Artie M

Agree if (i) the plan document states a plan year match with true-up rather than a per pay period match with true up and (ii) this occurred with all participants. It appears that this happened with all participants, but if instead it involved just a portion of the eligible participants arguably that match would need to be contributed at the same time the participants who received the match received their match. Otherwise, the plan sponsor could be administering the plan differently for one group of participants than another similarly (identically?) situated group of participants. Treating the participants differently could raise fiduciary duty issues.

I think that yes is generally the better answer if, as you state, the spin-off is intended to be a continuation of the portion of the existing plan attributable to those entities, rather than the establishment of a brand-new plan. So, for the first year Form 5500: In Part I For … fiscal plan year beginning “spin off date (whatever that is)” and ending “end of plan year (whatever that is) Part IB mark “first return/report”, Part II 1b likely “001” 1c “original effective date (whatever that is). Nothing in line 4 That coincides with your facts: • 401(k) plan has existed for years. • The two entities participated in that plan for years. • Assets and liabilities attributable to those subsidiaries are being spun off into a separate plan under IRC §414(l). • The intent is for the new plan to be a continuation of the benefits previously maintained for those employees, not a newly established retirement program. I mean the Form 5500 instructions ask for the plan's effective date, not the date the plan first filed a Form 5500 or the date of the spin-off. Legally, speaking, a §414(l) spin-off is generally treated as a continuation of the transferred portion of the original plan. Participants' accrued benefits, vesting service, distribution restrictions, and other plan rights continue uninterrupted. The spin-off itself is a transfer of assets and liabilities, not the creation of new retirement benefits. The legal substance (a continuation through a §414(l) spin-off) should drive the reporting, rather than simply using the date on which the separate trust or separate Form 5500 first comes into existence. That said, make sure the plan documents are drafted consistently. E.g., the AA says: Effective Date: January 1, 2026 while the Form 5500 reports: Original Effective Date: January 1, 2012 (or whatever the date is), that inconsistency could create unnecessary questions. So, make sure ALL plan documents and communications are consistent (AA or individual plan doc, first 5500, new SPD, resos, etc. If an IRS agent or DOL investigator were to ask: "Your Form 5500 says the plan's original effective date was ___ Why is this the first Form 5500 filed on ____?" the answer is straightforward: "Because, this is the first annual return filed for the XYZ 401(k) Plan as a separately maintained plan following a spin-off from the ABC Corporation 401(k) Plan effective January 1, 2026. Prior to the spin-off, the transferred participants and assets were reported as part of the ABC Plan on Forms 5500 filed for plan years 20__ through 2025. The XYZ Plan is just a continuation of the prior ABC Plan"… or something like that. That's a perfectly logical explanation. Perhaps memorialize this statement to put in the administrative file (for those folks who work on this 5 years from now when everyone has forgotten the transaction). No authority for anything said above except maybe the Form 5500 instructions.... so caveat emptor

As reflected in @Peter Gulia posts, whether a plan can or should accept a POA is not at all clear. There is definitely no requirement under ERISA to accept a POA and there is split views on what can be permitted even if the POA is accepted. The conservative view (which I usually adhere to) is that unless the plan's governing documents state that a POA will be accepted, it won't be... and most, if not all, of the plans we deal with do not address them. We do, however, often suggest and include in plans we work provisions that may permit benefits to be paid on behalf of physically or mentally incapacitated participant. Perhaps look to see if your plan has a provision of that type. For example, we recently assisted with payments under a POA, not due to the POA itself, but because that person qualified as a person who could receive that benefit under the plan's incapacity provision. Here is their provision: Even in this case, there is essentially totally in the Committee's discretion to provide benefits. In the instance, we concluded that "it would be reasonable for the Committee to determine" this paragraph fits we insisted on the POA providing quite a bit of documentation including driver's licenses or official ID of participant and agent, electric, sewage, utility bills showing residence address, the POA, signed letter from physician supporting disability claim, any SSA determination, any LTD determination, affidavits from POA, children of POA, release/hold harmless/indemnity from POA and children of POA, affidavits from supervising manager of participant. Also, we requested and received a HIPAA Privacy Authorization because we called the physician to confirm the facts i their letter. Like your case, part of the affiants' and the doctor's statements stated that the participant was in some type or nursing home. Even after the Committee determined that it would make payments under this provision, the recordkeeper was directed that the agent could not change the beneficiary and distributions could be made only an account held solely in the participant's name, an IRA fbo participant, or another qualified plan account in the participant's name. I believe the account was transferred in part to the participant's bank account and the remainder an IRA fbo participant. Our view was that if the POA is valid (it met all of the requirements of the state in which they resided..PA, which was amended sometime in the 2000s to add the requirements of specific direction), the agent would have access to the bank account (which they did). There was a little less fear here since the agent was the spousal primary beneficiary. However, we had the children, the contingent beneficiaries, sign releases in the event the spouse primary beneficiary died before the participant (the participant was 71 and the spouse was 74). Alot of time was spent on this but the participant did have around $4M in the account. Last but not least, a substantial file was built on this process and is being retained by the plan just in case.

