Bri

I was under the impression that the contributions are deductible (and part of the typical 25% limitation) but the earnings amounts are not.

Well, you can, but you do have to amend the plan up to 9.5 months after the end of the year to provide for the increased benefit.

Not necessarily, as I'm finding out. (Oh we hired someone and didn't realize they became eligible at their anniversary date last year....)

I figured this was a brand new plan, maybe with a 3% or smaller ICR, and then the employees are all more than 20 years from NRA, so their benefits are getting discounted at 3.29% for Dec. 2021. So you end up with a smaller normal cost than the allocation credit, because all those years at 3.29% are going to overcome the 3.00% rate expected to tack on for such a long time.

This also could blow up if the HCEs haven't gotten to 6,500 yet but are 1973 or later births.

Ooh! Is there profit sharing? How about doing a QNEC to some HCEs in order to cause the actually-refunded amounts to be correct after all and see if you still pass 401a4? (I'm not feeling well today. Sorry.)

Why can't they return the excess?

I've heard a couple of webinars in the past month or so indicating this was something that "didn't make it in" to "Secure 2.0".

Yes, unless you have a weirdo plan that defines the testing compensation explicitly as something else. If your plan just says "anything satisfying 414s" then full year pay is certainly compliant along those lines.

I always thought of the solution as you do make them "match eligible" - but if the formula means they get zero, then that's still the case. The real impact now is your denominator for your ACP test has gone up by 3.......which may or may not have a material impact on your results. Come to think of it, if your plan has voluntary after-tax, then wasn't everyone covered under 401(m)? So that even if some people didn't get a match, that wasn't the only way they benefited....

Sure - and some plans might even say you don't need a formal amendment. (Definitely an SMM, though.) The document/trust agreement should have language for how to address the removal of a trustee. I typically see it where either the company or the trustee gives the other party 30 days' notice, with the ability to agree to shrink that 30 day period by mutual agreement. The company could send the old guy such a notice that he'll be officially "out" as trustee.

It's been a while since I used RGF, but it might have to do with creating a "repeating" page 2 (which has a different suffix for its import routine compared to a static one "page 2" only situation)....

Tell the recordkeeper they're wrong and that they will fix it. Returns can be amended. Thursday edit: Sorry, I didn't see the part about it being the TPA's error rather than the recordkeeper's. TPA should do a make-good for the guy until he can get his big tax refund!

That sounds like a nondiscriminatory manner, so it might be okay in the document/trust language.

It's going to depend if the document defines it. Assuming we get to use "any definition satisfying 414(s)" (or words to that effect), then it's Administrator's choice whether to use "only as a participant" versus "full year." That being said, I think if you have different ENTRY dates for the two plans.....well, since the normal coverage/nondiscrim rules say to use the lowest eligibility rule between the plans, I'd probably use the pay starting from the earlier of the dates. But that's because I don't bother to calculate the separate plans' portions to the overall test under their own definition of pay and then add the resulting EBARs for each participant. (Does anyone do it that way, as in.....can you?)

Unfortunately you can't use the otherwise excludable rule on someone who has the standard 1, but not 2, years of service. The contributions basis won't work - her contribution rate was already lower than the owners'. I had thought about if she were to get the same 6, then they're good with a uniform contribution rate, at least. (And as the DB plan is still non-PBGC unless/until she enters, the 6 plays nice with the deduction limit still.) (If the "greater of normal/TH" formula argument allows her "well, it's more than zero" to be nondiscriminatory even though it's less than the owners' amount, then my only problem may indeed be the failing 401k test. Ahh, takeovers.)

Just want to make sure I've thought this through properly: Company has 3 employees. 2 owners employed since 2015 or so, and staff person hired May 2020. Staff person is a decade older than the owners, so cross-testing is not a consideration. [Edit: Dang, I wish I hadn't posted this in the cross-tested section then, I suppose.] Generous CB plan has 2 year eligibility, so it has been owners-only so far. 401(k) plan has 2 year eligibility for company contributions, 1 year for 401(k). For 2021, prior TPA was allocating only 3% to the staff person, as a nonelective top heavy minimum (listed as profit sharing on their report). But 6% each PS to the owners. The nonelective test has me thinking. Even though the staff person is nominally only eligible for the 401(k), the TH requirement is forcing her to get a nonelective contribution, one that is not going to pass when compared to the owners' amounts. (As in, it doesn't matter if staff person hasn't met the normal eligibility for a nonelective contribution, the top heavy requirement made her eligible anyway, and now she's subject to her rate against theirs.) If they give her 6%, I suppose they could be done. She's not eligible for the CB plan at all. And the DC plan would have a uniform allocation rate, so each plan passes coverage/nondiscrimination separately. Or are they indeed okay with just the 3? The allocation to the staff person is the greater of the plan's "normal" formula, or the top heavy formula. (Since her share would legitimately have been zero.) And the plan would pass coverage if she were not considered benefiting, because she's not "normally" part of the coverage test anyway, having not gotten the 2 years of service yet. (Thinking 1.401(a)(4)-2(b)(4)(vi)(D)(3) here) Thanks. (Never mind that they didn't get any deferrals out of the staff person, so I've got other reasons to think about the 3% and how it's not enough of a QNEC if they allocated it that way.) --bri

Check your language on participant benefit elections. If there's any reference to the requirement that the election be made on a form "approved by the Administrator" or words to that effect, that could be something that allows the Plan Administrator to require it. (Perhaps an attorney wants to chime in that requiring spousal consent places an undue burden on a participant with a valid claim for benefits.)

I believe so - the correction is under EPCRS, so the 1099 gets a code E which means taxable in the year of the refund, rather than the year the actual deduction was claimed for the deferral.

It depends on the distribution timing policy. If your plan allows for withdrawals of rollover money at any time, then you'd have a benefit offset. If the rollover money isn't distributable until some later time (age/service/etc.) then you have a deemed distribution instead.

Why wouldn't the AP simply elect what they want from the various available options, rather than trying to force the form of payment through the Order?

If the true-up were going to be offset by a payable, then I'd say there would have had to be 1099s for those amounts issued before they were even deposited. update to Nate's emoji: I just meant that if the claim is that they have a contribution receivable but also that amount as a corresponding payable, in an attempt to "deem them no longer a participant as of 12/31", then that payable should have generated a 1099-R for the same year, so that the participant indeed is deemed to have a zero balance from the offsetting receivable/payable.

Safe harbor amounts aren't late until the 12/31 following the end of the plan year - that's the date they were truly due. (As CBZ noted, that's independent of their deductibility.) Unless we're talking about a payroll-based SH match, those are due by each succeeding quarter-end.

It should be just a matter of your document's definition of an Eligible Employee. It sounds as though you've got a specialized plan document, and those can say they invalidate upon specific common-law employees becoming eligible. But if the document restricts its use only when there's an owner and family members, you might be okay.

I saw something that the old 2020 expiration date for the HATFA disclosure was just updated to 2034 for the ARPA impact. Since 2022 is the required start for the 15-year amort, and my plan didn't have a 2021 stabilized shortfall, I'm thern doing my unstabilized MRC based on 7-year because it's all hypothetical, indeed.