Peter Gulia

The Treasury department’s notice of proposed rulemaking about qualified tips is scheduled to be published Monday, September 22. The prepublication text is available at https://public-inspection.federalregister.gov/2025-18278.pdf. For my law practice, this is only an academic curiosity. Does anything in this proposed rule change how a retirement plan’s administrator counts an employee’s compensation?

For 2026, Internal Revenue Code § 414(v)(7) burdens only a participant who: had, from the employer that maintains the plan, 2025 Social Security wages (not counting self-employment income as a partner or member) more than $150,000 [$145,000 inflation-adjusted]; and is (or by the end of 2026 will be) 50 or older; and might need an age-based catch-up to support her deferrals. For your typical client plan, how many people is this?

David D, thank you, this is what I'm asking about. Does the IRS require that the (2) information to participants be a separate notice? Or is it enough to put it in the summary plan description?

Lois Baker, thank you for useful information on the “buried” difficulties I’m working on about other plan provisions and communications. BenefitsLink neighbors, any help on my question about whether the way one discloses information about a discretionary contribution affects reliance on an IRS opinion letter?

In a package of documents accompanying an adoption agreement to use a set of IRS-preapproved documents, a service provider furnished a “Discretionary Matching Contribution Notice” with this description: “This form describes the formula used if any discretionary matching contributions are made to the plan. This notice must be provided to each participant who received a discretionary matching contribution no later than 60 days following the date the last contribution is made to the plan for the plan year.” The plan’s sponsor/administrator does not use the service provider’s assembled summary plan description. Also, it does not use a summary of material modifications. Instead, we write and deliver an updated summary plan description before each year, and more often than yearly if there is a change. Rather than a distinct “notice”, the plan’s sponsor/administrator would prefer to include the content about discretionary matching contributions in the SPD (and omit anything separate). Does anything about reliance on the IRS’s opinion letter preclude delivering the information that way? Does anything about in a basic plan document preclude delivering the information that way? Is there another reason it would be unwise to deliver the information that way?

The tax-law condition that a higher-wage participant’s age-based catch-up deferrals must be Roth contributions (or not made) begins with 2026.

I’ve worked on reconstructing (at once) more than 20 years’ plan reports and financial statements. If one decides to use the delinquent-filer program, assemble completely all years’ Form 5500 reports before submitting anything. Test them for EBSA’s edit checks. Test each report for internal logical consistency. Test all the reports for year-to-year and other logical consistency. Consider also that the business organization and its owner (and we guess those are the plan’s administrator and trustee) might want the evidence-law privilege for confidential lawyer-client communications. For work not done by the lawyer or her assistants, consider having the lawyer engage the third-party administrator under a Kovel arrangement to preserve that confidential-communications privilege. This is not advice to anyone.

While this is off-topic for austin3515’s query, a quick observation: EBSA’s approach to abandoned-plan situations is to suggest almost anything that persuades someone to cause payments to participants and beneficiaries. In my experience, that includes ignoring plan or trust provisions that might slow down a path to that desired end. And in a way your and my experiences might suggest something about how austin3515’s client evaluates its choices. If there is a course of action that flunks a tax-qualification condition but is unlikely to be examined by the Internal Revenue Service, a decision-maker might find it’s worth taking a risk that the plan later is found to have been tax-disqualified if doing so speeds the plan’s termination and final distribution. While it’s fact-sensitive, there might be circumstances in which an imperfect allocation from forfeitures could be a reasoned choice. This is not advice to anyone.

BenefitsLink neighbors, we’re imagining a situation we don’t really know the full facts and background of. Sometimes, our bits of reasoning might help an inquirer think through a problem and a range of potential ways of thinking about it. But it's not advice.