I guess I failed to respond to this but it seems like full vesting should just be given. Why wouldn't the employer want to vest them? Seems like their issue would be the ongoing contributions not partial term.

Assuming this person is an employee... why not just leave them an employee and add a 1,000 hour allocation condition prospectively? or excludable employee classification.... but would have to pass 410(b), nondiscriminatory classification rules (wouldn't pass ratio so probably have to look at average benefits testing)? These by pass the contractor classification issue. also assuming no gateway or cross-tested allocation issues (think not but just in case)

Seems that this should be corrected as an operational failure and an amended 1099R issued showing a $0 distribution. I mean the participant remained continuously employed and was in compliance with the loan repayment schedule until the TPA or payroll screwed up the coding. Then, payroll deductions stopped and the loan was incorrectly processed as a default with a QPLO solely because of the TPA's or payroll's error. Sure, the participant has some responsibility in this but they simply missed the opportunit(ies) to correct the error at an earlier date; but, not picking up the error doesn't mean the error didn't occur. Seems that the error was an operational failure in plan administration, not a participant default. Look at Rev. Proc. 2021-30 §6.07(3) which let's you restore the loan to the status it would be in absent the administrative error through reamortization, etc. Since the participant didn't terminate or have any other distributable event (at least not under your facts), issuing Form 1099-R and reporting a QPLO (Code M) was wrong and should be corrected. A Code L wouldn't be appropriate either because there should never have been a distribution at all. The cure period issue concerns me but arguably there really was never a "failure to repay" under 72p legally speaking. The cure period exists because a participant fails to make a scheduled payment. Your facts seem different as the participant didn't stop the payments, the error caused it. The Plan (through the TPA or payroll) unilaterally made the participant's performance impossible. If just looking at the regs, this argument might not float (doesn't matter what reason the payments stopped) but this should be looked at under EPCRS. There go back to 6.07(3). IIRC there is an example in there about not starting payroll payments when a participant enters into a loan, the cure period runs out, and EPCRS still stated to correct by reamortization, make-ups, etc. I think that what happens..... sorry no time to check. If so, you should be able to fix after cure period. Experientially we have no support for this as we've never submitted a VCP or Audit Cap on that, nor have we ever taken part in an audit where SCP was used after the cure period. So,, again just thoughts....

