Post
They took off points for 2 for not mentioning that you can't use a negative parameter? The question says use the method. Didn't ask whether it's appropriate.
Don't recall whether I read it in the source or a manual, but I think it was the source, that when doing Least Squares if the a parameter is calculated as less than 0 you use the link ratio method. If the b parameter is calculated as less than 0 you use the budgeted loss method.

Assuming I read it in the source, it's an example of where using a manual will cost you marks since the manual will teach you least squares but doesn't cover that least squares actually includes these exceptions.
Reply With Quote
  #2102  
Old 07-14-2017, 01:04 PM
FCAS_to_be FCAS_to_be is offline
Member
CAS
 
Join Date: Oct 2015
Posts: 321
Default

Quote:
Originally Posted by ExamsRock View Post
Also in 24 you lost points if in part c you repackaged your answer from part a? It's the same question! Did they specifically say it needs to be different?
I agree with this so hard. I put this in the exam survey - that this question had two subparts that asked exactly the same thing. Part c) of the question should have just been thrown out.

No where in the instructions does it say that I can't reuse my answers from other questions (or earlier in the same question). If you're going to ask the same question twice and I'm confident I got it the first time, I'm going to double down and answer the same correct answer the second time. The only reason you'd put anything different is because you're not confident and want to maximize your chance of not scoring a 0 on the question.

a) and c) should have been combined to ask for 4 reasons, and to label them as advantages to c or disadvantages to a.
Reply With Quote
  #2103  
Old 07-14-2017, 01:28 PM
Furious_Falcon Furious_Falcon is offline
Member
CAS
 
Join Date: Apr 2013
College: Bowling Green State University Alumni
Posts: 96
Default

Quote:
Originally Posted by FCAS_to_be View Post
Don't recall whether I read it in the source or a manual, but I think it was the source, that when doing Least Squares if the a parameter is calculated as less than 0 you use the link ratio method. If the b parameter is calculated as less than 0 you use the budgeted loss method.

Assuming I read it in the source, it's an example of where using a manual will cost you marks since the manual will teach you least squares but doesn't cover that least squares actually includes these exceptions.
I used two different manuals and they both covered this. I think the issue was at that point you're changing your methodology and the question explicitly told you to use LS. Also, you get the negative a when using LS on 2013, and you're only doing that so you can have another data point to use LS for 2014. Deviating from the methodology for one year to aid you in using it on another seems illogical; a good actuary would just change the methodology.

I noticed the negative and used LS anyway, and explicitly pointed out to the grader that I was doing it because I was told even though CL might be better. The second model answer does this too, so I'm assuming I got full credit. It's annoying that they throw curve balls like this in there, but I've learned not to waste any time thinking about what's "right", just state your assumption and go.

Last edited by Furious_Falcon; 07-14-2017 at 02:00 PM..
Reply With Quote
  #2104  
Old 07-14-2017, 01:41 PM
cmoibenlepro's Avatar
cmoibenlepro cmoibenlepro is offline
Member
CAS SOA
 
Join Date: May 2007
Location: Montreal
Studying for 8
College: UQAM
Posts: 606
Default

Quote:
Originally Posted by ExamsRock View Post
They took off points for 2 for not mentioning that you can't use a negative parameter? The question says use the method. Didn't ask whether it's appropriate.
They did the same in previous years so no surprise here.

I explicitely said that I used LS even if the parameter was negative and CL was better so I got my points. Easy question.
__________________
SOA:P FM M C FAP1 FA ASA
CAS: OC1 OC2 5 6C ACAS 7 8 9 FCAS

Last edited by cmoibenlepro; 07-14-2017 at 01:44 PM..
Reply With Quote
  #2105  
Old 07-14-2017, 04:11 PM
Polak's Avatar
Polak Polak is offline
Member
CAS
 
Join Date: May 2013
Posts: 324
Default

Yeah I don't really think the negative a or b values in LS is a curve ball. It's been tested multiple times and it's a clear recommendation to use the alternate methods when one of them is negative. Everyone should have known the issues related to using negative a or b going in to that test, and when it happened it seems obvious that they were looking for you to comment on it somehow.
__________________
Exams: P FM 3L 3F 4C 5 6 7 8 9
VEE: Econ Finance Statistics
Online Courses: CAS1 CAS2
COP
Reply With Quote
  #2106  
Old 07-14-2017, 05:59 PM
CuriousGeorge CuriousGeorge is offline
Member
CAS SOA
 
Join Date: Dec 2005
Posts: 929
Default

I don't disagree too much, but when they asked this exact same question in 2011, the solution did use a negative b in the the sample solution. And then they asked a follow up question for you to assess the appropriateness of LS in that situation.

