Luke Bailey

Agreed.

This has been the leading solution ever since the excise tax was put in the Code in the 80's.

Reminds me of when the iPhone 4 came out and, in the midst of "antennagate," Steve Jobs said, "It's just a phone." It's just a 12-month loan extension. It was problematic when it was ambiguous and inconsistent, but now it's not. I have to think with everything going on any further issues with it are not going to make it on Congress's or IRS's to do list.

Plan disqualification would affect all years, and going forwarded. The IRS could only tax the taxable persons (participants, plan trust, and possibly disallowance of employer deductions) for open years, so in theory the SOL enters into calculation of the maximum payment amount, but if you are disqualified for any open year (and of course you would be), all the vested account balances are income to the participants, so it's such a big number that the SOL is not all that relevant. You mean the "recordkeeper," because the "administrator" is in fact the employer. I don't think this enters into the IRS's thinking at all. It might work for some agents, but others might actually react negatively that plan sponsor is trying to avoid responsibility. And sometimes they will assume you can be indemnified by the recordkeeper if it's their fault. This could be not that big a deal if it is a small account. If it is a significant owner and the RMDs would have been large, exam agent will probably view it as a significant failure. Would depend a lot on the other errors and why the occurred, the detail for the RMD error, and also the size of the plan, the extent to which it benefits HCEs as compared to NHCEs, and the income levels of owners.

Yes. I'm not convinced anyone can make complete sense out of the statute. Do you want to try to explain this to a client, including telling them that the IRS has provided an easy, "safe harbor" way that you are telling them they might not want to follow?

Bird, you raise some really good points. Regarding the federal income tax issues, luckily 2018 is still an open year. I can only speak hypothetically, i.e., treating the facts as provided as a hypothetical, since we don't have "real" facts. But speaking hypothetically, I think there are probably two ways you could view this. In the first, the transfer of the shares would be treated as not having fulfilled the client's payment obligation. The capital gain would be the client's, and the client, after receiving the property and earnings back, would then fulfill it's payment obligation to the sponsor and the sponsor would have 2020 business income in the amount paid. In the second treatment, you would view the plan as the sponsor's agent, so the business income would be on an amended 2018 return of the sponsor, and the gain would be taxable to the plan sponsor. The latter is probably more complicated and perhaps further from the facts. It also raises potential PT issues. A mess to be sure. Completely agree with you regarding the actuarial impact for DB plan, but we of course don't really know the facts regarding that at all.

Scott, although your post contains only a brief summary of the facts, I think the "textbook" answer is that the plan needs to make a VCP application to IRS. Depending on things like the clarity of the plan language, what the SPD said, and the statements the participants have received, the IRS might well allow you to correct by forfeiting the amounts now and reallocating the forfeitures retroactively to the active and inactive participants to whose accounts the amounts should have been allocated over the years if the forfeitures had occurred timely. Note that there is at least a good argument that in this case the forfeitures always exceeded the contribution obligation, which is why the amounts would might need to be reallocated to prior years. If based on facts and circumstances that is an unduly burdensome correction, the IRS might let you shortcut it. Would depend on many factors.

Yeah, but what they mostly said is that they are providing you with an easy peasy safe harbor of (a) suspending through end of 2020, (b) starting repayments again in January, 2021, (c) extending the loan term out for the lesser of a full 12 months or the date 12 months from the date its original 5-year term would have expired, and (d) doing just one new amortization schedule so that all the payments beginning with the January, 2021 resumption date are the same amount. They noted that other approaches could be defended as consistent with the statutory language, but would be more complex and they did not bother describing them, since the safe harbor will be better in almost every instance.

Right. As I've stated before on this topic, I believe there is pretty soild precedent coming out of the October, 1987 crash (when a much larger percentage of plans were pooled) that support the principle of doing special valuations where there has been a dramatic market change. Don't have the time right now to dig up the cases, but I recall their being out there.

Agreed. But then third party will immediately turn over to plan sponsor. But the form is important for reasons you stated, Alonzo Church.

Something from the IRS saying what they think would definitely be in order.

In Notice 2020-50 the IRS took the elegant way out of the confusion created by the, dare I say it, gaps in CARES Act Section 2202(b)'s wording. It provides a safe harbor in the second paragraph of Section 5(B) of the Notice. You (obviously) suspend any payments otherwise required in 2020. Then, you restart in 2021, say with first payroll in 2021, or end of January or end of quarter. Whatever you would normally do. If the loan before suspension would have termed out after December 31, 2020, you simply add a year to the loan term. If it would have termed out earlier (e.g., November 30, 2020), you take it out to the anniversary of when it would have termed out, but for the suspension. In either case, you create a new amortization schedule with a single uniform, level payment amount from the restart date to the end of the new term. In the last paragraph of Section 5(B), the IRS acknowledges that "there may be additional, if more complex, ways to administer Section 2202(b) of the CARES Act," and lets you use those if you want, if consistent with the statute, but who would want?

Not really. It does not seem to address the issue at all, one way or the other. There is in 2020-50 the statement (which is repeated at least a couple of times) that "only a coroanavirus-related distribution that is eligible for tax-free rollover ... is permitted to be recontributed," but Roth conversions are eligible to be rolled over tax-free, and in fact amounts can't be rolled to Roths as conversions unless they could be rolled over tax-free. They just aren't rolled over tax-free.

