Definitely Determinable Allocation Formula

[LANDO]

So I have a plan sponsor that wants to draft their profit sharing allocation formula to allow them to decide on an annual basis if they’re going to make either a pro-rata based on comp contribution, a contribution using permitted disparity (two tiered), or both.

Seems to me, with all this discretion this “formula” with all its flexibility would violate the definitely determinable allocation formula requirements. In my mind, I’ve always thought discretion meant a sponsor could determine the amount they will contribute each year, and a participant should be able to determine from the formula how that amount would be allocated. That said, you can put each participant in their own allocation group, so does “definitely determinable” really mean anything anymore?

Since an allocation using any of the possibilities this sponsor is contemplating (pro-rata, integrated, or both) would meet the permitted disparity rules, seems like they would still have a safe harbor allocation formula. Agree?

They also are expecting us to jam this formula onto our VS document…a long describe line! What prevents a sponsor from selecting all the allocation options in a VS document and saying they’ll choose one at the end of each plan year?

Maybe I just don’t like stuff that is out of the box, but this doesn’t smell right to me.

Anyone care to share their thoughts?

[Tom Poje]

well, you could put each person in their own group.

let's say year one you give everyone 5%.

If you were to test on an allocation basis, everyone would be at 5%.

Now year 2 you decide to allocate on a formula that works out to be 100% TWB (using 5.7% as the max excess percent)

you again test on an allocation basis (imputing disparity)

lo and behold, everyone will have the same 'allocation % plus disparity %' simply because that is the way the testing formula works.

so the net effect is that, yes, you sort of do have a safe harbor allocation formula.

it always cracks me up the regs say you don't have to test (i.e. you have a safe harbor formula) if you allocate the same % to everyone, or if you allocate integrated provided no one receives more than 5.7% excess, because if you wanted to you could test on an allocation and achieve the same results.

[Bird]

Pretty much what Tom said - you can effectively do what they want by just having everyone in their own group and general testing it. It might seem silly that that is definitely determinable (it's not, by my definition, but the IRS says it is so it is), but IMO having multiple formulas available and choosing one for the year is not definitely determinable...or if it is, then it is not a safe harbor and allocations must be general tested.

If I may...how do sponsors come up with these weird ideas? I think someone (mis)planted an idea and it grew like a weed that must be plucked out. In my (small plan) world, they know less than nothing; I get a general idea of what they want to do, and I tell them how we're going to do it. We do not have these conversations...

[MoJo]

Just a thought - why could you not have TWO profit sharing contributions authorized in the plan, one pro-rata, and one integrated, both "discretionary, and then each year, the employer makes a determination of how much, if any, to contribute to each of the two sources? It would require two "buckets" on the record keeping system (a Profit Sharing "A" (non-integrated) and a Profit Sharing "B" (integrated).

Haven't rolled up my sleeves to thoroughly "vet" this option - but I've seen various schemes that border on this....

[LANDO]

Bird, in an ideal world that's how it should work, but on my larger plans sponsors sometimes engage outside counsel to "review" their amendments and restatements, and of course they need to justify their fees and show why they should be included in every minor decision and amendment, and I guess that's what I'd do if that's how I got paid, but that was the genesis of this debacle. Fortunately it doesn't happen that often. The funny part is, the way the outside counsel drafted the permitted dispartiy language they want included, it would fail the 401(l) safe harbor. Sheesh!

MoJo, I don't know that you'd need to recordkeep the two contribution types separately since all the non-elective contributions are going to be tested together anyway. It does make some sense from the perspective that you could certainly set up a separate plan for each contribution. I guess if you can do something with two separate plans, you should be able to do it with a single plan???

Tom, I haven't really thought through the math, but if a pro-rata or integrated plan had to pass 401(a)(4) I think it still might be possible to fail non-discrimination if they have allocation conditions. As in, they pass coverage, but becasue you bring zeros into the averages for testing 401(a)(4), you could fail non-discrimination. So I could be wrong, but I believe there is still an advantage to having a formula that qualifies as a fail safe. Maybe that wasn't even your point, but I thought is was in interesting intelectual exercise.

