Testing option consistency for (k)/cross tested PS plan

[AndyH]

As I understand it, if separately testing otherwise excludable employees is used for ADP testing purposes, we must do the same for 401(a)(4) testing purposes. True?

If true, is it also true that you may not separately test otherwise excludables for 401(a)(4) if you do not test for ADP/ACP on the same basis?

What if the ADP/ACP testing is done by using prior year results, does this have any effect on your 401(a)(4) options?

[Mike Preston]

The only a4 test applicable to the 401(k) plan is the ADP test. Any other plan (profit sharing contributions) is mandatorily disaggregated from the 401(k) for a4 testing.

So I'm hardpressed to come up with any tie between the two. Maybe some application of benefits rights and features, but not regular a4 testing of non-deferrals, non-matching company contributions.

Am I misunderstanding?

[Tom Poje]

I would agree with Mike.

my understanding would be the following:

for 410(B) I have 3 tests

401(k)

401(m)

non-elective

I can statutory test on all of those or one of those or any combination.

but whatever ones I statutory exclusion test I would also have to test sin the same fashion the similar amounts test, e.g. ADP, ACP or a(4)

The main question is what happens when you get to the average benefits % test. I recall reading how to do this in the ERISA Outline Book (or at least I think I do) but at the moment I am a bit bogged down to look it up! it seems to be the busy time of year!

[AndyH]

Thanks, both, for your response. My question wasn't as clear as it could have been. Tom's final comments hit it on the nose (and believe me I do understand time constraints this time of year!).

What I was trying to get to is the average benefits portion of the 401(a)(4) test. For some reason (and it's the book Tom is referencing), I'm under the impression that if you permissively disaggregate for ADP, you need to do the same for the average benefits portion of the a(4) test. Is this true, and is the contrary situation also true is what I was trying to ask.

[AndyH]

Now I'm second guessing myself (with some help from both of you, which I very much appreciate); maybe this is not an issue.

The profit sharing a(4) test shouldn't be affected by the 401(k) 410(B) test, even if the k peice has to be brought into the average benefits test portion of the profit sharing a(4) test, I suppose, because it's the profit sharing portion that's being tested.

[Mike Preston]

You got it, Andy. The much more interesting question is whether there can be multiple AB tests when a plan sponsor tests under 410(B) by separating into those statutorily eligible and those not statutorily eligible. I don't want to re-write the long message I wrote on this subject a while back, but suffice it to say that the IRS has informally indicated that there aren't, while a strict reading of the regs indicates that there must be.