Peter Gulia

The “reasonable period” does not refer to how long the failure remained undiscovered. Rather, it’s about how promptly the failure is self-corrected “after such failure is identified.” SECURE 2022 § 305 undoes the Internal Revenue Service’s time limit on which failures are eligible (if otherwise eligible) for self-correction. Congress’s statute provides no special definition for the word “inadvertent”. Merriam-Webster says inadvertent means unintentional or inattentive. https://www.merriam-webster.com/dictionary/inadvertent In VirtualTPA’s story, one might imagine the plan’s tax-qualification failure could have resulted from its administrator’s unintentional or inattentive lack of knowledge of the plan’s provisions. (Isn’t that a way many failures happen?) The administrator (the one responsible under ERISA and the tax Code, not the TPA) might not have known the plan compels an involuntary minimum distribution to a participant who was at a relevant time a more-than-5% owner. (I observe nothing about how responsibilities sort out between and among the participant, the administrator, and the third-person service provider.) If the Internal Revenue Service later pursues something under the IRS’s finding that a plan was tax-disqualified and not self-corrected, whoever asserts the failure was self-corrected must persuade a finder of law and fact that the failure was eligible for self-correction. One can imagine at least plausible, and perhaps persuasive, arguments that a failure of a kind VirtualTPA’s story describes was inadvertent. If it was, the passing of a few or many years does not by itself make a failure that otherwise was inadvertent necessarily less so. A plan’s administrator that errs by not knowing the plan’s provision that applies to a participant who was a more-than-5% owner might continue its ignorance for years or decades. Likewise, inattentiveness too sometimes persists over stretches of time. I concede there is a separate problem about whether the plan’s administrator had procedures reasonably designed to cause the administrator to administer the plan correctly. If an organization really wants rules obeyed, one must supplement written procedures with compensating controls designed under an assumption that some or many people won’t read the written procedures they are told to follow, especially if the rules are many or complex (or, worse, both). But I’ve never seen the IRS push such a point. Instead, the IRS treats the procedures condition as met, even if everyone strongly suspects no one read the procedures. We’re not getting the full facts of the story. If we had them, there could be a discussion about whether the failure was inadvertent, not egregious, and otherwise fits conditions for a failure that could be a subject of self-correction. But that a failure happened more than two or three years ago does not by itself make the failure ineligible for self-correction.

If the plan’s administrator might consider anything submitted to the broker-dealer, among many factors one might consider: What do the plan’s governing documents require? What do the plan’s governing documents permit? How much discretion do the plan’s governing documents grant the administrator? Is there a written agreement, one the administrator has rights to enforce, that makes the broker-dealer the administrator’s agent? What records-retention obligations does that agreement mandate? What controls does the broker-dealer impose on the beneficiary-designation form to lower the risk of fakes? Does the broker-dealer reliably date-and-time stamp the writings it receives? What reporting does the broker-dealer provide to the administrator? When the administrator instructs the broker-dealer to furnish the writing it received, will it be the original showing the participant’s ink-on-paper handwriting? Or is it merely a scan of the writing received? For a beneficiary designation that requires the spouse’s consent, will the administrator be able to examine the notary’s seal or stamp? If the plan’s administrator does not itself have enough expertise to evaluate the soundness, legal effect, and prudence of what would be its agreement with the broker-dealer, has the administrator used outside help to evaluate the agreement? Does the broker-dealer have enough financial capacity and casualty and other liability insurance to respond to the inevitable errors and lapses in its performance?

Gilmore, thank you for your sortable spreadsheet!

In 2019, I helped an administrator file twenty old years’ Form 5500 reports. EFAST worked. I found it useful to begin with the oldest year, and proceed in chronological order. Doing so set a preceding year’s ending balance as the next year’s presumed opening balance.

If Congress abolished an individual-account retirement plan’s coverage, nondiscrimination, and top-heavy rules would that solve most of the difficulties about long-term part-time employees?

I like some of the new opportunities. (I especially like removing, almost completely, the employer/administrator from evaluating a participant’s circumstances.) Yet, I recognize many complexities, and imagine there are some I won’t see until there is a problem in play. Yesterday, a client decided to do nothing on all permitted provisions until after it concludes its search for a recordkeeper.

