roy819

Plan sponsor fails ADP/ACP testing for the 1/1/2024 to 12/31/2024 plan year and refunds are issued to HCEs on 3/1/2025 (refunds issued by the plan's recordkeeper). On 4/1/2025, the plan sponsor realizes that they mistakenly approved the wrong test and intended to rely on a permissively disaggregated ADP/ACP test (disaggregating otherwise excludable employees). The permissively disaggregated test still failed, but had better results and less refunds to HCEs. Any idea how the plan sponsor goes about this given that refunds have already been issued and cashed? If they intend on relying on the permissively disaggregated results, then are the original refunds deemed impermissible distributions? Do they try to collect the overpayments from the HCEs by following the EPCRS/Secure 2.0 overpayment guidance? For example - assume HCE Jane received an ADP refund on 3/1/2025 in the amount of $1,500 and cashed the check. But, based on the results of the permissively disaggregated test, her ADP refund was only $800. Does the plan sponsor follow the overpayment guidance in terms of dealing with the $700 "overpayment"?

bump

What if all of the deferrals prior to the in-service distribution (including those made in 2023) were pre-tax, but all deferrals made subsequent to the in-service distribution (accounting for his current $5k balance) were Roth. Does that change anything in your opinion?

He's still an employee of the employer. He took an in-service (age 59 1/2) distribution in June of 2024, but then continued to defer into the same 401(k) plan. And now he has built up around $5k into his 401(k) account. It just doesn't seem practical that an excess contribution refund to correct a failed 2023 ADP test is going to consist of deferrals that were made from June of 2024 to current date. But those are currently the only deferrals in his account.

Thanks Lou. So just to be clear ---- do the regulations prohibit the excess contribution refund (for the 2023 ADP test) to consist of deferrals made in 2024? If all of his 2023 deferrals were rolled over to an IRA, I can certainly understand that. But on the other hand, he has enough in his 401(k) account (from 2024 deferrals) to process the excess contribution refund.

1/1/2023 to 12/31/2023 ADP test failed. HCE is due an ADP refund of $3,000. The HCE in question took a full in-service distribution in June of 2024. He rolled it over to an IRA. He continued to defer into the 401(k) plan, and has deferred approximately $5,000 from June of 2024 to current date. When correcting the failed ADP test for plan year 2023, is it permissible to refund the $3,000 based on his current account balance (of the $5k deferred in 2024)? Or does the rollover to the IRA need to designate that $3k was ineligible to be rolled over?

Multiple ER 401(k) with a 1/1 plan year has 8 adopting employers. Employers 1, 2, 3 & 4 are all part of the same controlled group and are tested together in what we'll call plan "A". Employers 5, 6, 7 & 8 are all part of the same controlled group (and all unrelated to employers 1, 2, 3 & 4) and are tested together in what we'll call plan "B". The MEP uses the prior year testing method. The 2022 NHCE ADP average for plan "A" was 3.5% The 2022 NHCE ADP average for plan "B" was 4.1% In February of 2023, ownership changes occur and employer 8 moves from plan "B" to plan "A". So as of 2/1/2023, Employer 8 is no longer related to employers 5, 6 & 7. They are, however, now related to employers 1, 2, 3 & 4. Therefore, 1/1/2023 to 12/31/2023 ADP testing looks like this: Plan "A" consists of employers 1, 2, 3, 4 & 8 Plan "B" consists of employers 5, 6 & 7 How does each plan determine the prior year averages on the 2023 test? Am I correct in saying that each plan only has 1 prior year subgroup, and therefore 3.5% will be used as the prior year average for plan "A", and 4.1% would be used for plan "B"?

Back on 9/13 the IRS released IR-2024-236 providing tax relief for Louisiana due to Hurricane Francine. The "IR" release says nothing about Rev. Proc. 2018-58. In most cases though, the IRS releases a state-specific release at the same time as the "IR" release - and this state-specific release will have an "Affected taxpayers" section that typically covers deadlines under 2018-58. But the IRS has yet to publish a state-specific release for Louisiana. Does anyone know why the IRS hasn't released this and is there any way to confirm whether 5500 and other deadlines related to a qualified retirement plan are extended to 2/3/25 for Louisiana? I tried calling the IRS disaster hotline but the individual I talked to did not understand what I was asking and just directed me to the FEMA page.

DB plan terminated a few years ago. An annuity provider was established in terms of transferring the pension benefits. It was discovered that an individual (who was already in pay status when the plan terminated) was overpaid. In terms of trying to collect those overpayments, do the provisions of SECURE 2.0 apply?

Here's another twist/question: Assume a HCE takes a full distribution from Plan A, and rolls it over to their new employer's plan, which we'll call Plan B. HCE is later due an ADP refund from Plan A. Given that there is no account balance left, Plan A adjusts the 1099-R for the rollover, and they also create a new 1099-R to reflect the ADP refund amount (with the applicable distribution code). The HCE contacts their new employer to inform them that a portion of their rollover was ineligible. Plan B proceeds with distributing this money back to the individual. Question: What if Plan B issues a 1099-R for the distribution of these excess contributions (from Plan A)? The individual is left with duplicate 1099-Rs. Should the HCE have informed their new ER that a 1099-R has already been issued (by Plan A)? Who should really be issuing the 1099-R for the ADP refund? Plan A or Plan B?

