2011 Exam Posted

[Vorian Atreides]

http://www.casact.org/admissions/studytools/exam5/11-5.pdf

[JasonScandopolous]

regarding question #5: can someone examine this and confirm that the answer, whatever it may be, based on 6-mo policies is the same answer if we're talking about 1-yr policies? I remember that I determined this on the exam, and I agree with it now, just checking that this was right.

[carrytheCrøss]

Took the reserving exam back in 2006, so it's my first time looking at this 2011 exam. Have no clue why the 2011 exam either implictly requires or explictly asks the candidates to calculate link ratios so many times. I have no idea what testing that concept over and over proves.

[Inconceivable]

Took the reserving exam back in 2006, so it's my first time looking at this 2011 exam. Have no clue why the 2011 exam either implictly requires or explictly asks the candidates to calculate link ratios so many times. I have no idea what testing that concept over and over proves.

It proves that you can spend too much time on it, probably make silly mistakes and therefore lose points twice over. And give you hand fatigue. So it proves they want to screw you.

[Vorian Atreides]

regarding question #5: can someone examine this and confirm that the answer, whatever it may be, based on 6-mo policies is the same answer if we're talking about 1-yr policies? I remember that I determined this on the exam, and I agree with it now, just checking that this was right.

trend to forecast period is 7/1/10 to 10/1/12 ==> 1.75 years.

09 trend = 395/375 * 1.05^1.75 - 1.0 = 13.3%
10 trend = 395/390 * 1.05^1.75 - 1.0 = 10.3%

If you were using annual policies, trend to forecast period would be the same since the average written date in the forecast period is the same in both cases.

[Vorian Atreides]

Took the reserving exam back in 2006, so it's my first time looking at this 2011 exam. Have no clue why the 2011 exam either implictly requires or explictly asks the candidates to calculate link ratios so many times. I have no idea what testing that concept over and over proves.
CAS was wanting to provide an easy Exam for the transition period.

[booyah81]

regarding question #5: can someone examine this and confirm that the answer, whatever it may be, based on 6-mo policies is the same answer if we're talking about 1-yr policies? I remember that I determined this on the exam, and I agree with it now, just checking that this was right.

I believe that's right. The only trick was to notice that the projected period isn't just 12 months as it usually is but rather 18 months, so the AWD s/b 10/1/2012.

[booyah81]

trend to forecast period is 7/1/10 to 1/1/13 ==> 2.5 years.

09 trend = 395/375 * 1.05^2.5 - 1.0 = 19.0%
10 trend = 395/390 * 1.05^2.5 - 1.0 = 14.4%

If you were using annual policies, trend to forecast period would be 7/1/10 to 4/1/13 ==> 2.75 years ==> different answer than above.

Maybe I'm crazy but I thought the "trend to" date didn't depend on the length of the policy term, as long as you're trending to the AWD. AED would require the policy term length, but not the AWD.

Thus, if you're trending to the AWD it's the midpoint of the projected period which is 10/1/12. If it's the AED, it's half the length of the policy term after that, or 10/1/12 + policy term in months/2 = 10/1/12 + 3 months = 1/1/13.

[Vorian Atreides]

Maybe I'm crazy but I thought the "trend to" date didn't depend on the length of the policy term, as long as you're trending to the AWD. AED would require the policy term length, but not the AWD.

Thus, if you're trending to the AWD it's the midpoint of the projected period which is 10/1/12. If it's the AED, it's half the length of the policy term after that, or 10/1/12 + policy term in months/2 = 10/1/12 + 3 months = 1/1/13.
Yeah, I was thinking AED to AED in my first post . . . I've since fixed it.

[booyah81]

Yeah, I was thinking AED to AED in my first post . . . I've since fixed it.

dude, you nearly gave me a heart attack... :swear:

[booyah81]

trend to forecast period is 7/1/10 to 10/1/12 ==> 1.75 years.

09 trend = 395/375 * 1.05^1.75 - 1.0 = 13.3%
10 trend = 395/390 * 1.05^1.75 - 1.0 = 10.3%

If you were using annual policies, trend to forecast period would be the same since the average written date in the forecast period is the same in both cases.

7/1/10 - 10/1/12 = 2.25 years

[Vorian Atreides]

Glad I didn't sit for 5A.

[Vorian Atreides]

dude, you nearly gave me a heart attack... :swear:
Isn't that what AEDs are for? ;-)

[JasonScandopolous]

7/1/10 - 10/1/12 = 2.25 years

Yes yes, this is what I did, thank you both for the responses.