And yet, at least some of the lobbyists were people who say they speak for retirement-services providers.

If an ERISA-governed retirement plan provides participant-directed investment, the 404a-5 rule specifies three kinds of information changes that call for 30 days’ notice “unless the inability to provide such advance notice is due to events that were unforeseeable or circumstances beyond the control of the plan administrator[.]” 29 C.F.R. § 2550.404a-5 https://www.ecfr.gov/current/title-29/subtitle-B/chapter-XXV/subchapter-F/part-2550/section-2550.404a-5#p-2550.404a-5(b) Often, recordkeepers trot out this rule—even when it doesn’t apply. An unadvised plan administrator just falls in with what its recordkeeper says. Further, many service providers’ agreements require notice (often, 60 days) to the service provider before it has an obligation to provide a service regarding an investment-related change. A plan sponsor or consultant with purchasing power can negotiate the notice period and which circumstances invoke it; others are stuck with a rack provision.

MoJo, your recent BenefitsLink explanation that being available as a recordkeeper is a deselection game—because one must build every new feature lest any desired customer or to-be-satisfied consultant want it or even just ask for it—is right on. Do you see the software build you describe (and the people power needed for the inevitable data weaknesses and processing errors) as another factor that could move some (more) insurance, investment-management, and other financial-services businesses to sell off recordkeeping businesses? (I don’t seek any nonpublic information; I ask only about a general trend.) Could we be headed yet a little more toward having a “big five” do a vast majority of the recordkeeping?

If the plan has a tax-qualification failure to correct, does SECURE 2022 § 305 [attached and highlighted below] expand some opportunities for a correction? To get the protection of this new law, one must correct an eligible inadvertent failure promptly after it is discovered. But is there a time limit on failures otherwise eligible to be corrected? This section of the statute became effective on December 29, 2022. self-correction of eligible inadvertent failure.pdf

Thanks. Assume a plan created before SECURE 2022, so auto- provisions are not required (and will not be added).

If an executive lacks a right to a nonelective credit until the employer declares it, such a credit counts against the executive’s deferral limit for her tax year in which the credit was declared.

Which SECURE 2022 changes are in effect now? Assume a § 401(a) plan with a § 401(k) arrangement for elective deferrals (non-Roth and Roth), but no matching or nonelective contribution. Assume the plan has no automatic-contribution arrangement, and no auto-escalation provision. Assume the plan’s year began January 1, 2023. Assume the plan’s sponsor will amend and restate the documents within the remedial-amendment period. Which SECURE 2022 changes now are mandated as a tax-qualification condition? Which SECURE 2022 changes are permitted? Let’s use our BenefitsLink community to see that we have a thorough list. Remember, even if there’s a long delay about plan documents, we operate the plans now.

If an employer doesn’t want an automatic-contribution arrangement, does Internal Revenue Code of 1986 § 414A set up incentives for mergers-and-acquisitions and other deal makers to arrange things so no new plan is created? If so, will the deal makers be smart enough that the Internal Revenue Service will find no new plan is created?