[Tom Poje]

Lando - well, lets suppose you had ps only.

10 NHCE, 1 HCE (a real simple case, but changing the numbers shouldn't matter.

3 NHCEs get nothing.

so you pass coverage at 70%, and rate group testing would also pass at 70%.

now suppose 4 NHCE get nothing

so you fail ratio % for coverage, but because of deferrals you pass avg ben pct test.

so you pass coverage.

now you run the rate group test and - same results!

I'm not sure there is a scenario if the formula is comp to comp or integrated where it could pass coverage and not pass nondiscrim.

technically if you have a bunch of people getting top heavy only you could pass coverage and fail nondiscrim, but then you have changed the rules because not everyone is receiving the same %.

[LANDO]

I guess that's right Tom. If they didn't benefit because of an allocation condition they'd get counted as not benefiting under the coverage test, AND they wouldn't be in the HCE's rate group...same same. That's a good point about a top heavy plan, and a point in favor of a safe harbor allocation formula...free pass.

Not to get too deep in the weeds, but we're pretty far in the weeds already. What if you have 2 HCEs and only 5 or 6 NHCEs benefiting out of the 10 total NHCEs, but I'm still able to pass coverage using ABPT? Am I still always going to pass non-discrimination testing, or does that depend on the demographics? If I'm thinking about it right, I might pass coverage, but still fail nondiscrimination if I had to pass the general test.

[Tom Poje]

without seeing actual numbers, I still think if the allocation was pro rata.

(or integrated) and you pass coverage (because of the avg ben pct test) then the plan would most likely pass nondiscrim. (again, most likely with testing on an allocation basis)

certainly if the allocation was different for different folks, then no, because that is what drives cross testing.

[Bird]

Bird, in an ideal world that's how it should work, but on my larger plans sponsors sometimes engage outside counsel to "review" their amendments and restatements, and of course they need to justify their fees and show why they should be included in every minor decision and amendment, and I guess that's what I'd do if that's how I got paid, but that was the genesis of this debacle. Fortunately it doesn't happen that often. The funny part is, the way the outside counsel drafted the permitted dispartiy language they want included, it would fail the 401(l) safe harbor. Sheesh!

Thanks for the feedback/sorry to question you. I've run into stuff like that too, and wondered how much the outside advisor/counsel was getting paid to be flat-out wrong.

[LANDO]

Bird, no sweat. Seemed like a reasonable point/question.

Tom, I think this is a pretty simple example where a pro-rata allocation would pass coverage and fail non-discrimination...

HCE 1 - Age 25, Comp $250,000

HCE 2 - Age 55, Comp $250,000

NHCEs 1-10, all Age 30, all Comp $50,000

5% Pro-Rata PS allocation, no other allocations.

Four of the NHCE's don't meet last day requirement and don't share

Plan can only pass Coverage on a benefits basis. Plan Fails non-discrimination on contributions basis (one rate group, Ratio Percentage 60%). Plan fails non-discrimination on benefits basis because there is no NHCE's in the 25 year old's rate group.

Same allocation for everyone...passes coverage and fails non-discrimination. Again, a moot point since we were talking about a safe harbor allocation formula in the first point, but an interesting exercise.

I think that brings me back to one of my original questions. Would the allocation formula they are proposing (a pro-rata allocation, or a two tier integrated allocation, or both) be considered a safe harbor allocation formula?

I get it if you're all exhausted with this post!

