Non Discrimination Testing for Law firm plan

[Gary]

A couple of points I am going to present and would like either verification or difference of opinion if any exist. I'll try to keep things straight forward.

Background

Law firm has a 401k only plan for associate attorneys and a

401k, 401m and profit sharing plan for the partners and rank and file employees. Only NHCEs receive a 401m match.

It is intended that the partners' plan passes without aggregation.

We'll assume that the partners plan by itself is top heavy.

1. Regarding the ADP test, my understanding is that we include all non excludable employees in the company for the coverage ratio test, but only include the non excludable employees in the partners plan when performing the ADP test?

2. The partners plan passed the rate group non discrimination testing by means of the average benefit test. While the ABT is based on rates for all non excludable employees in order to test for the average benefit percentage portion of the test, it is still not deemed a required aggregation for purposes of providing top heavy allocations for the associates, thus associates would not be required to receive a TH allocation?

3. And finally, while the partners plan is being tested on a cross tested basis, the assoicates plan should not be required to receive a gateway minimum? The ABT uses cross testing for the associates plan as well of course.

4. To make a long story short, meeting coverage and/or non discrimination by means of the ABT does not warrant required aggregation as far as I can see? Otherwise it appears that the partners plan is being tested as a separate and distinct plan.

Thank you.

[Mike Preston]

Correct on all counts, as long as the associates plan doesn't cover *any* key employees for even an instant.

[ak2ary]

Devils Advocate for a second....

Lets assume that the partners plan fails the Ratio Pct Test and must use the average benefits test

Assume that the ABPT is passed when the the plans (including the associates plan) are aggregated, as is required

Assume that the ABPT would fail if the rate for all associates is zero

So, but for the deferrals in the associates plan, the partners plan would fail 410(b)

T-6 of the 416 regs says:

In addition, each other plan of the employer which, during this period, enables any plan in which a key employee participates to meet the requirements of section 401(a)(4) or 410 is part of the required aggregation group.

So since the partners plan would fail 410(b) without the associates plan, there is a strong arguement that it enables the partners plan to meet the requirements of 410 and is part of a top heavy required aggregation group..

We have never gotten an answer from the IRS on this situation as to whether there is a required aggregation group. In fact, at one point at a national meeting Wick said "Believe me, you do not want guidance on this question"

I would be concerned that, in court, this position of non-aggregation would not hold up

(Of course, some of those associates will become partners. Some of those partners will be key. If the associate makes his way to key within 5 years of becoming a partner there is an argument that the associates plan benefited a ket employee in the last 5 years ...since we have never gotten guidance on what that means either)

[KJohnson]

The associates becoming partners is an issue. I think some plan designs mandate a plan to plan transfer of account balances once an associate becomes a partner. Not sure if this completely solves the top heavy problem but at least you will never have the same individual with an account balance in both plans.

[jpod]

Generally (not always, but generally), this structure is used when all or the vast majority of the associates are HCEs, and the associates cannot be Keys.

[KJohnson]

jpod,

How are you going to pass coverage in the Associate plan if all the Associates are HCEs? I assume that they are not HCEs' even if they are over the comp limit because of a top 20% election.

[jpod]

My understanding is that they pass coverage through aggregation, but none of the associates are keys so there would be no th contribution.

[KJohnson]

I think if you have to aggregate to pass 410(b) that kicks in the TH contribution for the associates assuming (as is highly likekly) that the plan is top heavy on an aggregated basis.

[jpod]

KJohnson: I'm not sure about your last observation. The associates plan does not help the partner's plan pass 410(b) or 401(a)(4). To the contrary, the partner's plan helps the associates plan pass 410(b).

[KJohnson]

Jpod, I see your point. I had it in my head that if you aggregate for 410(b) you aggregate for top heavy, but that might not be the case.

[Mike Preston]

Aggregating (as required) under the ABT is not aggregating under 410(b) for purposes of determining whether the associates plan is a "helper" plan as to the coverage/non-discrimination rules.

As you just indicated, in the vast majority of cases the plans, exclusive of the associates plan, would satisfy 410(b) just fine. The only reason the associates plan is included in the ABT is that it is required to be included pursuant to the rules for calculating the ABT.