Thank you, all, for the many helpful suggestions. CuseFan, on my list of potential paths is getting an inside researcher at the plan’s administrator to check the probate court’s records for the county in which the decedent was domiciled to see whether anything was filed. MoJo, if the administrator finds a personal representative of the decedent’s estate, we’d likely seek information, but be unlikely to reveal information. AKowalski and EBP, regrettably the plan’s governing documents lack a provision anything like those you mention. AKowalski, I am considering a fact-finding aid that a writing not received with a reasonable mailing time from the date of death is presumed not genuine, unless evidence submitted to the administrator supports its finding that the writing was the participant’s authentic act. The plan grants discretionary authority not only to interpret the plan but also for findings of facts. I like working with the plan’s time limit on claims. An interpleader is possible without waiting for competing claims. The statute grants jurisdiction “if [t]wo or more adverse claimants, of diverse citizenship . . . , are claiming or may claim to be entitled[.]” 28 U.S.C. § 1335 (emphasis added). But there must be at least would-be or could-be takers. At least one must be “adverse” to another. Further, a could-be claimant must be sufficiently identified that the plan’s litigator could get service of process—else, who would be bound by the court’s decision? The plan’s administrator has not yet identified anyone with authority to act for the decedent’s estate. Even if the administrator finds adverse claimants, we worry they might spend little or nothing on discovery. (Worse, a personal representative of the decedent’s estate might lack money to file anything.) A ticked-off Federal judge might order the interpleader petitioner, the plan’s administrator, to develop the facts. In those circumstances, it’s simpler, and less expensive, to pay an at-risk distribution. Seeking a declaratory judgment is available only “[i]n a case of actual controversy” and only if the case is “within [the Federal court’s] jurisdiction[.]” 28 U.S.C. § 2201(a). (The plan’s administrator would not consider proceeding in any State’s court. And if anyone asks a State’s court for a declaratory judgment, the administrator would challenge jurisdiction anywhere but the State in which it is organized and has its principal office.) fmsinc, thank you for describing suspicions about a writing ostensibly signed a few days before the participant’s death. That’s what started my query. All, a discretionary finding is a straightforward path (and likely is so no matter which finding the administrator chooses). A disappointed person is unlikely to find a lawyer to pursue anything. And showing the administrator abused its discretion with an unreasoned (and so capricious) finding is an uphill climb. This administrator will put efforts on both the reasoning and the written record. I really appreciate everyone’s wonderful help. Even if there is no one clear and simple solution (and how would there be with missing and ambiguous facts?), you’ve helped me think this through.

That the remaining account is less than $100,000 hampers how much effort is prudent. The writing received is on the proper form the plan’s administrator, with its recordkeeper, specifies. The form is not retrievable from the plan’s before-login website. But the individual who submitted the form is also a participant under the plan, and so could have retrieved the form after using his login credentials. The ostensibly named beneficiaries seem to be friends of the decedent. None is described as a spouse, quasi-spouse, child, or other relative of the decedent. Her obituary mentions no spouse, child, brother, or sister; it mentions one cousin (who is not named on the form). The form was mailed by the first-named of the ostensible beneficiaries, who also is participant under the plan, and the return address is the same as his plan address of record. The decedent’s estate has not submitted any claim. An interpleader is premature because there is yet no competing claimant. Or if the plan’s administrator pursues a declaratory judgment, the judge would insist on the administrator submitting its finding about what it believes is the correct result. If the plan pays any ostensibly named beneficiary, the plan risks that a later claimant asserts that the form was a forgery and, absent a beneficiary designation, the decedent’s estate was entitled (as the plan’s default beneficiary) to the remaining account. Thank you for thinking with me.

Here’s the situation (with some facts adjusted slightly to protect my client’s and others’ privacy): About four weeks after a 78-year-old participant’s death, the plan’s administrator receives a document the sender presents as the participant’s beneficiary designation. It is dated a few days before the participant’s death. Nothing about the form is witnessed, by a notary or anyone else. But the employer has no record that its former employee ever had a spouse (or any child or other dependent), and the obituary mentions no spouse or former spouse and no child. The employer/administrator worries that the ostensible beneficiary-designation form might not be the participant’s act. Here’s the difficulty: Because the participant retired 16 years ago, the employer discarded records that might have showed its former employee’s handwriting. The retiree’s request, a few years ago, for automated minimum-distribution payments was processed through the plan’s website. The recordkeeper too has nothing that shows the participant’s handwriting. No one now working for the employer knows anything about the retiree beyond what’s in a computer system record from when she retired. (The employer has tens of thousands of employees, and many retirees.) What information would you want to form a discretionary finding about whether the form submitted as the participant’s beneficiary designation likely is the participant’s act? What information might suggest to you that the ostensible beneficiary designation is not genuine?

