It looks too good!

[Guest saeissler]

I am proposing 2 plans. The 2 HCEs are in the DB plan and the 3 NHCEs are in the DC plan. The plan passes 401(a)(26) because 40% of the nonexcludable employees are in the DB plan. The plans together pass 410(b) because everyone is in one plan or the other. The plans together pass 401(a)(4) because the NHCEs have the 7.5% minimum gateway for DB/DC plans and each of the two rate groups is over 70%. The DC allocations are converted to accrual rates using 8.5% and 1983GAM. Everyone is tested at the plan's normal retirement age of 56.

Did I miss anything?

[AndyH]

Hi Susan:

1. Does the DC plan also have a NRD of 56? If not, your testing age is 65.

2. You need to test the MVAR, and in most people's opinion, that would reflect the subsidy caused by 417(e).

Other than that, I think you are ok (albeit high maintenance).

[Guest saeissler]

1. I can make the NRA 56 for both plans.

2. If the actuarial equivalence in the plan document will be per 417(e) rates, and I am testing on a benefits basis, and there is no early retirement, then the MVAR and EBAR will be the same at each age, correct? Since the DC contributions are tested using standard assumptions, including an 8.5% interest rate, I see that I get some leverage out of the difference in assumptions, but I don't see anything to prohibit that.

[AndyH]

No, not correct on the MVAR. As I understand it, in the simplest case assuming SLA and no preretirement mortality, the MVAR would be calculated at the participant's current age as follows:

Actual lump sum payable (calculated using greater of plan rates or 417(e)) based upon benefit earned during measurement period, divided by plan's J&50 APR at current age, multiplied by testing APR as current age, projected to Testing Age using testing interest, divided by Testing Age APR using testing assumptions, divided by testing comp, and divided by testing service. That will normally produce a much higher MVAR than EBAR.

This assumes that J&50 is the Plan's QJSA.

Not everyone agrees that this is required. Most do.

[Guest saeissler]

I thought that the MVAB calculation was to take into account any additional value caused by early retirement benefits or subsidized benefits, like J & S. But if no benefits are subsidized, it doesn't seem like this calculation would be required. If it is required, it would seem like the same calculation would be required for the EBAR?

[AndyH]

You wouldn't consider a discount rate of 5% or less a subsidy when testing assumptions are 7.50% to 8.50%? Therein lies the issue.

Some people maintain that only a fixed rate lump sum calculation need be included in the MVAR.

Due to the uncertainty, I would make sure that it passes either way, and go for a FDL,

[Blinky the 3-eyed Fish]

I second Andy's comments.

Also with these type of designs, which I assume you are doing to avoid 404(a)(7), you need to be very careful in how you word eligibility. One new eligible participant can ruin the whole plan if not properly placed in the appropriate plan (DB in this case so 401(a)(26) continues to pass). Also that new person can wreck your tests, so that what if scenario should be considered and communicated to the client.

[Guest penman]

Andy,

In your first post on this thread, in 1. you said "If not, your testing age is 65". That is if the DC plans NRA is 65 otherwise you use the later of the normal retirement ages. In other words if it's DB NRA = 56 and DC NRA = 60 then the testing age is 60.

Do you agree?

[AndyH]

Yes, good point penman.

[Guest merlin]

Relius now calculates the MVAR by projecting to TA w/ interest and mortality, even where the plan definition of AE uses interest only pre-retirement. According to Frank Prager this is based on Q&As from the 2004 and 2005 EA Conference. Is anyone else aware of this?

[AndyH]

Thanks for pointing that out, Merlin. I have not seen any discussion of that. Yes, the Q&A's (2005-31 and 2004-25) seem to clearly require pre-retirement mortality (for DB testing only-not DC).