Average Benefits Test and Mortality

[Guest mdelin]

Do I need to use APR when calculating EBAR in my ABT as long as I am using 8.5 interest rate?

[Tom Poje]

well, 1.401(a)(4)-8(b)(2)(ii)(B)

Normalization says ...A standard interest rate, and a straight life annuity factor that is based on the same or a different standard interest rate and on a standard mortality table must be used....

under the definition of mortality tables found in 1.401(a)(4)-12 there is a list of permissible mortality tables. I don't see 'none' as an option

[Guest mdelin]

Hi Tom,

That was my understanding as well. Thank you.

I currently have a bit of a conundrum:)

Here is the scenario:

plan is part of ASG. document was written so that at end of year allocate any contributions necessary to NHCEs in Co #2 in order to pass ABT. An actuary that has since left our company did the ABT calculations in excel. I am running the ABT in Relius and when manually applying the formula to calculate the EBARS, I come to the same results as Relius.

When plugging my numbers into the spreadsheet that has been used in the past, I get a different result. This is what the two results look like:

Relius: Allocation(1.085^23*12/95.38)/Compensation

51,000(6.5295*.1258)/255,000

EBAR = 16.43

Spreadsheet: Allocation rate*(1.085^22.45)/10

.2*6.435289940/10

EBAR = 12.49

This is a Safe Harbor plan and I am not imputing permitted disparity. This is not a cross tested plan.

I can follow the formula used by the spreadsheet except for the division by 10. The 10 is a constant denominator for all. As you can see there is a slight rounding for what is being used as years to retirement with the spreadsheet being more detailed as Relius rounds. Using the spreadsheet I only need to allocate $1,426 to each NHCE in Co# 2 vs 1,879 in order to pass ABT in Relius.

Can I use the spreadsheet calculations?

Thanks!!

HAPPY LABOR DAY:)

[Tom Poje]

if everyone has the same retirement age and you do not impute disparity choice of mortality makes no difference - everyone would have their e-bar calculated by dividing by the same value.

e.g. the basic formula is

contribution * 1.0I ^ yrs to retire * 12 / APR

but if everyone has ret age 65 then the APR is the same for everyone - it is a constant.

the only time mortality table makes a difference is if you have someone past retirement or if you impute disparity.

you indicate Relius is using 23 years to retirement, the spreadsheet 22.45

that implies to me on Relius you are using age def last and if you used age def nearest Relius would use 22 years instead of 23

I have never found anything in the regs stating requiring using one method or another, as long as you are consistent amongs everyone.

[Guest mdelin]

Thanks Tom,

In your response I see that you use the formula equal to what I stated in my "Relius" example which yields an EBAR of 16.43. What I am trying to figure out is if I can use the formula used in the past in the spreadsheet. Not sure what the constant 10 stands for...... Does the second formula look appropriate to you?

As you can see the difference in the formula is that the *12/APR is not used but instead it is simply: contribution * 1.0l ^ yrs to retire / 10

Thanks!

[Tom Poje]

since the factor '10' is used on everyone, it is only a constant. you could have used 100 or 3.14159 for that matter.
without looking, I think Relius uses a similar factor for it's age weighted calculation to determine the number of points - I think for the only reason that you end up with a 'reasonable' value..e.g. points = 345.67 instead of something like 34567.890123 or 3.456789

again, if you don't impute disparity AND everyone is at the same testing age it really doesn't matter what you use as long as you have
contribution * 1.0l ^ yrs to retire as part of your formula. Note: if you use age definition nearest then people born in the first half of the plan year have 1 less year to retirement, so that will make a difference

everything becomes a constant. it wouldn't matter if you use UP84 which has an APR of 95.3829 or 1983 IAF 115.387 or a mortality table that produced a value of 10 (except that such a 'mortality' doesn't exists and therefore wouldn't be one of the tables permitted to be used under the regs.

but again, if everyone has the same testing age it makes no difference.