An employer sponsors a 401(k) plan for their employees, plan year is 1/1 to 12/31. We'll call this Plan A. They spun-off a certain division of employees into a separate (new) plan on 9/1/2023. We'll call this Plan B. These individuals did NOT incur a termination of employment. When testing Plan A for 1/1/2023 to 12/31/2023, the spin-off individuals will be included, but it will only include their contributions from 1/1 to 8/31, correct? As for the compensation used on the test for Plan A (for the spin-off group), I think it is going to be dependent on the provisions of Plan A, and how it defines compensation for the ADP/ACP test. Would that be correct as well? Depending on the plan's provisions - the compensation used for the spin-off group on the Plan A test could be from 1/1/2023 to 12/31/2023, or it could be from 1/1/2023 to 8/31/2023, if the plan excludes compensation while the participant is not an eligible employee (for Plan A). Thoughts?

Plan document facts: The only pre-retirement death benefit allowed is what is required by law (the minimum spouse's death benefit) It can only be paid as a life annuity to the spouse The plan contains language allowing "election to allow participants or beneficiaries to elect 5-year rule". This language can be found on page 132 of the IRS' Defined Benefit Plan LRM package. The obvious issue here is that the plan does not provide for a death benefit where the entire interest can be distributed by the end of the calendar year which contains the 5th anniversary of the date of death. Is this just a situation where, for this particular qualified plan, this election is not applicable? Or am I missing something?

A single participant retired from a qualified, ERISA DB plan at age 55 with a life annuity with 10 years of payments (120 months) guaranteed. He named his sister as beneficiary. 3 years after the annuity starting date, he got married. I presume the spouse has no right to the guaranteed benefits if the retiree dies before 120 payments have been made, correct? Any guaranteed payments will go to the named beneficiary (the sister). Are there any issues if the retiree wants to change the beneficiary? I presume he could contact the plan sponsor to make this change - but I'm not certain if it would violate a code or regulation. I don't believe it would change the benefit amount - as the calculation of the life annuity with 10 years guaranteed does not factor in the beneficiary's age/sex. Does it create a new annuity starting date? We're assuming the plan document is silent on the matter. It doesn't prohibit a retiree to change the beneficiary after the ASD, but it doesn't specifically allow it either.

Assume a DB plan's NRD is age 65. It provides for early retirement at age 55 with 10 years of service, where the benefit is reduced at 3% per year. Also assumes the plan allows a participant to take a distribution from the plan at any time after termination. They can take a lump sum, or they can take an immediate annuity. If a participant takes an immediate annuity prior to their early retirement age, then they won't get the subsidized reduction. Their benefit is reduced based on the plan's definition of actuarial equivalence. Further assume that the only death benefit provided under the plan is the required QPSA. An active participant dies at age 48 (with 15 years of service). The regulations in §1.401(a)-20 say to calculate the QPSA as of the "earliest retirement age", and A-17 in the regulations defines the earliest retirement age. It says that if the plan provides for voluntary distributions that commence upon termination of employment, then the earliest retirement age is "the earliest age at which the participant could separate form service and receive a distribution". If that's the case, then the QPSA is calculated at age 48, in the example above. It sounds like you would take the member's NRD benefit and reduce it to age 48 based on the plan's definition of actuarial equivalent. If the spouse chooses to defer the benefit - the regulations say that "the plan must make reasonable actuarial adjustments to reflect payment earlier or later than the earliest retirement age". So if the spouse defers the benefit, then it sounds like you take the QPSA (again, calculated at the member's age 48 with a reduction based on the plan's definition of actuarial equivalent) and then increase it actuarially to the date of commencement. Whereas, if you calculate the benefit at the member's age 55 (with a subsidized reduction) and then further reduce it to age 48 based on the definition of actuarial equivalent, you'll get a much larger QPSA benefit because it includes the early retirement subsidy. You can make the argument that age 55 is the earliest "retirement" age - but then again that doesn't appear to be what the regulations state. Any comments on how you've seen this calculated for plans that allow benefits at termination? Seems like the spouse could get a much smaller benefit if you treat age 48 as the earliest retirement age in which case they are not getting the advantage of the subsidized early retirement reduction.

Thanks for the reply & confirmation. 1.401(a)-20 Q&A 37 states that a plan is NOT required to provide the QPSA written explanation if it "does not allow a participant to waive such QPSA or to select a non spouse beneficiary". Since the Relius VS does allow a participant to waive the QPSA, then it appears a plan sponsor using this document (where the QPSA is the only death benefit) would still be required to provide the written explanation. I presume that the plan would need to be modified to remove such ability to waive the QPSA if the plan sponsor did not want to provide the QPSA notices. Thanks again.