Controlled Group -- separate plans

[R. Butler]

Company A owns 80% of Company B. Company A & Company B have separate 401(k) plans. Company A has 4 benefitting HCE's and 44 benefitting NHCE's. Company B has 5 benefitting HCE's and 6 benefitting NHCE's. It seems to me that Company B's plan can't pass 410(B) on its own and thus the plans must be aggregated for testing.

We have just taken over the administration of Company A's plan only; Company B remained with the prior administrator. I have nondiscrim. tests from the past 5 years. The plans have never been aggregated. I have a copy of Company B's test for 2001 and the administrator did not aggregate. (The other administrator involved is national investment firm. I know they are aware of both plans; the same people actually administered both plans until now.)

Am I missing something? Don't they have to be aggregated?

[Tom Poje]

I would agree, sounds like you have a controlled group. I also agree that plan B will fail if tested separately.

hopefully the ADP tests will still pass if aggregated

[Mike Preston]

Seems like it is right on the cusp, though. Assuming your numbers are 100% correct (more later), the 410(B) percentages appear to be (for the second plan):

(6/50) / (5/9) = 21.6%.

Assuming the average benefits test is satisfied (highly likely) and assuming the concentration percentage is:

50/ 59 = 84.746% == > 84%,

then the Safe harbor percentage is 32% and the un-safe harbor is 22%.

21.6% is very close to 22%.

In this case, I would think that facts and circumstances might very well lead to the use of the unsafe harbor.

So, check your numbers very carefully (especially the concentration percentage) because if you can move the concentration percentage to 85%, the unsafe harbor becomes 21.25% and, with facts and circumstances, the plans just might pass 410(B) on their own, and therefore be able to be tested separately.

[R. Butler]

I'm going through the Average Benefits Test on this. Just to be clear:

1. Are the components (401(k), 401(m), 401(a)) tested separately for the modified facts and circumstances test? I believe so, but Relius confuses me appearing to only show results for 401(a) component.

2. I know that all components are aggregated for ABP test.

3. Assuming I am correct on #1 and #2, 401(k) portion meets the unsafe harbor percentage, but 401(m) does not? Do I aggregate for ACP and test separately for ADP?

Thanks for any guidance.

[Tom Poje]

if you pass the unsafe harbor but fail the safe harbor, then the IRS has the final say, though the following would be considered:

a letter from your mommy saying its ok

% of total employees benefitting under the plan

difference between plan's safe harbor % and plan's ratio percentage (how close to passing were you)

passing margin of average benefits percentage test

and whatever classifications might come into play

..........

for example, we had a controlled group scenario in which one plan had a 100% match up to 3% and the other had a 50% up to 6%. Turned out all but one or two people in the 100% match plan deferred over 6%, so in effect the two plans had similar formulas. If one more nhce had benefitted the plan would have passed safe harbor. The plan had never failed before, this year there were 2 new hces (had barely made 80,000 in prior year) and next year they would be nhces again because they didn't make 85,000 in the current year. Lots of various facts and circumstances to consider, huh? Of course, the IRS could still always say no.

.............

it becomes more confusing on Relius if you are using the fields

'number of non excludables nonbenfitting ees'. then the coverage

#s are correct for 401(a)(4), but if I recall, they will not be correct for 401(k) and 401(m)