With a nongovernmental tax-exempt organization’s unfunded plan for select-group executives, there is no contribution; rather, there is a credit to the account the parties use to measure the organization’s unfunded contract obligation to its executive. (That many practitioners describe a credit as a contribution is understandable; even the Treasury department’s rules describe it that way.) A deferral under a § 457(b) plan counts against § 457(b)’s deferral limit for the participant’s tax year “in which the amount of compensation deferred is no longer subject to a substantial risk of forfeiture.” 26 C.F.R. § 1.457-2(b)(1). If an amount is not immediately vested, the amount “must be adjusted to reflect gain or loss allocable to the compensation deferred until the substantial risk of forfeiture lapses.” 26 C.F.R. § 1.457-2(b)(2). Unless a plan provides immediate vesting or a separate vesting time on each year’s forfeitable credits, either rule (or a combination of them) might result in a bunching of what counts against a year’s deferral limit. See 26 C.F.R. § 1.457-4(c)(1)(iv) example 3 (five years’ nonelective credits, adjusted for an investment gain, counted as a deferral for the year in which the amount becomes nonforfeitable). In my experience, this easily might overwhelm the vesting year’s deferral limit. To consider your question about whether a nonelective credit not paid over to a rabbi trust or other measurement investment until 2023 counts for 2022’s limit, one would evaluate whether the participant’s contract right to the deferred compensation became fixed and vested in 2022. What did (or does) the written plan (which 26 C.F.R. § 1.457-3(a) requires) provide? https://www.ecfr.gov/current/title-26/chapter-I/subchapter-A/part-1/subject-group-ECFRf2be51fac065c2d/section-1.457-2 https://www.ecfr.gov/current/title-26/chapter-I/subchapter-A/part-1/subject-group-ECFRf2be51fac065c2d/section-1.457-3 https://www.ecfr.gov/current/title-26/chapter-I/subchapter-A/part-1/subject-group-ECFRf2be51fac065c2d/section-1.457-4#p-1.457-4(c)(1)(iv)

Consider also which laws against age discrimination (if any) apply to this employer.

CuseFan, thanks. The two new early-out distributions, and a hardship distribution, can be self-certifying; and my assumption (which I neglected to state in my originating post) is that the plan or its administrator would make them self-certifying. Might a plan’s sponsor or administrator set a lower charge on some kinds of distributions? For example, imagine the plan’s undifferentiated charge for a distribution is $100. That charge might not attract attention when applied on a $200,000 severance-from-employment distribution. But imagine a participant wants to use an emergency personal expense distribution to meet her need to raise $200. The participant would need to claim $300 to raise $200. (Assume no withholding for income taxes.) The $100 processing charge then is 50% of the net amount the distributee receives. Because this kind of distribution is for $1,000 or less, some processing charges participants tolerate for bigger distributions might seem disproportionate. Others might say that’s what happens when one uses a retirement plan as an any-purpose savings account. I don’t advocate for or against any view. Rather, my queries are about illustrating the difficult choices service providers and plan fiduciaries face, and that there might be many (and differing) interests to consider.

Yes, absent a plan provision otherwise, one might assume the gross-up amount is money wages (and not a part of the fringe benefit). Even if the plan’s administrator will take advice from another practitioner, consider reminding your client that a finding should not treat highly-compensated employees more favorably than similarly situated (if any) non-highly-compensated employees. And beyond Internal Revenue Code §§ 401-414, an employer/administrator might consider whether its finding would be fair regarding similar gross-ups, and perhaps other kinds of gross-ups. Further, the employer might evaluate whether anything about the gross-up or a treatment of it violates the employer’s provisions for, or a desired tax treatment of, the fringe benefit, including as it applies regarding other employees.

I’ve never needed to apply a § 401(a) plan’s provisions on your question (because every gross-up I’ve seen has been for someone with compensation that, no matter how counted, would exceed the § 401(a)(17) limit). But unless a Read-The-Fabulous-Document exercise points in a different direction, I might assume the gross-up amount is money wages (and not a part of the fringe benefit), and then apply the plan’s provisions following that finding. That wouldn’t completely answer your question because a plan might restrict or limit which payments of money wages count in compensation. If the plan’s sponsor prefers excluding this gross-up from this employee’s (or some similarly situated employees’) compensation, consider that the Treasury department’s rule to interpret and implement Internal Revenue Code § 414(s) favors exclusions that lower the compensation of a highly-compensated employee. See, for example, 26 C.F.R. § 1.414(s)-1(c)(5) https://www.ecfr.gov/current/title-26/chapter-I/subchapter-A/part-1/subject-group-ECFR686e4ad80b3ad70/section-1.414(s)-1#p-1.414(s)-1(c)(5), § 1.414(s)-1(d)(3)(ii)(B) https://www.ecfr.gov/current/title-26/chapter-I/subchapter-A/part-1/subject-group-ECFR686e4ad80b3ad70/section-1.414(s)-1#p-1.414(s)-1(d)(3)(ii)(B).