I am assuming TSA = transition services agreement. Either way, first place to start are the transaction documents (including the TSA). The threshold question really boils down to who is the common-law employer during the TSA period? or like @CuseFan suggests, maybe the buyer/newco adopted the seller plan for a transition period. I would first look to the APA, as it usually states, in black-and-white, "Effective as of the Closing Date, Buyer shall offer employment and Seller shall terminate employment." If that language exists, it likely overrides any inferences you may be able to draw from the TSA's payroll provisions. Then you would need a buyer/newco adoption agreement for the seller plan for those employees to participate in the seller plan. If it doesn’t have those types of provisions, then look at TSA or perhaps an Employee Matters Agreement. If the TSA, EMA and/or APA expressly provide that seller remains the employer through a transition period, that could be the foundation for continuing participation in seller's safe harbor 401(k) through the end of the TSA period. Those documents—not the payroll mechanics—are likely to determine the answer. Note that many TSAs (or employee leasing/secondment arrangements) state buyer acquires the business, seller continues to employ workers through the TSA, seller runs payroll, seller issues W-2s, buyer reimburses seller. Employees are "assigned" to buyer operationally but legally remain seller's employees until a later transition date. All those facts would provide evidence that the seller remains common law employer (the "best case" scenario). If seller remains the common-law employer through say December 31, 2026, then seller's plan can continue to cover seller's employees til then at which time transition to buyer's plan at end of TSA period--becomes a normal plan transition. This is by far the cleanest outcome. If not, back to prior discussion. Point of concern in your last post, "the buyer then hires all the employees of seller..." that sounds like common law employer changed—i.e., employees terminated employment with seller on 4/2 and entered into employment with newco on 4/2—then TSA may not overcome that. A TSA can outsource payroll and HR functions, but it generally doesn’t change who the common-law employer is.

What does the transaction doc say? EPCRS = how to correct the mistake, but APA = what the correct result should be. In my view, that should drive the correction analysis. Not sure what should be done but I would work toward putting everyone in the position they should be in if this problem didn't occur. The big question I have is not provided in your facts: does buyer or newco sponsor a 401(k) plan as of 4/2? If yes, were the affected employees eligible for the seller/newco 401k or after a short waiting period? If they should (or perhaps ever could) be participating in buyer/newco plan, I would ask seller to transfer the elective deferrals (and earnings), the matching contributions (and earnings to the buyer/newco plan. (Seems like someone thought deferral elections carry over.) If there is a buyer/newco plan but a transfer is not technically available, the seller should return the contributions and buyer/newco should make corrective contributions to a buyer/newco plan. If newco/buyer doesn't have a plan, the deferrals should not remain in seller's plan... seller's plan has an operational failure. Those funds don't belong in seller's plan. Why should seller profit off this? Also, seems like it cant leave the money in the trust... presumably, that would be inconsistent with plan docs, transaction docs, and the parties' intent. For example, if buyer accidently wired $500K to the seller 401k trust, seller can't say... oh well, it's in the trust now. The trust only exists to hold assets for participants who are entitled to benefits under the plan--these individuals weren't participants. (I think there is a line of authority and trust law that distinguishes between where money is a plan asset and whether the plan has a right to retain it. but i have not looked at those in a long time. there is a subtle distinction here, especially from a fiduciary perspective.) If there is no qualified plan available seems like the deferrals (plus earnings) should be returned to the affected employees and are simply taxable comp to them subject to income and employment taxes. The employer match (plus earnings) should be returned to newco/buyer. So my view the objective should be to restore all the parties to the positions they intended and would have been in absent this screw up. If can't because no buyer/newco plan, unwind the transaction. This is an asset acquisition. Participation normally ends due to termination of employment. Payroll screwed up. (another thought...Was there some type of transition services agreement in place (sometimes the seller might continue providing payroll services so one could easily seem them not changing their systems to reflect these employees were no longer eligible)?) As always, just throwing darts....

Agree with @CuseFan, EPCRS can still apply to terminated plans under a literal reading of the rev. proc. The interesting question is whose money is the excess that should be returned? That would drive the question of whether they need to seek recovery. What did the plan do with forfeitures? presumably the forfeitures were used to pay expenses? have those all been paid? If there was a reallocation would that have to be done now? or would those funds simply "revert" back to the employer? if they go back to employer then its just a business decision whether to seek recovery or not. Sorry for just raising questions but I haven't ever looked at this.