For those of us who keep the CCH Pension Plan Guide hardbound volumes of before-1986 rulings, Revenue Ruling 81-10 is in the Pre-1986 IRS Revenue Rulings volume at ¶ 19,554. The ruling’s assumed facts describe an ongoing plan, not a discontinued plan. The ruling does not mention § 415’s annual-additions limit. Why does someone seek to allocate forfeitures to participants who lack compensation? If a forfeiture allocation—whether in relation to compensation, or to account balances—were restricted to participants with compensation in the limitation year in which the allocation is annual additions, are there not enough participants (or insufficient § 415 limits) to use the remaining forfeitures balance? This is not advice to anyone.

If a plan followed the IRS’s nonenforcement announcement, 2026 is the first year to apply a provision that a higher-wage participant’s age-based catch-up must be a Roth contribution. Whether a participant is § 414(v)(7)-burdened for 2026 turns on 2025 Social Security wages. Do BenefitsLink neighbors concur in my estimate that § 414(v)(7)’s $145,000 will be inflation-adjusted to $150,000 for 2025 Social Security wages that drive whether a participant is § 414(v)(7)-burdened for 2026?

In addition to the Bakers’ BenefitsLink posting, here in context is today’s Federal Register publication of the rulemaking. https://www.govinfo.gov/content/pkg/FR-2025-09-16/pdf/2025-17865.pdf This states the effective date (not any applicability date) is November 17. (When I counted 60 days from September 16, I didn’t look at a calendar and so didn’t see that November 15 is a Saturday.)

I don’t know whether EBSA’s software is smart enough to process as not an error the situation you describe. Consider attention to some details to help lower the risk of a query. Part 1 line B: “Box for First Return/Report. Check this box if an annual return/report has not been previously filed for this plan. For the purpose of completing this box, the Form 5500-EZ is not considered an annual return/report.” Form 5500-SF Instructions, page 8 left column. Does this suggest the software might be smart enough to see that an opening balance greater than $0 is not necessarily inconsistent with a first Form 5500-SF report because the preceding year might have permitted a Form 5500-EZ report (or none at all)? Part II line 1c: “the date the plan first became effective”. Part II line 5 about the counts of participants. While this might not prevent a query, a count of 1 or 2 might support an explanation that the plan was a one-participant non-ERISA plan for the preceding year. This is not advice to anyone.

If there are forfeitures unused, perhaps austin3515 didn’t bill enough. To find an allocation is good enough, let it be the administrator’s decision, not austin3515’s risk (however slight). Remember, a recordkeeper proclaims that it does not provide tax or other legal advice. And courts have followed those warnings and contract provisions, putting the responsibility and liability on the plan’s administrator. That said, I often favor letting a plan’s administrator (knowingly) accept risks, especially if an expense for advice would be disproportionate to the exposure. This is not advice to anyone.

Even if the suggested interpretation of ERISA’s title I and the Internal Revenue Code might be reasonable (and I don't suggest that it is), what does the plan’s governing document provide?

The statute, Internal Revenue Code § 414(v)(7), applies to contributions in a participant’s tax year that began or begins after December 31, 2023. (The Internal Revenue Service announced a nonenforcement for 2024 and 2025.) The Treasury’s rule “applies [assuming the rule is published on September 16 and becomes effective on November 15] to contributions in taxable years beginning after December 31, 2026.” For most participants that means 2027. “For prior taxable years [including 2026], a reasonable, good faith interpretation standard applies with respect to [I.R.C. §] 414(v)(7).” A Treasury rule might interpret the statute, but is not itself the source of the law.

"The final regulations do not extend or modify the administrative transition period provided under Notice 2023-62."

austin3515, thank you for your helpful observations. I can use them to improve my advice. A fiduciary deciding a default target-year fund for a class of individuals might know nothing about any individual beyond the common fact of the year in which they were born. So, even if a fiduciary considers mortality or longevity, it’s for the class of participants born in a year. And guessing an age at which an imagined person “plan[s] to retire and leave the workforce” might involve a range of plausible assumptions.

Observe that the rule’s applicability dates refer to “taxable years”—a participant’s tax year.

How much of what the plan sponsor seeks might be accomplished by setting as allocation conditions for a nonelective or matching contribution 1,000 hours of service in the year and employed on the last day of the year? (I ask this as an open-ended question; I don't know the workforce Connor describes or why the plan sponsor seeks to restrict reentry.)