Appleby, where? I searched the 8915 instructions and it seemed to me that all of the Roth references were regarding treatment of disaster-related Roth distributions, not contributions. The only thing I will grant you is that if, to take the simple example I used earlier, an individual (a) took a $100k CVD, (b) does not contribute it for almost three years to either a regular or Roth IRA, (c) reports $33,333.33 for each of years 2020, 2021, and 2022, and then (d) before the end of the three-year period puts the $100k into a Roth IRA, he or she would not report that as a "contribution" on Form 8915, even though the instructions, seemingly assuming that the "contribution" will be to a pre-tax vehicle, tell you to put "contributions" on a particular line and then subtract that from prior included amounts. But that can't be the IRS's subtle way of saying that you can't roll a DRD (and now CRD) to a Roth, because if that is the result they think is right they really would need to just come out and say it and (somehow) defend it.

I think IAWMP as well. I did not read the OP as closely as I should have. Would of course need more detailed facts about why the error occurred, but sure, if a third party somehow deposited money into the trust, the right answer is probably that the third party should get the money back, with earnings, and give it to the right person (here, employer). Of course, we all agreed I think, that the facts as presented have elements of implausibility, so again would need to know had all relevant facts before reaching conclusion.

Can't (at least without researching) add much to what Borsley said, but usually when you correct under EPCRS, the deposits are treated as if they had been timely made. This sounds like technically it could not be self-corrected, so maybe was covered in the VCP compliance statement, if your employer got one. Otherwise, at best unclear, probably.

Mike, these plans appear to be governed by ERISA and it's been over a year. How can you do that?

Catch22PGM, I think that's right. Under 402(c), the surviving spouse can roll over anything that is an eligible rollover distribution (ERD). Of course, under 402(c)(4)(B), the RMDs that would otherwise have been required for 2019 would not be part of the ERD, but CARES Act 2203(a)(I)(ii) says the 2019 amount that would have had to be distributed in 2020 does not exist (legally) as an RMD.

Good to know you are well, RBG. It is hard to know where to draw the line, but I do think you can do it if you have it in your plan document for anyone who has not yet attained 401(a)(14) date. The policy issues are complicated. I guess while you may have a working spouse who takes a lump sum or loan for an irresponsible purpose, you could also have one who is in an abusive marriage and needs money to get out. Anyway, for now it is what it is and very few employers will want to do anything more than what law requires, which is no spousal consent for DC distributions and loans, unless plan contains QJSA as default.

RBG, you make a good point, but I think most of them say or can be interpreted as saying that spousal consent is required for any distribution before the 401(a)(14) date (latest of age 65/NRA, 10th anniversary of participation, or termination of employment). I completely agree with austin3515 on the policy argument. Sure, there are lots of inconsistencies in policy, and requiring spousal consent where not required by law is not going to help or be necessary in many cases, but if someone can't get their spouse's consent for a loan or distribution, they probably should not get it.

So RatherBeGolfing, you are positing that if a 401(k) plan document with only a lump sum distribution said that spousal consent was required for a distribution or loan, that might violate ERISA? Now that you mention it, I can see the argument, and I can even see someone arguing that it violates IRC sec. 401(a)(13)'s prohibition on alienation. It surprises me that there is no guidance on it, at least that I know of. Maybe someone knows of some and will chime in. I'm pretty sure there are plans out there that do require spousal consent even though not required by the Code or ERISA and they have received DL's. My guess is that it is within the employer's authority as settlor of the trust/plan to determine to include such a provision, e.g. in theory a 401(k) plan could make a terminated employee wait until normal retirement age to receive a distribution, and that would be OK (although of course no one does that).

Thank you both, and that makes sense to me, but is there any guidance that nails it down? In theory, the fee should be viewed for tax purposes as paid by the tax cognizable person receiving the service. So it seems like the theoretical issue is whether the service is performed for the plan (and then allocable to the participant's account) or for the participant as recipient of the distribution. Is there any guidance that nails this down that you are aware of?

Larry, the 1099-R (assuming no other distributions to participant in 2020) will show distribution of $100k, right? Does the participant in your view get $100k check and the $50 fee is taken from account post-distribution, or does participant get a check for $99,950?

This is not that complicated. P either meets the CRD requirements, which at this point (without IRS "gloss" by way of a Notice or other guidance) are about as near to computer code as you can get (P diagnosed w/Covid or Dep diagnosed with Covid or P has [loss of job or hours] because of Covid) as you can get. And that little bit of logic is between P and his or her tax return. The employer's role is simply "If P certifies and ER has no contrary knowledge, then OK." In this case, depending on the employer's and plan's governance, P and ER may be one and same, so the retirement adviser needs to make clear what the requirements are and then stand back.

Agree with MWeddell. For there to be a problem the plan would have to say that the match cap is calculated based on comp up to a separate portion (e.g, 1/24th) of the comp limit each pay period, which is a rare to nonexistent plan provision, I think. If your plan says you match, e.g., 100% of deferral up to 6% of comp, and (a) the "Compensation" or "Eligible Compensation" definition (as referenced in the match formula) is just the usual "W-2 with adjustments, limited by 401(a)(17))," or the like, and(b) the participant's deferrals are adequately spread out to capture the maximum match, then the fact that the participant may reach the 401(a)(17) limit before hitting either the 402(g) limit or the maximum match isn't going to stop them from getting the maximum match later, even without a true-up, because the "compensation" plugged into the formula is in effect annual comp.