What am i missing? I don't see it that way. For an equity transaction: target remains the same legal employer after closing. becomes a controlled-group member at closing. Post-closing service with that target/now sub generally must be recognized by the buyer ESOP under the related-employer rules. pre-closing service is not automatically imported into the buyer’s ESOP merely because the seller’s 401(k) previously recognized it--this falls under predecessor service rules. I don't think there is a special ESOP rule changing that result (but I am no ESOP expert); my recollection is that ESOP eligibility and vesting follow the rules applicable to qualified plans generally. I know that service must be counted beginning at closing, but just because they are in a new controlled-group relationship I do not believe pre-closing seller service for ESOP eligibility or vesting must be counted. I think 414a (1 or 2?) says that predecessor plan service is only required when the employer maintains the predecessor's plan. To me pre-closing service would be credited only if required by one of the following: buyer’s ESOP document (see definitions of employer, Related Employer, employment commencement date, elapsed-time service, and predecessor service); SPA or other transaction covenant; a continuation, merger, or assumption of the seller’s plan that implicate 414a or protected-benefit rules (didn't see that in the facts). If this is pre-closing and still negotiating terms, seems the buyer should not necessarily have to adopt the seller 401k's historical service determinations wholesale. If the ESOP grants prior service, perhaps an amendment could specify whether it recognizes actual employment with the target under the ESOP’s own service-counting rules, or instead adopts service as credited under the seller plan. Those can produce different results. Of course, I defer to anyone providing contrary authorities as I have not taken the time to look for any.... sorry.

Permitting voluntary payroll deductions only is inexpensive and employee-friendly but it seems to provide little incremental value to the employer. That is, as an employee "benefit" it is relative weak, at least in my view (e.g., it works "better" if the employer has a relatively younger workforce as older employees won't benefit from it) Thus far when consulted by our clients we have been framing this as not simply looking at whether they should adopt a 128 plan (being politically neutral in the nomenclature) but looking at it as whether they should adopt the 128 plan in favor of another tax-favored benefit. We of course go through the legal aspects but if the employer has a fixed annual benefits budget (which they all do), should the question be is better to spend the up-to-$2500 per participant cost on a 128 plan or is it better to spend it on an increased 401k match, expanding dependent care, contributing to HSAs, etc.? That is, some/many/a few? employees would simply prefer a pay bump. or another broadly available benefit.

I think that is the latest date. One other thought but likely impractical, especially if the plan has a lot of loans outstanding. Economically, simply reducing the interest rate on the participant's outstanding loans should get you close to what he wants. An interest rate-only amendment to the promissory note should not be treated as a new loan or refinancing if the same loan remains outstanding and only the rate is reduced--that is, no other changes are made to the loans' provisions. That is, the amendment would not increase principal, extend the maturity date, provide additional proceeds, or replace the existing loans. The repayment schedule for each loan would be recalculated to amortize the existing outstanding balance, with interest at the revised rate, over the remaining original term. This shouldn't be considered a new loan or refi under the Regs--the loan wasn't replaced. There doesn't appear to be any contrary authority. Of course, this is based on the assumption that the interest rate is a reasonable rate of interest for DOL PT relief. The problem-- can this or should this be done for only one participant? That could create worse optics/favoritism/inconsistent administration issues than perhaps the tax issue. Likely wouldn't do this unless the plan adopted a rule to apply for all outstanding loans where the loan rate drops at least X basis points lower than the original rate and the remaining term is at least Y months or the committee applies the modification consistently in some manner. Again, this sounds impractical.... though perhaps technically permissible. (Also, depends on if the RK can do it.) Alas, I am just brainstorming....