Artie M. and Paul I, thank you for helping me. That a person is highly educated or is a knowledge worker (or is both) doesn’t always mean one knows how to select investments. Or, someone who does know might choose, whether rationally or otherwise, not to put one’s time on that activity. Moreover, even if 99% of the participants deliver affirmative investment directions, a plan’s fiduciary might set a default for the 1% who don’t. And should have a reasoning for the default one sets. In my not-so-hypothetical, the fiduciary that selected the menu of designated investment alternatives also is the fiduciary that must decide the default investment. The target-year funds’ investment strategy described above is from a disclosure for the Vanguard Target Retirement Trust funds. Those funds’ disclosures don’t say ‘this fund is for someone who will turn 65 (or some other age Vanguard assumed) in or near the target year.” Rather, Vanguard says: ‘The fund invests according to an asset-allocation strategy designed for investors planning to retire and leave the workforce in or within a few years of 20yy (the target year).’ If an individual affirmatively directs investment, the individual may decide her guess of when she expects to leave the workforce. But for a default-invested participant, the plan’s fiduciary must form its estimate or assumption. If the fiduciary that decides which target-year fund in the set is a participant’s default assumes that the collective trustee’s description of the funds is truthful, I’m imagining it might be prudent to follow that description’s logic. While that way is not the only way to meet 29 C.F.R. § 2550.404c-5, it at least has an explainable logic. And all it asks of the default-deciding fiduciary is to form a reasoned guess about when a typical default-invested participant leaves the workforce.

Paul I, thank you. In my hypo, the fiduciary has selected which manager’s set of target-year funds are included in the plan’s designated investment alternatives. The fiduciary’s remaining decision is about how to sort a default-invested participant into a target year. I’m imagining a fiduciary might find it’s prudent to use a target-year fund as the fund was designed to be used. (My hypo’s invented quotation about a fund’s investment strategy is adapted from a disclosure of the collective investment trusts of the investment manager with the biggest market share for target-year funds.) Many fiduciaries’ defaults seem to aim at age 65 as a presumed normal retirement age. But for many individual-account retirement plans, ERISA § 3(24)’s construct of a normal retirement age has almost no practical effect. BenefitsLink neighbors, even if 65 might approximate a retirement age for the U.S. general population, would you use a different assumption if you know the workforce you’re planning for all are highly educated knowledge workers? If so, what age would you assume for the target?

The fiduciary is considering setting which target year applies to a default-invested participant by assuming a few might plan to leave the workforce or otherwise begin a payout as young as 73 (even if most people work a few years longer). Thus, a participant born in 1997 (now 28) would be defaulted into the 2070 fund. A participant born in 1962 (now 63) would be defaulted into the 2035 fund. Can anyone suggest a reason why the fiduciary should not decide that?

It might be too hard even to think about a question like this. The expense for even a short bit of a professional’s time to consider even a partially reasoned course of action might be horribly disproportionate to the probability-discounted risk exposure. This is not advice to anyone.

Imagine the responsible fiduciary of an individual-account retirement plan with participant-directed investment decides to use a set of target-year funds for the plan’s qualified default investment alternative. Each of those funds describes its investment strategy with this: “The fund invests according to an asset-allocation strategy designed for investors planning to retire and leave the workforce in or within a few years of 20yy (the target year).” How should a fiduciary select the age at which a default-invested participant is assumed to leave the workforce? 60? 62? 65? 67? 70? 73? Assume the fiduciary does not know when the plan’s participants leave the workforce because almost all people who leave the employer go to work for another employer. If the fiduciary knows that the plan’s participants all are knowledge workers, does that suggest anything about what leaving-work age the fiduciary ought to assume? Whatever else a fiduciary might consider, is there some advantage to falling in with a recordkeeper’s norm? Do recordkeepers have a norm? Am I imagining a choice a plan’s fiduciary doesn’t practically have because a recordkeeper will require its customer to use the recordkeeper’s regime for sorting default-invested participants?