I know maybe I should have commented, but when I already can't finish in the time allotted, I've got to cut somewhere.
Reply With Quote
  #2107  
Old 07-14-2017, 06:08 PM
Polak's Avatar
Polak Polak is offline
Member
CAS
 
Join Date: May 2013
Posts: 324
Default

Quote:
Originally Posted by CuriousGeorge View Post
I don't disagree too much, but when they asked this exact same question in 2011, the solution did use a negative b in the the sample solution. And then they asked a follow up question for you to assess the appropriateness of LS in that situation.

I know maybe I should have commented, but when I already can't finish in the time allotted, I've got to cut somewhere.
Okay, this is a fair response, in the past they did have a part specifically referring to the negative and the sample answer used the negative in the solution.
__________________
Exams: P FM 3L 3F 4C 5 6 7 8 9
VEE: Econ Finance Statistics
Online Courses: CAS1 CAS2
COP
Reply With Quote
  #2108  
Old 07-14-2017, 10:08 PM
Tacoactuary's Avatar
Tacoactuary Tacoactuary is offline
Member
CAS Non-Actuary
 
Join Date: Nov 2014
Location: Des Moines, IA
College: Vanderbilt, UIUC
Favorite beer: Yazoo Sue
Posts: 512
Default

Question 10, I thought they might do this: providing an inappropriate number of residuals by calendar year. I think you could argue against this. What if residuals were the same? In a given CY and thus overlapped?
__________________
ACAS 7 8 9
Reply With Quote
  #2109  
Old 07-17-2017, 12:31 PM
FCAS_to_be FCAS_to_be is offline
Member
CAS
 
Join Date: Oct 2015
Posts: 321
Default

Quote:
Originally Posted by Tacoactuary View Post
Question 10, I thought they might do this: providing an inappropriate number of residuals by calendar year. I think you could argue against this. What if residuals were the same? In a given CY and thus overlapped?
I actually commented in my exam something along the lines of:
I've drawn this quickly to show the general trend and did not count the appropriate number of residuals by CY.

I wonder if I was penalized for that or not. I stated my "assumption" pretty clearly, however likewise I didn't bother saying the number that each year should have had, etc...

As for the appeal argument: it'd be possible on that grounds for someone lucky enough to have just missed 1 or 2 somewhere. If you drew the same number for each year, etc... I doubt they'd accept that argument (i.e. they won't believe that there was 1 at the same level in one year, 2 in the next year, 3 in the next, etc...).
Reply With Quote
  #2110  
Old 07-18-2017, 09:12 AM
|B|rad |B|rad is offline
Member
CAS
 
Join Date: Jul 2009
Posts: 912
Default

Quote:
Originally Posted by Tacoactuary View Post
Question 10, I thought they might do this: providing an inappropriate number of residuals by calendar year. I think you could argue against this. What if residuals were the same? In a given CY and thus overlapped?
I think that's a little weak, but counting the # of dots on a free-handed graph with such a vague question is also pretty weak.

I can see arguments against their count position on both graphs.

In part (a), if you have a triangle that only includes AYs 2001-2005 (i.e., your company stopped writing these policies), but you have calendar data through 2010. Then if your residual graph x-axis goes 2005, 2006, 2007, 2008, 2009, 2010 -- each year would have an equal number of residual points (5). This would be even further justified if your triangle was missing transactional data prior to 2005 so that your "triangle" is actually a parallelogram.

For part (b), I think a similar argument can be made. At no point does the question lay out exactly which years/ages you need to include or that we are talking about a "standard" triangle.
Reply With Quote
Reply

Tags
gbmisag, thankswilly, we shall pass, willy id do anything

Thread Tools
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 06:40 PM.


Powered by vBulletin®
Copyright ©2000 - 2017, 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 0.34144 seconds with 10 queries