If, OTOH, the associates plan is required to be aggregated with the other plans for purposes of enabling the other plans to satisfy 410(b), then it is highly likely that the entire she-bang (technical term, that) would then be top-heavy.

As indicated, if the associates plan does not stand on its own and is aggregated with the other plans in order to enable it to satisfy 410(b), then it is not required to be aggregated for TH purposes.

Again, the ABT portion of all of this is irrelevant to the determination of RAG's and th minimums.

[jpod]

[As indicated, if the associates plan does not stand on its own and is aggregated with the other plans in order to enable it to satisfy 410(b), then it is not required to be aggregated for TH purposes.]

This is the scenario I was talking about.

[ak2ary]

Aggregating (as required) under the ABT is not aggregating under 410(b) for purposes of determining whether the associates plan is a "helper" plan as to the coverage/non-discrimination rules.

As you just indicated, in the vast majority of cases the plans, exclusive of the associates plan, would satisfy 410(b) just fine. The only reason the associates plan is included in the ABT is that it is required to be included pursuant to the rules for calculating the ABT.

If, OTOH, the associates plan is required to be aggregated with the other plans for purposes of enabling the other plans to satisfy 410(b), then it is highly likely that the entire she-bang (technical term, that) would then be top-heavy.

As indicated, if the associates plan does not stand on its own and is aggregated with the other plans in order to enable it to satisfy 410(b), then it is not required to be aggregated for TH purposes.

Again, the ABT portion of all of this is irrelevant to the determination of RAG's and th minimums.

Mike,

I understand your point that required aggregation under the ABPT is not aggregation for 410(b) and that the IRS has said as much in gray books and the like when the question is posed as ".. and the only reason for the aggregation is that it is required under the rules of the ABPT" However, they have been asked if their answer would change in a situation where the ABPT would fail BUT FOR the deferrals to the associates plan and they have, on several occassions, declined to answer said question. As a result, my group tends to run a separate ABPT excluding the associates to ensure that the arguement is never raised by, oh, say a plaintiff's attorney.

I also agree that actual 410(b) aggregation will, indeed, make the whole she-bang (excellent usage of the technical term BTW) a RAG, regardless of which plan would fail without the aggregation. Once aggregation is chosen, the combined plan is a single plan and is (or is not) top-heavy as a whole

[ak2ary]

As indicated, if the associates plan does not stand on its own and is aggregated with the other plans in order to enable it to satisfy 410(b), then it is not required to be aggregated for TH purposes.

.

It is NOT required to be aggregated?

[jpod]

ak2ary: The word "NOT" is used correctly in Mike Preston's response to my scenario. None of the associates are key (because they are neither owners nor officers), but all or the vast majority of them are HCEs (i.e., no top 20% election). The TH partners' plan helps the associates plan pass 410(b), rather than vice versa.

[ak2ary]

The "plan" for 410(b) purposes is the aggregation of the partners and associates plans. If that "plan" is top heavy, the whole she-bang is top heavy. The employer makes an election to aggrgate the two plans for coverage, once that election is made they rely on each other to meet 410(b)...there is no concept of a separate 410(b) test for the Partner's plan.

You can't take the position that in the first scenario above, where the associates plan clearly allows the partners plan to meet 410(b), that there is no RAG because ABPT aggregation is not 410(B) aggregation and then say well, 410(b) aggregation isn't enough either, you have to look at which plan would pass 410(b) if they were disaggregated to determine the TH RAG but you dont have to actually disaggregate.

I could be wrong, but the current version of 410(b) has been around for nearly 20 years and never have I seen in any gray book , or heard at any conference or in any discussion with any IRS representative that two plans that are aggregated for 410(b) can disaggregate for the TH determination to determine if the key plan is the plan that needs help or is the helper. Once aggregation is chosen, there is no separate 410(b) identity.

Does anyone have any kind of cite for this?

[jpod]

ak2ary: cite = 416(g)(2)(A)(i)(I) and (II). Look at (I) and (II) carefully, then apply these facts: the partners' plan, which obviously has keys but also covers all non-lawyers (who are 99% NHCEs), helps the associates' plan satisfy 410(b), not vice versa; and the associates' plan has NO keys.

[jpod]

I know that common sense does not always carry the day when applying the qualification rules, but just for fun, consider these points.

