Actuarial Outpost > SoA FA Can-Do Task 7
 Register Blogs Wiki FAQ Calendar Search Today's Posts Mark Forums Read
 FlashChat Actuarial Discussion Preliminary Exams CAS/SOA Exams Cyberchat Around the World Suggestions

 DW Simpson Global Actuarial & Analytics Recruitment Download our 2017 Actuarial Salary Survey now with state-by-state salary information!

#1
07-20-2012, 06:03 PM
 JJG86 SOA Join Date: Jul 2012 Posts: 11

Does anyone who has recently passed want to PM me to discuss this topic? I received feedback after my first two attempts. This was the question I completely redid; the rest I just updated. I was told on my third fail that it was one of the two needing the most improvement.

What I did was look at the following assumptions based on actual experience:

1. Exchange Rate. We were told this was equal to 1.11 instead of 1.04, so I recommended that 1.11 be used to reflect actual experience.

2. Inflation. The way I did this was I subtracted out labor costs from total costs, since labor costs were not impacted by inflation. Then I took the ratio of the prior quarter's to the current quarter's costs to estimate inflation, which was on average around 0.9% per quarter. I recommended this amount be used.

3. Decay rate. I backed into this number using "g" as it was defined in the instructions and found the decay rate using the following equation: 1-(g(current)/g(prior))^(1/#periods in prior quarter). This was around 4%, which I suggested be used.

4. Labor costs. I tried to see if there was a linear relationship. To do this, I took the difference between the current and prior quarters' costs, then divided by the change in "g" over the quarter. There was not a clear relationship by doing this. Variable amounts were all over the place from quarter to quarter. So then I decided to see if there was a strictly variable cost without a fixed cost. To do this, I took the current quarter's labor costs and divided them by the current quarter's "g", and found there was a relationship of around \$6k for two shifts and around \$2.5k for one shift, consistently by quarter. I suggested this formula be used in the model.

5. "Other" costs - Delivery, equipment, and energy. I can tell that these have been lower than expected, but since the breakout was not provided for the individual components, I can't determine which costs are not in line with expected. I was able to determine "other" costs were lower than expected by taking the sum of \$4.5million for two shifts or \$2.5million for one shift (energy costs), \$1.00 per oz of gold excavated (delivery cost), and \$100k (maintenance and repair), then multiplied them by expected inflation per quarter.

I then stated that I wasn't able to determine a good way to analyze investment yield or cost of capital using the data provided. If additional data is provided on those items, I would be able to conduct an experience study on them.

What am I doing wrong?
#2
07-20-2012, 06:19 PM
 JJG86 SOA Join Date: Jul 2012 Posts: 11

I didn't include any graphs, although I did have an appendix showing my numerical results. I.e. what I expected vs. actual (with "other" costs broken out, etc). Do you think this was part of the reason I failed?

I'm really at a loss here, so any advice at all as to what people who passed did would be appreciated.
#3
07-28-2012, 08:58 PM
 tormato SOA CCA Join Date: Jul 2012 Posts: 6

Similar questions as the last ones....I really am stuck on how to approach Task 7 with only 6 columns of numbers.
#4
09-09-2012, 05:44 PM
 Tireburner10 Member SOA Join Date: Jul 2010 Favorite beer: Left Hand Nitro Milk Stout Posts: 191

I didn't pass the FA on my first attempt and they listed Task 7 as one of mine that needed the most improvement, so I'm interested in feedback from people who have passed the FA, not from people who are awaiting results.

After accounting for the fact that my Task 7 response was shorter than all of the others (might showing the grader that I spent less time on it?), and after accounting for the fact that I did all of the usual expected calculations on the data given, I can only find one area for improvement: explaining my methodology a bit more. Instead of saying "I did an actual to expected analysis", should I explicitly state for each variable I compared actual experience to the modeled experience and used the difference to make a recommendation?
__________________
Work to live, don't live to work. Also, http://blog.jalopnik.com/

ASA FHE Mod HF Mod, PRF Mod, DP-GH GH-Advanced, DMAC, FAC
#5
09-09-2012, 11:51 PM
 lucas_m Member SOA Join Date: Nov 2006 Studying for FSA modules Favorite beer: Alexander Keiths Posts: 256

