Instead of a typical last-day condition, may a qualified plan allocate a year’s contribution only among those employed the next February 15?

[Peter Gulia]

I’m wondering whether a § 401(a)-qualified retirement plan may provide an allocation condition that, instead of looking to the last day of each plan year, instead uses a somewhat later date, a little after a year closes.

For example, to share in a nonelective contribution declared for 2027 and allocated in relation to participants’ 2027 compensation, the participant would need to be employed on February 15, 2028.

The idea is the client’s, not mine.

Assumptions:

The employer’s accounting and tax year is the calendar year. The plan’s plan year and limitation year are the calendar year.

The plan uses no safe-harbor regime to meet any coverage, nondiscrimination, or top-heavy condition. The plan never has had difficulty meeting these conditions.

Of those participants who might not share in a nonelective contribution (if one is declared for a year) because the allocation-condition is the next February 15 rather than the last day of the year, that effect cuts against highly-compensated participants. (For the business and workforces involved, the executives are mobile and the nonexecutive workers are rooted.) And for further reasons, the employer is not the least bit worried about nondiscrimination.

The plan sponsor could change from using IRS-preapproved documents to an individually-designed plan.

Questions:

What qualified-plan rules, beyond nondiscrimination, are involved?

What am I missing?

Are there reasons beyond tax law why an employer should not do this?

Thank you for helping me think about this.

[justanotheradmin]

Is this allowed? and if yes, does it have to be written into the plan document? 

offhand I don't know the answer to your first question. I suspect there is something about definitely determinable as of the last day of the plan year, but I am unsure, but I suspect it is allowed, based on my answer to the second question.  

Does it need to be written into the plan document at all?

If the plan uses a "everyone in their own group" method of allocating the employer contribution, and testing passes, why would an allocation condition need to be written into the plan document at all? An employer could decide "everyone who wears blue shoes in November" gets an employer contribution this year, and the plan's TPA, recordkeeper, etc might never know that is the reason why some people got a contribution that year and not others, even if the TPA was the one who performed the testing. When the allocation groups are not a safe harbor classification - full testing has to pass anyways. So use whatever criteria or allocation restrictions the client wants and as long as it passes I would think its okay. 

[Bill Presson]

Agree with justanotheradmin with one caveat: if the plan is top heavy, anyone employed on 12/31 will likely have to get a contribution whether they are employed in February or not. 

[CuseFan]

Also agree with @justanotheradmin and @Bill Presson. Assuming no coverage and nondiscrimination concerns, I would use the everyone in their own group document provision and then utilize that 2/15 methodology for allocation determinations. What you lose in that is the ability to statutorily exclude from testing those who terminate during the plan year with 500 or fewer hours. I do not think you could write the 2/15 of the following year into the document as your allocation entitlement would not be determinable by the plan year end.

I think someone in this forum asked this same question (maybe with different date) within the last year or two, if memory serves.

[WCC]

CuseFan - excellent memory!

[Peter Gulia]

Justanotheradmin, Bill Presson, CuseFan, and WCC, thank you.

I’m grateful for the own-group suggestion. And thanks for the pointers to the earlier discussions.

For the situation I’m thinking about, there are too many participants to use, in the employer’s circumstances, many allocation groups.

I see the issues about when something counts for § 415(c) and for § 404.

A profit-sharing “plan must provide a definite predetermined formula for allocating the contributions made to the plan among the participants[.]” 26 C.F.R. § 1.401-1(b)(1)(ii) https://www.ecfr.gov/current/title-26/part-1/section-1.401-1#p-1.401-1(b)(1)(ii).

But is there a rule or an IRS interpretation that the facts that drive the “definite predetermined formula” must be facts that happened within the plan year?

My scope is only technical advice whether the proposed February 15 allocation condition would tax-disqualify the plan (or be contrary to an ERISA §§ 202-204 command about what a plan must provide). Others, including my client’s inside lawyers, its human-resources officer, and its chief administrative officer, would evaluate whether a proposal is worthwhile to present to my client’s executive, for him to evaluate whether to present the idea to the corporate parent.

