Cash Balance - Nondiscrimination testing

[flosfur]

Just took over a Cash Balance plan. The prior actuary, treated it as a DC plan and cross tested it based on benefits by accumulating the contribution credits to NRA and converting them to monthly benefits using 8.5%/8.5% & standard motality. Based on this, the plan passes 401(a)(4) with extreme ease! Is this is how a Cash Balance is tested?

A Cash Balance plan is a DB plan in the first place! So shouldn't the equivalent benefits based on the plan assumptions (5.5%/5.5%...) be used for testing? Based on the Normal Accrual Rates using these equivalent benefits, the plan falls fails the test miserably!!

[AndyH]

Just took over a Cash Balance plan. The prior actuary, treated it as a DC plan and cross tested it based on benefits by accumulating the contribution credits to NRA and converting them to monthly benefits using 8.5%/8.5% & standard motality. Based on this, the plan passes 401(a)(4) with extreme ease! Is this is how a Cash Balance is tested?

A Cash Balance plan is a DB plan in the first place! So shouldn't the equivalent benefits based on the plan assumptions (5.5%/5.5%...) be used for testing? Based on the Normal Accrual Rates using these equivalent benefits, the plan falls fails the test miserably!!

I think you are right but also consider that there is a "safe harbor" testing method in 1.401(a)(4) somewhere which I think allows testing similar to if it were a dc plan being tested on an allocations basis. Not the answer here, but something else to make sure they didn't rely upon.

[John Feldt ERPA CPC QPA]

Unless they test the DB on a DC basis (as AndyH mentions), for testing, I would think the cash balance hypothetical accounts should be projected at the plan's interest crediting rate, then converted using the plan's definition of actuarial equivalence.

I doubt the interest crediting rate is 8.5%, it is more likely based something from 96-8 or possibly the long-term corporate bond rates. The conversion of that projected balance at normal retirement age could be 8.5% if that's defined in the plan's definition of actuarial equivalance.

[flosfur]

Unless they test the DB on a DC basis (as AndyH mentions), for testing, I would think the cash balance hypothetical accounts should be projected at the plan's interest crediting rate, then converted using the plan's definition of actuarial equivalence.

I doubt the interest crediting rate is 8.5%, it is more likely based something from 96-8 or possibly the long-term corporate bond rates. The conversion of that projected balance at normal retirement age could be 8.5% if that's defined in the plan's definition of actuarial equivalance.

Testing of DB as DC involves converting accruals into equivalent lump sum cash, dividing that by comp to give allocation % - right?

If they were testing Cash Balance as DC then they should have used the contribution credit rates which are definitely discriminatory - 75% for owner and 5% for all others (1 HCE and 9 NHCEs).

Plan's interest credit is 5.5% and Plan's assumptions for converting balances are 5.5%/5.5% and Applicable mortality.

Do I have an agreement what the previous actuary did was incorrect ?

[Effen]

Do I have an agreement what the previous actuary did was incorrect ?

I'm sorry, I was laughing so hard when I first read it, I though you were just providing some comic relief after a difficult AFTAP season.

Yes, I agree - the previous actuarial person was WAY off.

[John Feldt ERPA CPC QPA]

Until the IRS (someday) comes out with the "fixed rate" guidance for cash balance plans, the 8.5% is not usable as you describe. I doubt the eventual guidance would allow it either.

[ak2ary]

Be careful here though. If a plan does not have a uniform NRA, in determinin the normal accrual rate, you would use the plan's interest credit rate to project to NRA and the a standard interest rate to project to a uniform testing age. Prior to the changes to the NRA rules, there were many many plans that used a concept of "early-normal" retirement age. Under this design, the NRA is usually set to the EARLIER of age 65 or the completion of 5 years service. Thus if all your employees had at least 5 years service, all of your projections would be at 8.5%, regardless of the plan's interest credit rate.

Since the finl regs on NRA put a damper on this NRA definition, these designs have gone away ...but were certainly around for 2006 tests...sp depending om plan terms the old actuary may be right

[John Feldt ERPA CPC QPA]

Seems like a stretch, but I'll concede that it is remotely possible.

