Each Prt Allocation Groups-Safe Harbor ?

[JAY21]

Our pre-approved Volume SUbmitter plan has the language in it that allows for the selection of "each participant being their own allocation group" (often used in conjunction with cross-testing). Now, if in a given year if client just wanted to do a regular integrated allocation (which normally is a safe-harbor formula) do my allocations have to be general tested given the afore stated plan doc structure (each participant their own allocation group) or can I claim it's still a safe-harbor contribution not subject to general testing ?

I guess the issue is whether the plan doc specifically needs the integrated formula specified in the plan doc in order for the integrated contribution to avoid the general test and be deemed a safe-harbor contribution/formula. Since I can "impute" permitted disparity into the general test, maybe the whole issue is a moot point if the general test (w/imputed permitted disparity) pretty much guarantees me to pass anyway using standard permitted disparity contributions. Thoughts ?

[Blinky the 3-eyed Fish]

You do need to general test because your document doesn't have a safe harbor formula. But as you point out in your last sentence, the question is moot if you are going to integrate using 100% of the taxable wage base because the plan will pass on a contributions basis imputing permitted disparity.

[JAY21]

Blinky, thanks for the info. It makes me wonder then if there's an argument for always going with the "each participant own allocation group" for virtually all PS plan designs. I realize you probably need some ER resolutions on contributions to define the contributions for each employee group, but other than that, would there be any reason NOT to use this as a default PS formula design since you can always fit a pro-rata or integrated contribution within this structure and be virtually assured of passing (same contribution rate for each participant) or some other more elaborate allocation as well.

[AndyH]

The reasons why IMHO this is a very bad idea could include the following:

1. The plan could fail coverage if it does not pass the ratio/percentage test.

2. It could be nightmarish to administer. Who decides the amounts, and when, and what happens when they change their mind? What happens if the allocations fail 401(a)(4)? How would it work for a partnership? What do new employees get? What about top heavy minimums?

[jquazza]

Plus the fact that you need a VS or ID document and then apply for a FDL or you can't use most of the self-correction methods when something goes wrong.

[Guest merlin]

Andy,

Could you please elaborate on your comment no.1 above? I thought it had been previously established on these boards that individual allocation classes, as opposed to eligiblity classes, are OK because the issue of a reasonable classification only arises when using the ABT to pass coverage , not nondicrimination. FWIW, Tom Finnegan's presentation on this issue (he called them "Dicretionary Bonus Profit Sharing Plans) at the Mid Atlantic Benefits Conference last May was very favorable towards this plan design. He called it "the ultimate flexibility in design", even allowing for zero allocation classes,as long as the plan is not top heavy. Jim Holland was on the speaker's platform with Tom, and didn't disagree with anything Tom said.

WRT your comment no. 2, these considerations apply to any class allocation plan, maybe to any plan at all. Clients always change their minds. Partnership calcs are always a bear. Classes by name don't add any more complexity. The client may end up with too many decisions to make between classes, but that's his problem , not mine. We have a client with 12 allocation classes, by real job categories, covering about 50 employees. The plan has been in existence 2 years and each year so far he's given the same % of pay to all employees, regardless of class. Before you ask, we didn't design the plan. It was done by another firm, but the reationship went sour early on. l

[AndyH]

Merlin, I refer to the argument that a 0 allocation has the same effect as an eligibility exclusion based upon name or some other individual reference that is not reasonable based upon objective business criteria. Therefore, the argument goes, the average benefits test is unavailable. Yes, this is a coverage issue only.

I recall this from a Q&A at an ALI-ABA session a couple of years ago. Positions and opinions do change of course, and I cannot speak to the conference to which you refer. But this position is in print and I happen to think that it has merit.

Regarding #2, yes, I agree that these issues are always difficult; I just think they become more difficult when there are multiple parties to the decisions and that more decisions need to be made. But this is just from a practical administrative perspective. Theoretically it is terrific. Maybe it belongs in a classroom like some of the economics theories I learned in college; not necessarily the real world. Just an opinon, that is all.

But thanks for the comments. Your views and comments are always appreciated.

[Guest merlin]

Andy,

Interesting about the 0 allocation, I hadn't heard about that. Was the question asked in reference to named classes, or class allocations in general? Most of my clients are top heavy anyway, so it's probably academic. But you never know...

If you're interested I'll send you Tom Finnegan's outline.

[AndyH]

I believe it was individual classes. I thought I still had the Q&A but instead I have a link to it that is no longer valid.

Sure, I'd love to see the outline.

[JAY21]

Good discussion. I hadn't heard about the 0 allocation issue either, but can see where that could be a real problem if that's the interpretation the service takes on that (unknown for sure it sounds like).

[Tom Poje]

see #3 at the following

http://www.abanet.org/jceb/2001/qa01irs.html

(and then will someone please organize all these notes I have so I won't have to keep looking them up)

[Guest merlin]

OK, there it is.

What about the IRS response to #4? Wouldn't that reasoning apply to any testing methodology? I don't think Corbel includes any testing language, except for ADP/ACP and (optional) 410b fail-safe language.

[Belgarath]

Jquazza - just one quick observation here - for VS plans, a current favorable advisory letter for the VS plan itself is considered a favorable letter for Voluntary Correction purposes. See Section 5(.01)(4) of Rev. Proc. 2003-44. I'm assuming, of course, tht there's no deviation from the approved VS doc, otherwise I agree that a FDL is required.

[jquazza]

Belgarath,

I agree with you, I should have qualified my statement. You have to admit some people do push it to the limit with the VS plans and I am not so sure the original doc provides much reliance, especially if you establish one class for each individual, to me, that sounds like a substantial deviation.

[Belgarath]

Ayup - and even though the IRS apparently doesn't have a problem with it, I still can't get myself to feel comfortable about it. We don't do it in our plans.