Quote:
 Originally Posted by JJG86 Does anyone who has recently passed want to PM me to discuss this topic? I received feedback after my first two attempts. This was the question I completely redid; the rest I just updated. I was told on my third fail that it was one of the two needing the most improvement. What I did was look at the following assumptions based on actual experience: 1. Exchange Rate. We were told this was equal to 1.11 instead of 1.04, so I recommended that 1.11 be used to reflect actual experience. 2. Inflation. The way I did this was I subtracted out labor costs from total costs, since labor costs were not impacted by inflation. Then I took the ratio of the prior quarter's to the current quarter's costs to estimate inflation, which was on average around 0.9% per quarter. I recommended this amount be used. 3. Decay rate. I backed into this number using "g" as it was defined in the instructions and found the decay rate using the following equation: 1-(g(current)/g(prior))^(1/#periods in prior quarter). This was around 4%, which I suggested be used. 4. Labor costs. I tried to see if there was a linear relationship. To do this, I took the difference between the current and prior quarters' costs, then divided by the change in "g" over the quarter. There was not a clear relationship by doing this. Variable amounts were all over the place from quarter to quarter. So then I decided to see if there was a strictly variable cost without a fixed cost. To do this, I took the current quarter's labor costs and divided them by the current quarter's "g", and found there was a relationship of around \$6k for two shifts and around \$2.5k for one shift, consistently by quarter. I suggested this formula be used in the model. 5. "Other" costs - Delivery, equipment, and energy. I can tell that these have been lower than expected, but since the breakout was not provided for the individual components, I can't determine which costs are not in line with expected. I was able to determine "other" costs were lower than expected by taking the sum of \$4.5million for two shifts or \$2.5million for one shift (energy costs), \$1.00 per oz of gold excavated (delivery cost), and \$100k (maintenance and repair), then multiplied them by expected inflation per quarter. I then stated that I wasn't able to determine a good way to analyze investment yield or cost of capital using the data provided. If additional data is provided on those items, I would be able to conduct an experience study on them. What am I doing wrong?
Here are a few tips.

What is the statistical tool or analysis usually used to confirm a linear relashionship between X and Y? You should use it to determine a and b.

The decay rate is stochastic. Is it possible to use experience data to calibrate the parameters of the normal distribution for future projections?

Try to reproduce the formulas from task 1-6 with the data of task 7. It will help you understand how the different parameters affect the final result.

Finally justify each of your recommendation, explain your method and summarize yours results.

Good luck!
__________________
P FM MLC MFE C VEE FAP FETE APMV FHE ERM DMAC FINREP FAC
#6
09-12-2012, 11:57 AM
 komorgan Member Non-Actuary Join Date: Mar 2012 Posts: 112

Quote:
 Originally Posted by lucas_m Here are a few tips. What is the statistical tool or analysis usually used to confirm a linear relashionship between X and Y? You should use it to determine a and b. The decay rate is stochastic. Is it possible to use experience data to calibrate the parameters of the normal distribution for future projections? Try to reproduce the formulas from task 1-6 with the data of task 7. It will help you understand how the different parameters affect the final result. Finally justify each of your recommendation, explain your method and summarize yours results. Good luck!
Are you saying that we should not update the parameters for the decay rate assumption?

It was my understanding that the original parameters were based on assays (which only involve small samples). On the other hand, the actual experience of the gold yield (over the past 5 years) involves extracting TONS of ore from the ground. Hence, I would have thought that the experience data would be more credible than what was used to estimate the original parameters.

Any thoughts?
#7
09-12-2012, 12:38 PM
 mathmajor Member SOA AAA Join Date: Dec 2010 Location: Nowhere in particular Studying for Japanese College: B.S. Applied Math Favorite beer: La Croix Grapefruit Posts: 9,790

Noteworthy: Decay rate is not on the scenarios and assumptions tab. I mentioned it would be worth looking at, but technically not part of the question.
__________________
FSA
If I can become one, anyone can, right?
#8
09-12-2012, 02:19 PM
 lucas_m Member SOA Join Date: Nov 2006 Studying for FSA modules Favorite beer: Alexander Keiths Posts: 256

Quote:
 Originally Posted by mathmajor Noteworthy: Decay rate is not on the scenarios and assumptions tab. I mentioned it would be worth looking at, but technically not part of the question.
Agreed but it costs nothing to add a few lines about it just to make sure you don't miss some important points. Better be safe than sorry.
__________________
P FM MLC MFE C VEE FAP FETE APMV FHE ERM DMAC FINREP FAC
#9
11-19-2012, 11:56 AM
 missactsci Member SOA Join Date: Aug 2007 Studying for LP Posts: 81

for those who have a hedging factor >0, did you take this into account when calculating "actuals" in task 7??
#10
11-19-2012, 01:07 PM
 Anitha Desai Member SOA AAA Join Date: Jul 2007 Studying for LRM Posts: 9,848

Quote:
 Originally Posted by missactsci for those who have a hedging factor >0, did you take this into account when calculating "actuals" in task 7??
The actuals are given. Do you mean calculating expected values?
What have you used the hedging factor for?
Is that variable or value being calculated here?
__________________
"YOU HUSSY!!! Do our AO vows mean nothing to you? To post and to quote, to RN and to :ctm: for better or worse? Till logout do us part so help you Tom? " - Inconceivable

 Tags can do, fa task 7, final assessment