@Peter Gulia Not sure on the exploitation question, though anything that can be exploited usually will be. For typical corporate 401ks, we view 2 concurrent loans as a sweet spot (one general-purpose and one residential or simply two total and don't distinguish). That strikes a reasonable balance between flexibility and complexity. Allowing 4 or 5, by contrast, often creates administrative burdens that outweigh the incremental participant benefit. This leads me to a topic we have been broaching with retirement plan committees. If a client brought this to us, we would ask: if a participant has reached the point of needing 4 simultaneous loans, should the Plan administrator/committee ask whether the plan is functioning more like a revolving credit facility than a retirement savings vehicle. Given your fiduciary background we thought you might be interested.... we have started recommending to the retirement committees we work with to periodically evaluate their loan policy as part of their fiduciary oversight. Even though the decision to offer loans (or how many loans to permit) is generally a settlor function, the committee can monitor operational experience. Metrics such as default rates, average number of loans per borrower, deemed distributions, refinancing frequency, and correction activity can help identify whether the program is operating as intended. Those data points can then be shared with the plan sponsor if the sponsor is considering changes to the loan policy. Not many committees do this, but it seems to us to possibly be a best governance practice as plan administration and fiduciary oversight continue to evolve. ERISA does not require plans to offer participant loans, offering loans is a settlor function, however, once offered.... (We note that the litigation risk here is not the same as say as the risk from class action investment litigation but likely more indirect from failures that may occur under the plan.) We believe that there can be many disadvantages of having a multiple loan program than many sponsors/committees appreciate. @Brenda WrenAs to your original question: Applying the Treas. Reg. math to your facts--Highest balance during last year $37,200, Current balance $37,200, statutory limit $50,000. Applying Q&A-20 the outstanding old loans $37,200 + the replacement loan $37,200 = amount tested $74,400 with limit of $50,000 so it seems the potential excess is $24,400. Under a "mechanical" reading of the regulation that seems to me to be the deemed distribution—not because the participant received $24,400 in cash, but because the replacement loan causes the tested amount to exceed the statutory limit. With that said, wondering what folks might think about an "alternative" line of thought using a "technical" reading of the Regs that once utilized because an independent RK was amenable (and able) to "consolidate" a set of similar loans (3 in that case) and separately amortize each loan within the "consolidated" loan. Under this alternative, using your fact, the original loans A, B, C, D, under the Regs, cannot become payable in 2031. However, the Regs also seem to state this cannot be done unless the replacement note can be viewed as containing multiple internal amortization schedules. Last sentence of Q20 of 1.72p states (also look at Examples): This paragraph (a)(2) does not apply to a replacement loan if the terms of the replacement loan would satisfy section 72(p)(2) and this section determined as if the replacement loan consisted of two separate loans, the replaced loan (amortized in substantially level payments over a period ending not later than the last day of the latest permissible term of the replaced loan) and, to the extent the amount of the replacement loan exceeds the amount of the replaced loan, a new loan that is also amortized in substantially level payments over a period ending not later than the last day of the latest permissible term of the replacement loan. Using your facts and a literal reading of the Regs: suppose the new note is drafted where outstanding principal: $37,200, with Component A $8,500, repayment schedule ends 4/30/28; B $4,000 ends 9/30/28; C $4,700 ends 5/15/28/; and D $20,000 ends 4/30/31. This would be structured under one promissory note and payroll would do one deduction (nothing in the regs would require 4 different ACH drafts). Operationally then it would be one loan but analytically Treas Regs seem like it would treat it as 4 loans. Does this fit within that quoted language above? Most recordkeepers won't structure it in this manner because administratively difficult.... they want the old loans to disappear, just one new loan, 5 years to start over, simple programming, and, granted, still potential tax risk. This was a technical legal application with no specific IRS authority. In our case, 2 loans were originally for 3 year terms and extended to 5 year terms (which is permitted under the last clause of the sentence quoted above) and one loan stayed within its original 5-year term with all having lower interest rates. Essentially, in our case, the payroll deductions were applied to loan A first, loan B second, and loan C last so they could be paid in the time frame required. Of course, this is not a recommendation or advice as I know of no specific authority that permits it. Just looking back and wondering if anyone else thinks this works (or could work) or is it an overly technical reading? @fmsincI don;t think it is the case but the size of your font as well as the bolded all caps at the end of your post makes it seem like you are upset with someone... maybe it was just a cut and paste.

Agree that the issue here is whether these people are employees. If they are "employees," as defined for purposes of the Internal Revenue Code, they are in the plan and the plan cannot be amended to make them ineligible for the contributions until the next plan year (assuming they can exclude and pass any required testing). If they are not "employees" under the Code definition, the plan has been providing contributions to ineligible individuals and there has been an operational error for every year in which these folks have participated in the plan. So, if they are not employees and are, in fact, contractors, they would be required to be kicked out mid-year... and, if they have worked for the employer for more than this year, should have corrections for prior year(s) also.