Again, thank you for helping me explore my own preliminary thinking.

[StephenD]

Hiya Peter!  Thanks for all you do for this community, I appreciate the way you step up to make sure correct information gets out there and questions get answered.  There are lots of other folks on this board that have also shared their deep experience, and perhaps they can weigh in here:  In all the years I've been using the General Test for Non-discrimination (Cross testing), I've developed a "wait and see" attitude.  I have been surprised many times when the math doesn't work.  One really knows nothing until the tests are performed.  I would advise a Plan Sponsor against jumping to conclusions about the likelihood of passing the tests.  I have also been surprised when a Plan passes testing when at first glance I thought that it never would.  

[Bill Presson]

1 hour ago, Peter Gulia said:

Justanotheradmin, Bill Presson, CuseFan, and WCC, thank you.

I’m grateful for the own-group suggestion. And thanks for the pointers to the earlier discussions.

For the situation I’m thinking about, there are too many participants to use, in the employer’s circumstances, many allocation groups.

I see the issues about when something counts for § 415(c) and for § 404.

A profit-sharing “plan must provide a definite predetermined formula for allocating the contributions made to the plan among the participants[.]” 26 C.F.R. § 1.401-1(b)(1)(ii) https://www.ecfr.gov/current/title-26/part-1/section-1.401-1#p-1.401-1(b)(1)(ii).

But is there a rule or an IRS interpretation that the facts that drive the “definite predetermined formula” must be facts that happened within the plan year?

My scope is only technical advice whether the proposed February 15 allocation condition would tax-disqualify the plan (or be contrary to an ERISA §§ 202-204 command about what a plan must provide). Others, including my client’s inside lawyers, its human-resources officer, and its chief administrative officer, would evaluate whether a proposal is worthwhile to present to my client’s executive, for him to evaluate whether to present the idea to the corporate parent.

Again, thank you for helping me explore my own preliminary thinking.

Peter, just because the document says everyone is in their own group, it’s incredibly rare to actually have to test more than a handful of groups. For example, everyone that is employed on the February date would likely be in the same group for testing. The plan definition just provides flexibility on how they are grouped. 

[Peter Gulia]

StephenD, thank you for your kind words, and thank you for a related point to think about.

This plan sponsor can’t let its plan design “wait and see” on what coverage and nondiscrimination tests reveal.

The plan’s sponsor is only one of many subsidiaries of a common parent. Each subsidiary maintains its own distinct plan, and each has its own distinct contributions and allocations. Each plan has its own recordkeeper. All coverage and nondiscrimination tests for all plans are run independently of all recordkeepers.

[RatherBeGolfing]

22 minutes ago, Bill Presson said:

just because the document says everyone is in their own group, it’s incredibly rare to actually have to test more than a handful of groups.

Said another way, its incredibly rare to test more than a handful of rates.  Everyone in their own group, maximize owner, and 5% to all others is very common (as long as it works for testing of course).

[Peter Gulia]

Bill Presson, thanks. My reasoning for disfavoring each-participant-is-an-allocation-group is unrelated to coverage or nondiscrimination tests.

About this plan, I have no involvement with, and do not provide advice about, coverage and nondiscrimination testing—except the advice to check-in with the parent.

(More than a few years ago, I advised my client not to change anything—even seemingly innocuous features—without its parent’s approval. And whenever they ask me anything about a point that even imaginably could touch coverage or nondiscrimination, I remind them about how one subsidiary’s plan provisions might others’.)

Could the proposed idea work with as few as two allocation groups: those who met the February 15 allocation condition, and those who did not?

(This is regarding just the one plan. The employer’s coverage and nondiscrimination tests would, I guess, consider contributions and all allocations under all plans of the § 414(b)-(c)-(m)-(n)-(o) employer.)

And if using allocation groups is allowed, doesn’t that leave me with the same question about whether it’s proper to have an allocation formula (or define, even if indirectly, an allocation group) that turns on a fact that won’t occur until after the year closes?