Actuarial Outpost RP-2014 Age Basis
 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

 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

#1
01-25-2016, 05:04 PM
 MathGeek92 Member Join Date: Jul 2004 Posts: 724
RP-2014 Age Basis

Does any one know how they classified the age basis in RP-2014? Age-last, exact age?

My question really is if I were using RP-2014 to value the expected future benefits of someone today that is age 62 and 3 months... would I assume the age 62 mortality rate for the next 9 months (assuming its an "exact" age table) and then go to 63, or is it an age last table, so I should use the age 62 rate for the next 12 months of my projection.

I've read through the SOAs description of how the table was constructed, but I couldn't find a reference to the method

tia
#2
01-26-2016, 01:49 AM
 zeus1233 Member Join Date: Nov 2003 Posts: 1,393

Quote:
 Originally Posted by MathGeek92 Does any one know how they classified the age basis in RP-2014? Age-last, exact age? My question really is if I were using RP-2014 to value the expected future benefits of someone today that is age 62 and 3 months... would I assume the age 62 mortality rate for the next 9 months (assuming its an "exact" age table) and then go to 63, or is it an age last table, so I should use the age 62 rate for the next 12 months of my projection. I've read through the SOAs description of how the table was constructed, but I couldn't find a reference to the method tia
Pension vals generally apply decrements once annually, and they aren't averaging them across ages (i.e. you don't use 9 months of the age 62 rate and 3 months of the age 63 rate for someone aged 62.25)

Generally, it is truncated age, not rounded age - one would treat a 62.9 year old as 62.
#3
01-26-2016, 01:12 PM
 MathGeek92 Member Join Date: Jul 2004 Posts: 724

thats what i would have guessed, but was looking for verification in the SOA write up and couldn't find a description.

Thanks!
#4
01-26-2016, 02:55 PM
 Duke of Chalfont Member SOA Join Date: Apr 2011 Posts: 38

Though the RP-2014 report itself surprisingly seems to be silent on this point, I am certain that the underlying age basis is "age nearest birthday" (rounded age). That was definitely the basis for RP-2000. So, someone aged 62 years 3 months would be treated as being age 62, while someone aged 62 years 10 months would be counted as being age 63.
#5
01-26-2016, 04:17 PM
 Helen Sass Member SOA AAA Join Date: Oct 2015 Location: Nearby Studying for Colorectal Exam College: SUNY Alumni Favorite beer: #1 - Free; #2 - Cold Posts: 2,715

Quote:
 Originally Posted by MathGeek92 Does any one know how they classified the age basis in RP-2014? Age-last, exact age? My question really is if I were using RP-2014 to value the expected future benefits of someone today that is age 62 and 3 months... would I assume the age 62 mortality rate for the next 9 months (assuming its an "exact" age table) and then go to 63, or is it an age last table, so I should use the age 62 rate for the next 12 months of my projection. I've read through the SOAs description of how the table was constructed, but I couldn't find a reference to the method tia
Neither of those. If you are doing a monthly rather than an annual calculation, you just interpolate between the ages. So 62 and 3 months is (age 62)*9/12 + (age 63)*3/12, etc.
#6
02-09-2016, 04:47 PM
 Fuzzy Member SOA AAA Join Date: Oct 2001 Location: Far from the madding crowd...but closer than I used to be Studying for the "final" exam Favorite beer: Carlsberg Posts: 623

Who says that linear interpolation is the way to go?
__________________
"Don't worry about the world coming to an end today. It's already tomorrow in Australia." (Charles Schulz)
"We created an environment where we didn't know what we were doing, but it was legal and making profits."(Bill Sharon, chief executive of Sorms)
"As soon as we solve one problem, another one appears. So let's try to keep this one going for as long as possible." (Pepper...and Salt, WSJ, 5/4/2011)
#7
02-09-2016, 05:48 PM
 Helen Sass Member SOA AAA Join Date: Oct 2015 Location: Nearby Studying for Colorectal Exam College: SUNY Alumni Favorite beer: #1 - Free; #2 - Cold Posts: 2,715

Quote:
 Originally Posted by Fuzzy Who says that linear interpolation is the way to go?
People on the internets.

Please feel free to use any interpolation method you prefer. My point was that if you are going to the trouble of valuing an annuity on a monthly basis rather than annual, I would interpolate between the ages for the monthly annuity factors rather than use the same annual last age (or whatever) factor for each of the 12 months.
#8
02-09-2016, 11:36 PM
 Colymbosathon ecplecticos Member Join Date: Dec 2003 Posts: 6,017

Quote:
 Originally Posted by Helen Sass People on the internets. Please feel free to use any interpolation method you prefer. My point was that if you are going to the trouble of valuing an annuity on a monthly basis rather than annual, I would interpolate between the ages for the monthly annuity factors rather than use the same annual last age (or whatever) factor for each of the 12 months.
The question is what do the ages in the table mean. For example, if (65) is the average of (64.5)+ to (65.5)-, then the average age for (65) is 65 (ignoring mortality between the endpoints) and your method makes sense.

On the other hand, if (65) us (65.0)+ to (66.0)-, the average is 65.5, and your method needs to be modified.

I think that this is the OP's original question.
__________________
"What do you mean I don't have the prerequisites for this class? I've failed it twice before!"

"I think that probably clarifies things pretty good by itself."
#9
02-10-2016, 10:29 AM
 Helen Sass Member SOA AAA Join Date: Oct 2015 Location: Nearby Studying for Colorectal Exam College: SUNY Alumni Favorite beer: #1 - Free; #2 - Cold Posts: 2,715

Quote:
 Originally Posted by Colymbosathon ecplecticos The question is what do the ages in the table mean. For example, if (65) is the average of (64.5)+ to (65.5)-, then the average age for (65) is 65 (ignoring mortality between the endpoints) and your method makes sense. On the other hand, if (65) us (65.0)+ to (66.0)-, the average is 65.5, and your method needs to be modified. I think that this is the OP's original question.
Quote:
 Originally Posted by MathGeek92 My question really is if I were using RP-2014 to value the expected future benefits of someone today that is age 62 and 3 months... would I assume the age 62 mortality rate for the next 9 months (assuming its an "exact" age table) and then go to 63, or is it an age last table, so I should use the age 62 rate for the next 12 months of my projection.
My apologies if I misunderstood the OP (and the Duke's response). When they referred to "using age x" I thought they were talking about what to use to value the benefit, not what was used to construct the table.

I agree with the Duke that the underlying basis for this table (and I believe for most "official" mortality tables) is assumed to be age nearest which I believe is consistent with your first paragraph.

I would still answer the bolded question above much the same way. To value the benefit for someone age 62 and 3 months, if I was calculating on a monthly basis I would interpolate between the years for each month as I described, rather than use the age 62 rate for either 9 or twelve months and then switch to the age 63 rate.
#10
02-12-2016, 01:25 PM
 MathGeek92 Member Join Date: Jul 2004 Posts: 724

In order to properly understand how to use the table I need to know how it was constructed. If the table is birthday driven in it's exposure and experience, then it's exact age, I should use exact age assumptions when valuing a given benefit stream. If the table is based upon a policy age at issue, then I would need to know if it was an age near or age last table.

I think how the table is constructed is paramount in being able to use the resulting q's correctly.

I would think given the RP table is a pesion plan table, the q's would have been an exact age exposure basis. Most insurance products are a "policy age" basis (near or last)

Sorry if I made my question more confusing.