First, I don't think the forfeiture is appropriate. It was a deferral that should not be forfeited. If forfeited, it will either be used to pay employer expenses, employer future contribution, or allocated to other participants. The first two are clearly advantageous to the employer and the latter is likely not what is in the plan. Plus the forfeiture doesn't give the participant what they should have received under the plan (or the employer simply took their money) Second, seems to me reopening the account should not be an issue. That seems like a system issue. If they can't reopen the account, how do they process any corrective allocations after a participant has taken a complete distribution. Seems like they could temporarily reestablish the account, allocate the $35 and earnings and immediately process the distribution. This is what we see most often. Your fact say all the amounts will end up going to the RK since the distribution cost is more than the deposit plus earnings. However, this would not be a participant-initiated distribution, it is a corrective distribution resulting from a Plan's operational failure. Presumably, the participant already paid the distribution fee when they took their full distribution. If I were the participant my question would be: why do I have to pay again for the company's failure? The Plan (if permitted under the plan's terms) or the employer should absorb the distribution/administration cost. If the RK absolutely cannot waive the distribution fee, I would recommend depositing $35 plus earnings, (agreeing with @Bri) paying the $ administrative charge with participant receiving the full corrective amount. Back to the RK, question them as to how they process post-distribution operation corrections, EPCRS corrections QDRO adjustments, late contribution corrections, etc. I would push the RK on the procedure, not the fee. They have to have a method to use in other instances. Last, an additional issue: If this is a participant contribution that was actually withheld from payroll, this is also a late deposit under the DOL rules. It seems they missed elective deferral deposit, need lost earnings, potentially a prohibited transaction, and possible Form 5330 considerations depending on timing and correction. The excise tax likely will be almost nothing but it likely still is subject to the excise tax.

These answers assume the excess deferrals were all made under this one plan... if it was under multiple plans there would be a different answer. But if one employer, 2026 1099R provides Box 1 gross distribution amount (including earnings), Box 2a same amount, Box 7 Code 8, then 2026 W-2 Box 1 wages (don't include the excess deferral plus earnings amount), Box 12 Code D for total amount deferred (which would include the excess amount). See W-2 instructions for example.

@fmsinc Typically a plan will provide that if a beneficiary fails to file a beneficiary designation or the beneficiary designation is determined to be "invalid" or "ineffective," the plan's default rules will decide the beneficiary. The plan would normally would not look to state law to determine the beneficiary (unless the terms of the plan specifically state to do so). Right, an agreement between the brothers would not normally supersede the order of precedence in the plan's default provisions. Also, I guess I cannot say that any ambiguity cannot form the basis of a distribution. As someone above stated. If a properly filed designation names "my brother" and there are two brothers, seems like some or all of it should go to at least one of the brothers... or did the participant accidentally leave off an "s" on his designation. Under this designation, if the plan's default rules say first to surviving spouse, if none, to children, if none, to estate, and there is no spouse but two children. Well neither of the two children are a brother. Here, if a settlement can be reached where the two brothers agree to split the account funds and the two children are okay with that then we would advise that the plan get a release from the brothers and children and, if they execute the releases, pay out 50/50 to the brothers. Here, I think it is possible that the agreement from the brothers could supersede the plan's default provision because it seems the intent is not for the kids to get the money but a brother would. Also, there was substantial compliance as we assumed the designation was properly filed. As far as costs go, not saying that it would happen but it is possible the fees could be paid through the interpleaded funds (though we always advise against interpleading the funds but only interpleading the parties, unless the court requires it, and we always note that fees can be requested but most likely not going to be paid). In ERISA cases, courts generally do not like to award fees from a participant's account but they might if there are truly competing adverse claimants, administrator completely neutral and acted promptly, dispute involves difficult factual or legal issues, and plan expressly allows payment of account specific legal expenses or extraordinary administrative expenses. They are less likely to award fees if the administrator created the ambiguity, they view there was poor administration, fees are substantial relative to interpleaded funds, and the court believes the determination of the beneficiary under the facts should have been a routine plan administrative decision.