[ThePerfectGame]

Took the reserving exam back in 2006, so it's my first time looking at this 2011 exam. Have no clue why the 2011 exam either implictly requires or explictly asks the candidates to calculate link ratios so many times. I have no idea what testing that concept over and over proves.

That's been the major complaint about this exam (5B). Glad to see it seems as pointless to someone outside the test-taking population as it does to everyone in it.

Although, perhaps the exam committee deemed the ability to feverishly divide 15 sets of two 5 digit numbers and then calculate a straight average of the results to be one of the most important aspects of setting loss reserves. Sort of a secret Learning Objective (15-25%). In which case I am golden, because I divided the hell out of those 5-digit numbers. And then I recorded the crap out of what my calculator said the results were.

[oblivious]

Glad I didn't sit for 5A.

:swear::swear::swear::swear::swear::swear::swear: :swear: :swear::swear::swear::swear::swear::swear::swear: :swear::swear::swear::swear::swear::swear:

[booyah81]

Here are my solutions for #3 and #8 - just want to see what points I'm losing:

(3)
(a) .91666666
(b) 1
(c) 1.1666666
(d) 1
(e) 3 (I wrote 1.5 b/c i thought it was in-force exposures)

(8)
Combined Ratio = LR + UW Expense / EP + LAE / WP
100% = LR + (33.6 + 9.8 + 36.96)/308 + (LAE/loss)(LR)
100% = LR(1+.082) + .26090909
LR = .6831

OER = UW Expense / EP + LAE / EP
= .26090909 + (LAE/loss)(LR)
= .26090909 + .082(.6831)
= .4169

[MightySchoop]

Here are my solutions for #3 and #8 - just want to see what points I'm losing:

(3)
(a) .91666666
(b) 1
(c) 1.1666666
(d) 1
(e) 3 (I wrote 1.5 b/c i thought it was in-force exposures)


3a is 0.75. The cancellation of policy D "unwrites" 3 months of premium.
The others look fine to me.

[chopsuey]

I think OER calc is this:

U/w expense = .275

that is:
[33.6 + 9.8]/280 + 36.96/308 = .275

1 - u/w expense = LR(1.082) = .725

=> LR = .67

and, OER = 1- LR = .33

[booyah81]

3a is 0.75. The cancellation of policy D "unwrites" 3 months of premium.
The others look fine to me.

You're right. I thought it was -1/12 because I thought 1 month into 2010. Not smart.

[booyah81]

I think OER calc is this:

U/w expense = .275

that is:
[33.6 + 9.8]/280 + 36.96/308 = .275

1 - u/w expense = LR(1.082) = .725

=> LR = .67

and, OER = 1- LR = .33

Yeah, I forgot that I split the gen/EP and all other/WP too.

OER + LR = 1? Gotcha. Yeah, didn't do that. Hopefully won't lose too many points here...

[herr blau]

Anyone remember what they got for the indicated rate change in question #9b (big 6.75 pt question)?

I'm pretty sure I screwed it up b/c of the 6-month policy term... :(

[actuarialista]

I think OER calc is this:

U/w expense = .275

that is:
[33.6 + 9.8]/280 + 36.96/308 = .275

1 - u/w expense = LR(1.082) = .725

=> LR = .67

and, OER = 1- LR = .33

That's what I got.

[actuarialista]

Anyone remember what they got for the indicated rate change in question #9b (big 6.75 pt question)?

I'm pretty sure I screwed it up b/c of the 6-month policy term... :(

On exam day, I got 12.4%, but then later on I realized I made a mistake---I added the 0.032 ULAE provision to ratio of loss & ALAE to premium rather than multiplying the LR by 1.032. So I think the answer should really be 11.2%. However, I think there's another thread where two people were saying they got something less than that.

[herr blau]

On exam day, I got 12.4%, but then later on I realized I made a mistake---I added the 0.032 ULAE provision to ratio of loss & ALAE to premium rather than multiplying the LR by 1.032. So I think the answer should really be 11.2%. However, I think there's another thread where two people were saying they got something less than that.

Hmmmm...can't find it. I got +9.25%, but I'm not sure about my answer, given the 6-month policies...

[booyah81]

Not sure that this sounds familiar to me but I got +7.32%.

Trend for premium and losses should be 3.0 and 2.0 years for CAY 2009 and 2010, respectively; premium trend is based on AWD to AWD whereas losses are based on AAD to AAD.