[Tom Poje]

there is no requirement a plan run the avg ben pct test the same way it runs the rate group test.

in fact, running the tests on different methods is part of what is known as component plan testing. one avg ben pct test is run, then the employees are split into 2 groups - one group tested on an allocation basis and the other group tested on a benefits basis.

so if the avg ben pct test is > 70% (tested on a benefits basis) and the rate group test for each NCE > 50% (tested on an allocation basis) then the plan would be deemed to pass nondiscrimination.

in other words, the statement "Plan can only pass Coverage on a benefits basis." is not quite true, as the coverage test can be done on either

1. ratio percentage test or

2. avg ben test, which consists of

a. avg ben pct test on a benefits basis

b. nondiscrim classification test on an allocation basis

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

the only possible exception I can think of is if each person is in their own group which could throw the nondiscrimination classification test out the window, due to 1.410(b) -4(b)

[LANDO]

Sorry about that Tom. I should have been clearer. Let me rephrase...based on the example I gave, the plan passes coverage using the average benefits percentage test on a benefits basis, and fails 401(a)(4). If you're saying the plan I described can be restructed into component plans, and then each component plan can pass coverage and nondiscrimination, I'm all ears, not to mention that we are a long...LONG...way from saying any plan that uses a uniform allocation formula (pro-rata or Integrated) would pass nondiscrimination testing, and the allocation safe harbors don't really get you anything.

Any thoughts about my safe harbor allocation formula question?

[Tom Poje]

I'm not really saying the example you gave even needs restructuring. I was using restructuring as an example to show it is possible to run the avg ben pct test on a benefits basis and the other test on an allocation basis.

you simply run the avg ben pct test on a benefits basis, e.g. it passes

then you run the rate group on an allocation basis, it equals 60%

because 60% > 32.75% [the safe harbor % of the nondiscrimination classification test - in this particular example because 10/12 = 32.75%] and because the plan passes avg ben pct test the plan passes nondiscrimination. and coverage.

as to whether it is a safe harbor formula, I vote no, the document does npt indicate if the formula is pro rate or integrated (ignoring the possibility you might be able to have a formula with 2 discretionary possibilities. I don't really think that is possible)

now, by choice, the actual allocation may imitate a safe harbor formula and as a result pass testing when performed on an allocation basis (or at least the rate group portion)

[LANDO]

Tom, that registers with me. Thanks for sticking with me through this exercise, I very much appreciate that.

Thanks again.

[Mike Preston]

That said, you can put each participant in their own allocation group, so does “definitely determinable” really mean anything anymore?

Sure. You are confusing determining a contribution (which is discretionary) with determining an allocation (which is not, and has never been, discretionary). In order to be definitely determinable, either the plan document, or a separate instrument which is contemporaneous with the contribution, must specify how the contribution is allocated. So, as Tom says, you can easily put everybody in their own group and then have the client decide, when a contribution is made, who gets how much. They can, at that time, decided to allocate on a pro-rata basis or on an integrated basis or on any other basis that doesn't violate another statute (ADEA, 415, etc.).

I see nothing wrong with you using a very long "describe" line to establish two allocation methods and then determining the amount that goes through allocation formula 1 (pro-rata) and the amount that goes through allocation formula 2 (integrated) in the same way that one would in the case of individual allocation groups: contemporaneously.

I would think the combined allocation would easily satisfy 401(a)(4) general testing, but it may not be necessary to go through the effort because the allocation is considered a safe-harbor if it satisfies the rules for multiple safe harbor formulas (must be the only formulas, etc.).

I'd treat it as satisfying the safe-harbor rules knowing that if the IRS challenges it, I can do the general test at that time to demonstrate 401(a)(4) compliance.

The risk isn't 401(a)(4), it is 410(b). Continuing the example of 4 NHCE terminees with > 500 hours resulting in an NHCE ratio percentage of < 70%, there is the slight risk that the ABT is unavailable should the IRS assert that the covered group does not constitute a reasonable business classification (only those employed). The IRS likes to rattle sabres about this issue at conferences but I've never heard of it resulting in a problem in the wild. And, IMO, I find it hard to believe that a judge will say that benefits flowing solely to those employed at EOY does not constitute a reasonable business classification.