I agree @QDROphile that the claims procedures should be used first. As queried above: what's ambiguous? Another fact that has not been provided is whether a claim has been filed. We never want to use an interpleader unless there is no other way to dispose of the claim(s). Interpleader is usually appropriate where multiple parties assert competing rights, or the administrator genuinely cannot determine the beneficiary after a prudent review. Interpleader isn't appropriate where no one has made a claim, the administrator simply has not finished interpreting the plan, or the administrator is avoiding making a fiduciary determination that the plan document clearly requires. In our experience, judges expect the administrators to interpret the plan, review beneficiary forms, review marriage/divorce records, apply the plan terms. Only when the administrator faces a real risk of multiple liability or genuinely irreconcilable claims should interpleader come into play.

Technically, no, using 12-month elapsed time for FT employees and 1 YOS (1000 hours) for PT employees doesn't automatically create a 410(b) coverage failure. But the design establishes separate eligibility conditions for different employee classifications so the matching contributions would need to satisfy 410(b) coverage testing, and the differing eligible rules could create a coverage issue if they disproportionately delay or excludes NHCEs. The IRS would question first if FT and PT are reasonable classifications with object criteria (presumably, yes) but then does the classification create a coverage problem. If full-time employees contain a disproportionate number of HCEs, the design could raise discrimination concerns. The issue is whether the distinction creates a coverage problem. If the PT group disproportionately consists of NHCEs (which is often the case), a larger percentage of NHCEs than HCEs could be excluded during a given year. Then your matching contribution component would have to satisfy §410(b). The coverage test would look at the benefiting employees for the match, not just the entire plan. I mean why is this being done? Is this a way to get around the LTPT employee rule under the SECURE Acts?

Though EPCRS was broadened to include corrections under SIMPLE IRAs and this would technically be an operational failure, the tricky part is that once valid salary reduction contributions have been deposited into a SIMPLE IRA, there is no SIMPLE IRA correction mechanism that allows the employer to simply pull the money back out of the employee's IRS. Unless things have changed, in the past, we have found that IRA custodians normally won't return the money because it is not an excess contribution under the Code (i.e., the contributions didn't exceed the SIMPLE IRA annual limits and were properly deposited). Like @justanotheradmin states, we see this type of mistake usually "corrected" under a practical approach. Here, that would be correct the payroll contribution settings going forward and leave the contributions in the SIMPLE IRA. If the employee is upset because the over-withholding caused cash-flow issues, reimburse the employee through payroll (some might gross this up, others wouldn't) and absorb the cost. Also, there should not be a "net out" going forward. The employee's salary reduction election controls future payrolls so there should not be a reduction in future contributions without the employee's consent. In the event of an audit, document this to include a description of the employee's actual election, the payroll error, the three affected payroll dates (if in one quarter, also in one year), the correction implemented, (i.e., any reimbursement provided to the employee outside the SIMPLE IRA, no net out going forward), and concluding with something like a failure to implement the election generally/technically is viewed as an operational failure, but, given the these facts, this isolated payroll error was corrected administratively outside the SIMPLE IRA. Really, since can't kick the money out and shouldn't reduce future contribution election, what else is there to do?

Agreed. My post was intended to apply only to a 457f arrangement

I meant to make clear that 10 years is not a bright line... the key is payment is made independent of severance or retirement.

@Peter Gulia starts his last post with "if a plan is ERISA-governed". I would suggest drafting the document so that it is not ERISA governed. It seems that you could comply with 457(f) without turning it into an ERISA pension plan. 3(2) pension plan must provide "retirement income" or "results in deferral of income... for periods extending to the termination of covered employment or beyond". Here put in a 5-year retention bonus, succession planning incentive or CEO transition arrangement. It's payable on a date specific and not termination of employment or retirement. We have a NQDC plan that was audited by the DOL just two years ago that is open to all 2000 employees of the client--it is not a pension plan or a top hat plan under ERISA because payments must be made no later than the 10th year after deferral (not til separation or retirement). (Also 10 years was as long as we felt we could push this,) I know this goes the opposite direction from your facts but it still covers 100% of the company's employees like your proposed plan. Our client's isn't a TH plan but it doesn't need to meet the exception because it isn't a pension plan. Along with that we use terminology like agreement (instead of plan), for retention (not retirement), no funding, unsecured promise, etc.... all self-serving.

