C. B. Zeller

Was there a partial plan termination? If so, all of the affected employees would become 100% vested.

I really like the Corrections eSource; it is like a searchable, indexed, and hyperlinked version of EPCRS with examples. I almost always find it much faster to find what I'm looking for in the Corrections eSource, then follow the reference from there into RP 2021-30. Basically, they said that it's their opinion that the corrective contribution takes the place of the missed contribution and wouldn't be counted in 401(a)(4) testing.

"Must" you say? Strong assertion there. Unfortunately EPCRS is silent on the matter. If you have access to ERISApedia, chapter 9.8.5 of the Plan Corrections eSource addresses this question.

No. To satisfy the combo gateway, the aggregate normal allocation rate (ANAR) has to be at least 7.5% for all NHCEs if the ANAR for any HCE is greater than 35%. Aggregate means the sum of the DB and DC allocation rates. You convert the annual accrual in the DB plan into an allocation using the interest rate and mortality selected for testing. For NHCEs getting an accrual in the DB plan, you can also substitute the average DB allocation rate for all NHCEs instead of using each individual's allocation rate; this avoids having different PS allocation rates needed for every employee to satisfy the gateway.

This should be fine. The group of employees who met age 21/1 year of service are all covered by the safe harbor, so no ADP test on that group. The otherwise excludable group has to be ADP tested, but you could count the SHNEC as a QNEC in the test if you wanted. None? Why safe harbor at all then? If they really want to make a 3%, fully-vested contribution, just do it as a discretionary contribution, that way they're not subject to the rules of 2016-16 if they want to change it later.

Don't forget about the ability to use 430(i) assumptions to determine the maximum deduction.....

The notice has to be provided within 45 days. I don't see anything that says the QNEC has to be made within 45 days. Rev Proc 2021-30 Appendix A .05(9)(c) (emphasis added) Your notice can contain language that says "If applicable, a corrective contribution pursuant to IRS Revenue Procedure 2021-30 will be made to your account" or something vague like that.

Then put in a limit of $0.01 instead.

Yes, they would be excluded, unless the plan specifies otherwise. They would be non-excludable for your coverage test if they met the age and service conditions. The plan document should contain a section that describes what happens when an employee moves from an eligible class to an excluded class.

Ok, then you have the option to allow them to participate early by retroactive amendment. Be mindful of the impact on your coverage test.

I agree with Lou that a $0 plan-imposed limit on deferrals for HCEs will accomplish what you want. If we're talking about 2022 then amend the plan now to put the limit in, as long as no HCE has already contributed more than $6,500. If we're talking about 2021, and the owner already contributed $6,500 then you are probably out of luck. One exception might be if none of the NHCEs deferred, because then the ADP maximum for HCEs would be 0%, so the HCE would have to take a refund of his entire contribution, which would then be reclassified as catch-up, which would then not trigger the top heavy minimum. It's ironic that NHCEs contributing less actually yields a better result for the owner in this case. I think there is an argument to be made that you can set up that same result intentionally, even if NHCEs deferred, by shifting all NHCE deferrals to the ACP test. The ACP test will pass both before and after the contributions are shifted - since the HCE ACP would be 0% in both cases - and once they are shifted, the remaining NHCE ADP is 0%, causing all of the HCE's contribution to be reclassified as catch-up. Mind you, this is just my pet argument and I have no idea if it would actually hold up if challenged. Amending the plan to put in a $0 limit for HCEs is a much safer and cleaner approach.

I don't think there is anything that requires the sponsor to make the correction immediately. It might make sense to wait until the end of the year to determine the missed deferral opportunity, so that the participant's actual contributions can be taken into account to determine the MDO and the corresponding QNEC.

Are either of the affected employees HCEs? If so, you do not have the option to let them in early. You would have to refund the contributions with earnings. If you allow the employee in who didn't meet the plan's age and service requirements, it will affect your coverage test. Make sure that will not be a problem if you decide to go that route.