If everyone has at least 5 years of service in your example, the are no years left for the "projection" at 8.5%, so it's merely a conversion at 8.5%.

[ak2ary]

I agree that its not likely. but it is something you should look at first before you call the prior actuary and accuse him of being wrong

[JBones]

I agree that for calculation of the Normal Accrual Rates, the hypothetical account balance would be projected and converted using the plan's rates, but for the calculation of Most Valuable Accrual Rates, if you are defining the most valuable benefit as lump sum, then you would use 8.50% to project and an allowable table other than the plan's AE to convert to an annuity at testing age. I have had some IRS reviewers insist that the MVAR is based on lump-sum and others insist that is is based on J&S.

[ak2ary]

I understand that there were reviewers questioning lump sum vs qjsa...but the regs are absolutely clear that its qjsa. The IRS national office has been asked this question and agreed its the qjsa. they apparently have communicated this to the field agents. That don't mean the field agents got the message but I understand that if you push back on the ossue, they cave.why are you using a mortality assumption other than the plans table ( assuming the plan has a standard table)?

[AndyH]

I agree that in the final analysis the regs do say that the lump sum is not reflected in the mvar, but with all due respect I disagree that it is clear. I think the path to that conclusion is murky and filled with potholes.

[Calavera]

I think the following is how the IRS looks at it. MVAR is based on QJSA payable at earliest retirement date. Cash Balance plans offer an immediate lump sum to terminated employees. In order to offer the immediate lump sum, they also need to offer the immediate QJSA. So for the MVAR purposes, there is the QJSA payable immediately that needs to be normalized with the standard rate.

[JBones]

why are you using a mortality assumption other than the plans table ( assuming the plan has a standard table)?

I usually 8.5% and one of the standard tables other than the DB plan's table as my testing assumptions. When normalizing the most valuable benefit, 1.401(a)(4)-12 says that the assumptions used to normalize the benefit must be reasonable and that standard interest and mortality tables are considered reasonable. In either case, qjsa or lump sum, the most valuable benefit is normalized using the testing assumptions. Not that it can be relied on very much, but this matches the results that I have gotten from several discrimination software systems.

[Effen]

I think this discussion is just going around in circles.

JBones - could you give a specific example on which we can comment.

In general, a cash balance plan is a defined benefit plan and the plan document must contain language regarding how the cash balance account is converted into an annuity. This annuity is the basis for your non-discrimination testing. You don't "normalize" the cash balance account into an annuity like you would in a DC plan.

"but this matches the results that I have gotten from several discrimination software systems." Do I really need to say it? Software generally does what you tell it, most software systems give you plenty of rope to hang yourself.

[JBones]

I’m not arguing whether J&S or lump sum is the correct way to calculate the MVAR. I would rather use J&S because the results are typically better in a cash balance plan. But here’s an example of how I have always calculated the MVAR when required by an IRS reviewer to use the lump sum. It is exactly the same calculation as if it were simply a profit sharing allocation.

Plan AE is Applicable Interest and Mortality (4.68% with 94 GAR for 2007)

Normal Retirement Age / Testing Age is 62

Testing Assumptions are 8.50% ‘83 IAM Male

Participant is age 52

Hypothetical Contribution Credit for 2007 is $50,000

Compensation is $225,000

APR’s:

4.68% 94 GAR age 62 sla 13.06008

4.68% 94 GAR age 52 sla 15.85085

8.50% 83 IAM age 62 sla 9.32642

8.50% 83 IAM age 52 sla 10.54998

If the immediate lump sum is assumed to be the most valuable form of benefit payable under the plan, then it is normalized by converting the benefit (the lump sum) to an actuarially equivalent straight life annuity commencing at the employee’s testing age using the testing assumptions per the definition of normalize in 1.401(a)(4)-12.

$50,000 x 1.085^10 ÷ 9.32642 = $12,121.39

MVAR = $12,121.39 ÷ $225,000 = 5.39%