I agree with the others. The IRS could make use of the SC arguments but aggressive IRS enforcement or application these arguments/rules is highly unlikely at this time. I mean SC was pretty fact specific, the IRS has not issued any guidance even insinuating its adoption or application of SC, and the consequences of adopting this stance would be enormous. If applied aggressively, the IRS would actually have to come up with some of its own standards just to administer its application. In discussions I have heard many practitioners state their view that the 414 regs would have to be rewritten to apply SC rules to regular retirement plans. That said, we have adjusted our due diligence when dealing with PE-backed clients to include a focus on who are the management/GP entitles, affiliated investment vehicles, level of operational control etc. noting a risk of potential future application. A low risk assessment (other than in multiemployer plans, PBGC issues, DOL investigations, and transactional due diligence) however is based on the typical structures being used currently. We all know that there are those out there who work day in and day out looking for an angle. I don't believe that PE groups are thinking about putting all their HCEs in one entity and all their NHCEs in another.... but who knows. If that were to happen, one could easily see the IRS pulling this nary used arrow out of its quiver.

I agree with @Peter Gulia. I am not aware of any general rule under ERISA or the Code that would render an individual ineligible to participate in a multiemployer plan solely because the individual lacked work authorization or provided an incorrect SSN. Rather, the relevant inquiry is whether the individual was a common-law employee performing covered employment for a contributing employer under the applicable collective bargaining agreement and otherwise satisfied the plan's eligibility requirements. Common-law employee? Likely yes. An individual may have violated immigration laws, but that does not automatically mean the individual is not an employee for plan purposes. The Code and ERISA do not condition employee status on lawful immigration status. The IRS and courts have long recognized that unauthorized workers may still be employees for federal tax and employment law purposes. E.g., wages paid to unauthorized workers are generally still wages for employment tax purposes, employers still have withholding obligations, and unauthorized workers may still be common-law employees. Covered employment under CBA? Presumably yes. Contributions required under the CBA? Presumably yes. False SSN? Administrative problem, e.g., W-2 reporting, payroll tax reporting, benefit administration, etc. And perhaps for the multiemployer plan a participation identification issue (wrong SSN = wrong Person). Sometimes multiemployer plans take the position that contributions cannot be credited because the SSN does not match SSA records. If so, the real issue is we cannot properly identify the participant. That is different from the participant was ineligible. Here, the correction may involve obtaining the correct identifying information and remapping contribution history. But those are generally administrative/reporting issues, not necessarily eligibility issues. Undocumented status? Potential immigration issue, but not obviously a plan eligibility issue. If the person was hired, performed services, paid wages, treated as an employee, direction/control etc, then the fact that the SSN later proves incorrect does not retroactively mean the person was never an employee. I mean what specific plan, CBA, participation agreement, trust, etc. provision makes these individuals ineligible? Some multiemployer plans contain eligibility language tied to covered employment, covered employees, bargaining unit status, participation agreements. Is there any language specifically addressing undocumented workers--I would be surprised to see it. Also, this isn’t an issue of first instance. Construction, hospitality, agriculture, and certain manufacturing industries multiemployer funds have had to have dealt with this We don’t usually see it in retirement plans but often there are provisions in welfare plan documents that state that fraud, misrepresentation to the company of material facts can vitiate eligibility. I have not looked at this but that might ve an avenue to look at. So at this point, I would be inclined to view this as primarily a participant identification and benefit administration issue, not an eligibility failure, unless the recordkeeper can point to specific plan language or legal authority that says otherwise. One other practical question: Are these individuals no longer employed and only now being discovered because they applied for benefits? That often reveals what the real issue is.