My LDFs were 1.15 at age 12 and 1.111 at age 24, and 1.01 at age 36. I actually calculated all the age-to-age factors just in case there was something screwy but there wasn't. They were all stable at each age so I took straight averages. Multiply to get ultimate factors.

LRs end up being .6970 for CAY09 and .7441 for CAY10. I selected the dollar weighted = sum trended ult L&LAE / sum proj'd OLEP = .7221 (I think that's what I got on the exam).

Indicate rate change = (.7221 + .056) / (1 - .24 - .035) - 1.0 = +7.32%.

Attached is my scratch work...

[chopsuey]

OLF for 2010 is a little off. ARRL_10 = (.97)*1.06/[.25*.97 + .75*1.0282]

all of your other work is correct.

I think final answer is somewhere around 9.68%? Although I round a lot.

[booyah81]

OLF for 2010 is a little off. ARRL_10 = (.97)*1.06/[.25*.97 + .75*1.0282]

all of your other work is correct.

I think final answer is somewhere around 9.68%? Although I round a lot.

Fudge. Yeah, I misread it the same way I misread it on the exam. 1/1/10 rate change, not 1/1/11. Hopefully I didn't but I think I did.

[actuarialista]

OLF for 2010 is a little off. ARRL_10 = (.97)*1.06/[.25*.97 + .75*1.0282]

all of your other work is correct.

I think final answer is somewhere around 9.68%? Although I round a lot.

I found my other mistake... I subtracted the years wrong and so my premium trend factors were 1.015^2 and 1.015 instead of 1.015^3 and 1.015^2. If I fix that mistake and also apply the ULAE provision correctly, I get 9.68%, agreeing with you, chopsuey!

[booyah81]

I found my other mistake... I subtracted the years wrong and so my premium trend factors were 1.015^2 and 1.015 instead of 1.015^3 and 1.015^2. If I fix that mistake and also apply the ULAE provision correctly, I get 9.68%, agreeing with you, chopsuey!

Did you make the same mistake for the loss trend?

[actuarialista]

Did you make the same mistake for the loss trend?

No, I don't think I did! My losses seem to have been right and my premium wrong. I know that's strange--chalk it up to the fact that I was trying to go really fast (and still didn't get through the whole exam).

[booyah81]

No, I don't think I did! My losses seem to have been right and my premium wrong. I know that's strange--chalk it up to the fact that I was trying to go really fast (and still didn't get through the whole exam).

I hate exams. :exams:

[JasonScandopolous]

Hmmmm...can't find it. I got +9.25%, but I'm not sure about my answer, given the 6-month policies...

I got something very similar to this.

EDIT: 9.68 actually seems very familiar, cool.

[Michael D Wang]

3a is 0.75. The cancellation of policy D "unwrites" 3 months of premium.
The others look fine to me.

Yeah, I got 0.75 for a) as well but put 1/2+1/2+1/12 = 1.0833 for part c)

:roll2:

[Test_Taker]

For #5...Since it's 2 Step Trending, the Step 2 Trend period should be the mid point of the latest data point, right? So with 6 month policies, you'd trend from 10/1/10 to 10/1/12?

[Vorian Atreides]

For #5...Since it's 2 Step Trending, the Step 2 Trend period should be the mid point of the latest data point, right? So with 6 month policies, you'd trend from 10/1/10 to 10/1/12?
It's not based on policy length, but the amount of data used to calculate that latest data point.

Since avg writ prem = (total written prem) / (total policy count) is based on a 12 month figure (1/1/10 to 12/31/10), you need the mid point of this set of dates: 7/1/10.

[Test_Taker]

Makes sense...thanks

[Vorian Atreides]

Makes sense...thanks
The amount of time the data is drawn over can be tricky as well.

Examples in W&M (IIRC) show both a set of data that is quarterly (so, avg date of last data point is over a 3 month time frame) and a data set that is also quarterly but on a "rolling 12 basis"--that is, the data for that final data point is over a 12 month period.

[shab]

" This 79.75 point examination consists of 37 problem and essay questions"

which questions were Essay(S) beside question 36 that I thought was essay...
doesn't the wording sound misleading..."37 problem and essay questions" sounds to me that there were more the 37 questions!

[MightySchoop]

" This 79.75 point examination consists of 37 problem and essay questions"

which questions were Essay(S) beside question 36 that I thought was essay...
doesn't the wording sound misleading..."37 problem and essay questions" sounds to me that there were more the 37 questions!

It's leftover wording from the days when it said

"This 100 point exam consists of 17 multiple-choice questions and 24 problem and essay questions."