1. As noted, because the associates are all or virtually all HCEs, they don't need to be in any plan for the partners' plan to pass 410(b). As a matter of fact, in many cases at law firms they have lobbied heavily to be excluded from all employer-funded retirement plan contributions and to receive extra cash compensation in place of that, but that's another story.

2. Therefore, the goal is to find a way to permit those associates who wish to make elective deferrals to do so without triggering an obligation on the part of the law firm to make TH contributions for them.

3. Clearly, if the associates were in the partners' plan, and if the partners' plan remained TH, the law firm would have to make TH contributions for them. The solution is that the associates can have their own, stand-alone 401k plan with no employer contribution, as long as the associates plan isn't needed to allow the partners' plan to satisfy 410(b).

4. If setting up a separate associates plan to permit them to make elective deferrals didn't work, then those associates who wished to make elective deferrals would not be able to do so. In my opinion, that would not make (common) sense.

[ak2ary]

The common solution in this situation is not to setup a plan for the Associates who are all HCE's, instead it is to make a top-paid group election the result of which is to make the vast majority of the Associates NHCE's.

The deferral only associates plan easily passes 410(b) and usually ADP. The effect of the TPG election is to hurt, however, the rate group testing of the Partners plan, which then generally requires the use of the ABT for coverage testing, rate group testing or both.

Your approach IMHO makes a very dangerous assumption that you can read 416 to say that if the partner's plan would pass alone, it does not have to be aggregated with the asociates plan for 416 purposes, even though it is aggregated with that plan for 410(b) purposes. The clearer reading, I believe is that, under 410(b) there is no Partner's Plan and there is no Associates plan there is a single, aggregated PARTNER/ASSOCIATE plan and that plan is top heavy and mins are owed to the non-keys.

Your approach reads the word "enables" in 416 to be less restrictive than "plans that are aggregated for 410(b). I believe that the IRS reads it to be equal to or, perhaps, more restrictive than aggregation.

Unless you have a specific letter on this issue relating to your plans or unless you can point to some IRS statement in a gray book at a conference or some other source, I believe this is very dangerous (and unnecessary) approach.

[KJohnson]

ak2ary--Went back and looked and Tripodi takes the same position as jpod.

[ak2ary]

Interestingly....without a cite

[Mike Preston]

jpod provided you with a cite. It is directly on point. And it does precisely what you say it does: it allows an employer that aggregates two plans together for 410(b) purposes to NOT aggregate them together for 416 purposes IF and ONLY IF the requirements of 416(g)(2)(A)(i)(II) are met. There really is no substitute for a careful reading of that sentence: "each other plan of the employer which enables any plan described in subclause (I) to meet the requirements of section 401(a)(4) or 410."

Note that it isn't "each plan of the employer aggregated under 401(a)(4) or 410". The bar is higher. It not only has to be aggregated, but it must "enable" the other plan "to meet the requirements". If the requirements of 401(a)(4) and 410 are already met without aggregating the plan in question, what else can it mean? By that, I mean, shouldn't the section be worded differently if it means what you think it does?

BTW, I disagree with your design alternative of making the top 20% election and then excluding the associates from all plans. It is much, much better to not make the 20% election if there are a smattering (it takes only a couple, in most cases) associates who would otherwise be NHCE's. This allows most of the associates to be treated as HCE's and, along with their "low benefits" greatly assists the ratios in the plan that provides benefits. This is true whether the associates plan is rich or poor, aggregated or not. If we make the assumption that whatever design we go with for the associates plan has to "work" on its own, then the mere fact that there are more HCE's in the overall ratio test for each rate group makes the non-associates plan much easier to pass.

BTW, I go back and forth on the issue of having a demonstration in file that the ABT is satisfied without aggregating the associates plan. While that is certainly expedient for the client in deflecting this issue at a low level if brought up by an IRS reviewer. The fact is that it is not required because the ABT isn't subject to bifurcation in this way. At least, I can't find it in the ABT regs anywhere.

Are we having fun, yet?

[ak2ary]

ok look...everything I said here...

I was just kidding

thanks mike, jpod et al...

[Mike Preston]

ok look...everything I said here...

I was just kidding

thanks mike, jpod et al...

We aim to please. ;-)

[AndyH]

Terrific DB discussion, BTW. Surprised pax didn't "prune" this to another board.

[Mike Preston]

You could just "prune" it back, no?