Control Group Testing

[Guest tcunagin]

I am needing clarification on testing for control groups. Is it true that if each separate entity covered by the control group passes coverage testing separately then they can process all other nondiscrimination tests separately as well?

[jaemmons]

generally, yes.

[Archimage]

You still must test all employees of the controlled group in each plan. You cannot test just the employees of the participating employer for that one plan.

[KJohnson]

As an example of what Archimage is talking about suppose that you had separate plans and coverage in Company A and Company B as follows:

Company A

1/1 non-excludable HCE's covered

3/3 non-excludable NHCE's covered.

Company B

1/1 of non-excludable HCEs covered

50/50 of non-excludable NHCE's covered.

If both Company A and Company B were treated as separate entities, your 410(B) ratio percentage would be 100% in each.

However, when you test coverage on a controlled group basis you would fail ratio percentage for Company A's plan because you would only be at 11.3%.

However, if both plan A and plan B passed coverage testing on a controlled group basis (for example if plan of Company A covered 50/50 non-excludable NHCES) then you would only apply your ADP test for those individuals who were actually eligible to participate in each Plan and not on a controlled group basis.

[jaemmons]

KJohnson,

If the plans had different profit sharing allocation formulas (e.g.- one is comp to comp the other is cross-tested), do you agree that all ee's of both corps would be counted in the 401(a)(4) testing? I have always treated them as such, but have received conflicting opinions on the matter.

thanks

[jaemmons]

KJohnson,

If the plans had different profit sharing allocation formulas (e.g.- one is comp to comp the other is cross-tested), do you agree that all ee's of both corps would be counted in the 401(a)(4) testing? I have always treated them as such, but have received conflicting opinions on the matter.

thanks

[Mike Preston]

jaemmons - yes.

[Mike Preston]

tcunagin - getting back to your original question, if we accept the fact that each plan satsifies the coverage rules (section 410(B)) then you can definitely test each separately under 401(a)(4).

However, sometimes testing under 401a4 will require you to collect information on plans not being tested. For example, if you need to know the results of the Average Benefits Test you will almost always need to collect information from other plans.

There is an exception to the exception, by the way. If one entity has a defined benefit plan and another entity has a defined contribution plan, you can test entirely separately even under the Average Benefits Test.

If you lay out the specifics of what your groups look like and what the plans look like, a quick determination may be possible.

[KJohnson]

Mike,

I just wanted to reconcile your last two answers. In your response to jaemmons does your response simply reflect the fact that because you have a cross-tested plan involved, each rate group in the cross-tested plan would have to satisfy 410(B) and therefore you would have to pick up information on employees from the entire controlled group.

However, if you, for example, had two plans in a controlled group that passed 410(B) and one had a "comp to comp" formula of 10% and the other had a "comp to comp formula" of 3% then each plan would be tested on its own populaiton for 401(a)(4) and that since each had a safe harbor formula there would be no issue 401(a)(4) issue?

[Mike Preston]

Yes, in jaemmons case that was posited there is a crosstested plan involved. So, as you noted, you must test the rate groups counting the bodies of all in the controlled group.

In the second case,where the formulas are safe-harbor, there is no 401(a)(4) testing to be done once the plans are shown to satisfy 410(B). I think we are doing it the same way, but I find the phrase "each plan would be tested on its own population for 401(a)(4)" a bit of a misstatement in the sense that there really isn't any 401(a)(4) testing. This is semantics, because I understand that one can use the existence of a safe-harbor formula as the mechanism one uses to satisfy the 401(a)(4) test. I usually resever the word "test", though, to mean something that I have to prove numerically. Ignore my ramblings, I know that you and I are on the same page!