Belgarath, thank you for your helpful thoughts, practical observations, and fact observation. For a big or mid-size plan, about a year is a project timeline for the many operational, legal, and communications steps needed to implement a change. I’m now negotiating some plans’ service agreements for terms that begin on or a little before January 1, 2024—just when the new early-out distributions take effect. That a sponsor or fiduciary should make informed choices is why their counsel is thinking about it now. Is there really much about the new distributions that’s unknown? For each kind, we know: whether it allows a before-severance payment; whether a claim is self-certifying; whether the distribution is treated, for tax reporting and withholding, as not an eligible rollover distribution; whether the distributee has a three-year option to restore the amount to the plan; and other essential points. Thank you for your observations about keeping charges for distributions constant, and not using a charge to influence participants’ behavior. BenefitsLink neighbors, if one does not use a charge to influence participants’ behavior but seeks only to allocate a cost to those who generate the cost, how would you feel about charging more for a particular kind of distribution if the recordkeeper’s cost accounting shows clearly that it costs substantially more to process that kind of distribution? Or, once a service provider has built capacity, do distinctions about costs become blended or at least less clearly attributable?

Carol’s client’s story might lead a designer of a plan’s governing documents to simplify what the plan’s administrator must, may, or must not do when a writing submitted as a participant’s beneficiary designation is submitted after the participant’s death. The US Government’s Thrift Savings Plan rejects such a writing: “To be valid and accepted by the TSP record keeper, a TSP designation of beneficiary must: (1) Be received by the TSP record keeper on or before the date of the participant’s death[.]” 5 C.F.R. § 1651.3(c)(1) https://www.ecfr.gov/current/title-5/chapter-VI/part-1651#p-1651.3(c)(1) (emphasis added). A plan provision like that could make it unnecessary to evaluate whether a writing submitted after its ostensible maker’s death is genuine. What do BenefitsLink neighbors think: Good idea? Bad idea? Or if one wants to allow a near-death beneficiary designation with some time for it to be delivered to the plan’s administrator or recordkeeper, what about a cutoff nn days after the participant’s death? And how many days would you allow?

Some BenefitsLink neighbors have observed that a recordkeeper will incur considerable expenses to tool up for new distributions and other features SECURE 2.0 permits. And some have observed that, even if one knew or estimated that many or most plan sponsors don’t want the newly permitted provisions, the expense to build a capability is almost unavoidable, because there will be some current and prospective service recipients that want a provision (or the opportunity and flexibility to choose it). Further, new kinds of distributions—such as, an emergency personal expense distribution (not to be confused, or cost-accounted for, with a distribution from an emergency savings account) and an eligible distribution to domestic abuse victim—might involve increased complexity, might generate increases in transactions, and so might increase attributable or allocable costs. (Without discussing specific amounts or anything else that could be price-fixing, collusion, or anti-competition:) How should plan fiduciaries and recordkeepers together seek to allocate these expenses? Compared to a fee for processing a normal distribution after severance from employment (or age 59½) and assuming one’s only reasoning is an attempt to allocate a cost to those who generate the cost: Should a fee for processing an emergency personal expense distribution be less than, the same as, or more than the fee for processing a normal distribution? Why? Should a fee for processing a domestic-abuse distribution be less than, the same as, or more than the fee for processing a normal distribution? Why? Should a fee for processing a hardship distribution be less than, the same as, or more than the fee for processing a normal distribution? Why? Should a plan’s sponsor—using its non-fiduciary settlor powers to decide a plan’s provisions, including charges—favor or disfavor some kinds of distributions? Should a sponsor disfavor an emergency personal expense distribution by charging a higher distribution fee? Should a sponsor disfavor a hardship distribution by charging a higher distribution fee?