Actuarial Outpost
 
Go Back   Actuarial Outpost > Exams - Please Limit Discussion to Exam-Related Topics > SoA > Modules 6-8
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 our jobs meet your skills and objectives.


Reply
 
Thread Tools Search this Thread Display Modes
  #1  
Old 01-23-2017, 10:11 PM
actuarialstudent13 actuarialstudent13 is offline
Member
SOA
 
Join Date: Oct 2015
Posts: 46
Default Selecting an Asset Mix and Tax Rate for CDEF

I failed the CDEF FA and my feedback was that Tasks 2 and 3 were the worst.

What was your strategy for selecting your asset mix? In my first attempt, I modeled all the possible mixes in 25% increments and went with the asset mix that minimized the CTE95. I ended up with an allocation of 25% T-bills, 50% Bonds and 25% equities.

I felt pretty good about this at the time (and still do), but I recognize that I'm open to SOA criticism on two fronts: First, 25% increments might be too large. The SOA might claim that I might be missing something. I doubt that narrowing the increments results in a great deal of new information, but the SOA might assert that I haven't tried hard enough here.

Second, allowing equities in the portfolio. Equities are high volatility, and they can't be used to duration match the liability, so the SOA might not want any equities in the portfolio. That having been said, equities reduce the expected payments that taxpayers are going to need to make, and furthermore, as I already mentioned, the model says that this portfolio with equities minimizes the CTE95, so clearly the model is suggesting that this extra risk is worth it. Thoughts?

To set the contribution rate, I took the mean (as a proxy for the median), as I neither wanted to be too conservative (in which case current tax payers would pay more than future tax payers), nor too aggressive (in which case the next generation of tax payers would face a greater burden). I think this was what the SOA really disliked about my Task 2 response. I think that the fact that it would require new legislative action to change the tax rate was hinting to pick a tax rate that was on the conservative side. Any suggestions for how to do this?

I was thinking of using the CTE95 as my risk metric and using the CTE75 as my tax rate. How does that sound? Thoughts? Alternatively, I could take the mean tax rate and add a standard deviation or two.

Thanks for the help. It's appreciated.
Reply With Quote
  #2  
Old 01-24-2017, 10:38 AM
Amyashi Amyashi is offline
Member
SOA
 
Join Date: Jul 2014
Location: Toronto
College: University of Waterloo (Alumnus)
Posts: 57
Default

I think it's less about what SOA wants, and more about what you think is the best fit for the company's objectives, and how you decide to balance risk vs. return.

For the asset mix, I don't think it's about getting the "right" answer. If you're putting 50% in bonds and 25% in T-bills, knowing that they have lower returns, can you support the extra tax burden on the population? If you're putting 25% in equities can you justify the level of risk? How do the different asset classes interact with each other, and how do you expect your portfolio to behave in markets that are performing well vs. low-interest environments?

I think you're on the right track with your thought process for setting the contribution rate. Whatever you choose you should be able to support it. Rather than picking an arbitrary point of CTEXX you should be looking at what that means for your population. Like for example, CTE75 is the average of the worst 25% scenarios through your testing. If you set the price at CTE75 you're effectively setting the price as if the worst 25% of scenarios from your testing will happen 100% of the time. Is that enough of a contingency? Is that too much?
Reply With Quote
  #3  
Old 01-24-2017, 10:13 PM
actuarialstudent13 actuarialstudent13 is offline
Member
SOA
 
Join Date: Oct 2015
Posts: 46
Default

Thanks for chiming in Amyashi.

Your point that we need to justify that the asset mix we're going with is the best one, is good. And your question, "Is that enough of a contingency?" is really good. The problem is that, I do not feel like the FAP course has prepared me to asset the amount of risk capital the CDEF should take on.

Do have suggestions for how to determine the level of contingency?
Reply With Quote
  #4  
Old 01-25-2017, 10:04 AM
Amyashi Amyashi is offline
Member
SOA
 
Join Date: Jul 2014
Location: Toronto
College: University of Waterloo (Alumnus)
Posts: 57
Default

I agree! I don't think the FAP course was designed to teach us exactly how to price these kind of risk capitals, and if it was, it wasn't done very well :P

I think you'll have to use some creativity and find a middle ground that's cohesive with your answers for the other tasks. I'd like to think that they're flexible enough to accept a range of answers, but I think definite no's are either setting it at the mean/median with no margin for adverse deviation (too risky), or setting it exactly at a high CTE (too conservative).