If it's a 401(a)(30) violation, e.g. they allowed the participant to exceed the limit in that plan (or in multiple plans within the same controlled group), then they have to distribute, as the plan never should have accepted the contribution in the first place. In other words, the plan document probably says that a participant will not be allowed to contribute more than the annual limit, so by allowing the participant to contribute in excess of the limit, there is an operational failure and that can be corrected by distributing the excess under EPCRS. If it's only a 402(g) failure, and not a 401(a)(30) failure in any plan, then there is no qualification issue and no need to distribute any of the contribution. In fact, there is no ability to distribute the excess unless the participant has a distributable event, such as age 59½ or termination of employment. There is a tax consequence for the participant but that is not the plan's problem.

402(g) is an individual limit. 401(a)(30) is a plan limit. You can have a 402(g) excess that is not a 401(a)(30) excess when an individual contributes to two or more plans of unrelated employers during the same year. Most plans will provide the ability for a participant to request a refund of their excess contributions in this scenario. The request usually has to be made by a reasonable deadline in advance of April 15.

Noninvolvement and community property are two different things. The bill as it currently sits in Congress would remove the attribution that causes a controlled group to exist solely because of state community property laws (or solely because of the existence of a minor child). Noninvolvement means that the spouses are not involved in each others' businesses, which is broad enough to cover almost anything. If they talk about work over dinner that could violate noninvolvement and cause a controlled group to exist, regardless of whether they live in a community property state. If the bill is signed into law in its current form, some businesses that are currently controlled groups will cease to be controlled groups. In most cases, that will be advantageous for the parties involved. If, for some reason, they want the controlled group to continue to exist, it should be easy enough for the spouses to become involved in each others' businesses in a way that the noninvolvement exception no longer apples, and they would maintain controlled group status. That said, I'm sure there will be some unanticipated edge cases where things are not so simple. It is likely that, if this provision becomes law, the IRS will provide some sort of transition relief to give people to get this sorted out without jeopardizing their plan's qualified status.

In the wife's plan yes, in the husband's plan no. You can treat an employee who terminates with less than 500 hours of service as excludable if they do not benefit, and they do not benefit solely because they terminated with less than 500 hours of service. In the case of the husband's plan, the reason they did not benefit was not solely because they terminated with less than 500 hours; they also did not benefit because their employer did not adopt the plan. Therefore the employee may not be treated as excludable with respect to the husband's plan.

I would avoid designing a plan this way, but I think you could read the regs to support that approach. 1.401(a)(4)-9(b)(2)(ii)(B) On the other hand, paragraph (b)(2)(iv)(A) suggests maybe you can't: Probably safest to use the earlier of the two entry dates for measuring pay as a participant, and using that for all purposes, if you are going to do this.

I'm not aware of any such requirement in the 401(a)(4) regs. What the regs say is that you have to use plan year compensation (if not using average annual compensation), and plan year compensation means 414(s) compensation measured over either the plan year, or over the period of participation occurring during the plan year (or a couple of other options).

Are you using a preapproved document? If so, I would recommend you talk to your document sponsor.

Our document contains this language:

ERISA 105(a). The term "participant" as used here is defined in ERISA 3(7) and includes "employee or former employee." That said, are you sure the plan doesn't have an immediate deemed cash-out provision for this situation?

You will generally get better results when testing using pay as a participant, assuming that the mid-year entrants are (predominantly) NHCEs. This is fine, even if benefits are based on full year compensation, as long as it is not abusive (for example, a participant who enters the plan on December 30 gets an allocation based on full year compensation which is tested on pay as a participant, resulting in an absurdly large allocation rate).

While a plan must define compensation to be used for allocation (DC) or benefit accrual (DB) purposes, in my experience it is unusual to see a plan that explicitly requires (or prohibits) a particular definition of compensation to be used for satisfying amounts testing under 401(a)(4). Does your plan have a provision that limits the choice of testing compensation?

Why is it the employee's responsibility to open an account? The plan has to have a trustee. This is the trustee's duty. DOL rules on this are quite clear - it has to be deposited as soon as administratively feasible. There is a 7-business day safe harbor for small plans.