For what it's worth, the risk measure I used to evaluate the recommended asset portfolio and what I used to set the contribution rate were different. I used a high CTE to evaluate the riskiness of the asset portfolio (and therefore arriving at a recommended asset mix), and then using that portfolio, mean + some margin derived from a risk measure as my recommended contribution rate.
Reply With Quote
  #5  
Old 01-25-2017, 07:05 PM
YHill YHill is offline
SOA
 
Join Date: Jan 2017
College: Baruch
Posts: 11
Default

Quote:
Originally Posted by Amyashi View Post
I agree! I don't think the FAP course was designed to teach us exactly how to price these kind of risk capitals, and if it was, it wasn't done very well :P

I think you'll have to use some creativity and find a middle ground that's cohesive with your answers for the other tasks. I'd like to think that they're flexible enough to accept a range of answers, but I think definite no's are either setting it at the mean/median with no margin for adverse deviation (too risky), or setting it exactly at a high CTE (too conservative).

For what it's worth, the risk measure I used to evaluate the recommended asset portfolio and what I used to set the contribution rate were different. I used a high CTE to evaluate the riskiness of the asset portfolio (and therefore arriving at a recommended asset mix), and then using that portfolio, mean + some margin derived from a risk measure as my recommended contribution rate.
This is super helpful feedback. When you say you used CTE to determine the asset mix, how did you link the CtE value to the other metrics (mean, min,max)? For example I was looking for a blend that has CTE90 about 10% higher than the mean in order to have a portfolio with lower variation in outcomes. But the 10% is very discretionary and hard to justify why 10% instead of 15. Any ideas?
Reply With Quote
  #6  
Old 01-25-2017, 08:48 PM
Amyashi Amyashi is offline
Member
SOA
 
Join Date: Jul 2014
Location: Toronto
College: University of Waterloo (Alumnus)
Posts: 57
Default

I had to think a bit about your method and why it might not make sense. I think, having a CTE90 about 10% higher than the mean tells me two things: that the variability is low, yes, but also your mean must be VERY high. Lower risk comes with lower rewards, and if I had to guess your asset blend I'd guess that it should be close to 100% in some mix of treasuries/bonds. If you are to propose such a risk measure as your guideline, can you justify the amount of extra tax burden on the population? I don't think it necessarily makes sense to link the mean and CTE, but that's just my opinion.

I approached it slightly differently. I started with the fact that bonds/equity are negatively correlated and therefore act as a hedge towards the other, then worked from there to minimize the CTE95 to a reasonable level and align the target ranges with my key risks and CDEF objectives.
Reply With Quote
  #7  
Old 01-25-2017, 09:20 PM
actuarialstudent13 actuarialstudent13 is offline
Member
SOA
 
Join Date: Oct 2015
Posts: 46
Default

Thanks for continuing to offer helpful suggestions, Amyashi.

Did you include your risk metric in your sensitivity testing? I suppose that you must have, given the way you selected your contribution rate.

Peter
Reply With Quote
  #8  
Old 01-25-2017, 10:05 PM
Amyashi Amyashi is offline
Member
SOA
 
Join Date: Jul 2014
Location: Toronto
College: University of Waterloo (Alumnus)
Posts: 57
Default

No problem, CDEF is quite new and I'm glad I can help a bit.

I remember looking at the sensitivity for all the risk measures listed in the worksheet. I think the impact was pretty consistent across the board.
Reply With Quote
  #9  
Old 01-25-2017, 11:17 PM
YHill YHill is offline
SOA
 
Join Date: Jan 2017
College: Baruch
Posts: 11
Default

Quote:
Originally Posted by Amyashi View Post
I had to think a bit about your method and why it might not make sense. I think, having a CTE90 about 10% higher than the mean tells me two things: that the variability is low, yes, but also your mean must be VERY high. Lower risk comes with lower rewards, and if I had to guess your asset blend I'd guess that it should be close to 100% in some mix of treasuries/bonds. If you are to propose such a risk measure as your guideline, can you justify the amount of extra tax burden on the population? I don't think it necessarily makes sense to link the mean and CTE, but that's just my opinion.

I approached it slightly differently. I started with the fact that bonds/equity are negatively correlated and therefore act as a hedge towards the other, then worked from there to minimize the CTE95 to a reasonable level and align the target ranges with my key risks and CDEF objectives.
The mix I ended up with was 10% equity, 15% treasury and the rest bonds. Did you end up with higher than 10% in equity?
Reply With Quote
  #10  
Old 01-25-2017, 11:19 PM
YHill YHill is offline
SOA
 
Join Date: Jan 2017
College: Baruch
Posts: 11
Default

Quote:
Originally Posted by actuarialstudent13 View Post
Thanks for continuing to offer helpful suggestions, Amyashi.

Did you include your risk metric in your sensitivity testing? I suppose that you must have, given the way you selected your contribution rate.

Peter
I am not sure I follow, are talking about the sensitivity test in task 4 or just the variation when choosing the portfolio mix?
Reply With Quote
Reply

Tags
cdef

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 03:48 AM.


Powered by vBulletin®
Copyright ©2000 - 2018, 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 0.47698 seconds with 9 queries