Luke Bailey

Treas. reg. 1.402(g)-1(e)(5) states that the allocable income is equal to the "allocable gain or loss." So the answer is $81, i.e. the cash and the 1099-R. Of course, he or she paid FICA on $84, and that is still the correct amount for that. You're not going to issue a W-2c.

I think if you had folks who still had accounts, and so distributed the excess amounts now, with earnings, you would of course report on 1099-R and there's acode you use to show it is a correction amount and not subject to 10% penalty or eligible for rollover. For the folks who have already taken the money, you do not issue a 1099-R CORRECTED, but rather you need to send them letters telling them that the portion of what they previously received that was excess did not qualify for rollover and should be withdrawn from any IRA or plan they rolled to. See Section 6.06(1) of Rev. Proc. 2018-52.

Kevin C, I'm confused. Suppose from prior years participant had $50k in after-tax account, consisting of $40k basis and $10k earnings. In current year, makes a $10k after-tax, then before ACP is run, and after there are an additional $2,000 in earnings on prior $50k and $200 in earnings on new $10k, participant does Roth conversion of whole $62,200. If ACP is not failed, you probably need one 1099-R to report the $10,000 after-tax contribution, in Box 5, and a second 1099-R to show the $62,200 IRR, again in Box 5 of 1099-R. On the 1099-R for the IRR, the $12,200 of income and $50,000 of basis is parsed out in boxes 1 and 2a. Let's suppose ACP is failed, however, and $5,000 of the $10,000 after-tax is distributed as excess aggregate, together with $100 of earnings. I'm thinking that the Roth rollover is reduced to $55k in Box 5 of a corrected 1099-R, and$5,100 is reported on another 1099-R using Code 8 in Box 7, with $5,100 in Box 1 and $100 in 2a. So this (if correct) is really just a complete unwind of the portion of the attempted after-tax that failed ACP, as if had never occurred.

Right. So even though it may be "processed from the Roth," it's just an excess ACP, as if the in-plan Roth rollover/conversion had not occurred. In effect, to back to the original language of BW's question, you are "unwinding" the conversion to extent of ACP excess amount.

I would add to what Larry has said that 414(b) and (c) (by way of 1563), 267 (used for 4975), and 318, all have different attribution rules.

Right. Completely agreed. Was my point. Glad we're in agreement in end, Lou S.

I don't know of any rule of thumb on marketability discount, but 35% seems in the ballpark. You just don't want it to be so large that the IRS could question whether there was a transfer of property to begin with. See Treas. reg. 1.83-3(a)(5). There are nontax issues regarding valuation, I guess, in theory, but if this is papered from the outset as a call option on part of company at time of grant to pay that price, seems bullet-proof. 409A may apply, but I'm not sure of your facts. The provision in 409A regs regarding 83 (1.409A-1(b)(6)) says in (a) that you don't have a deferral merely because an 83 property transfer is not includable at transfer because not vested, but then in (b) says that if you don't have an immediate transfer, but rather a promise to transfer, you have deferred comp. If the idea is that on a liquidity event everyone vests and includes 65% of 100% of the value of the shares (i.e., after taking the 35% marketability discount) in income, but then company buys the shares back over 5 years, then 409A does not apply, but rather you have an installment sale or a call option or put, however it's papered, and the payments will be nontaxable return of basis except for interest amount. How you paper it (installment sale, vs. call/put) may affect whether exec gets short-term or long-term cap gain. If the idea is rather something like, in the event of a liquidity event the stock never gets transferred and instead you pay cash equal to 65% of 20% of value of shares in each of next 5 years, and only report the payments as made on W-2, then you've turned an 83 arrangement into a 409A arrangement, and violated 409A at same time.

Stash026, I thought about doing this once many years ago for a small plan I was drafting. Seemed like a good idea in theory, but when I thought about the fact that the "backup trustee" provision might not be triggered for 20 years, that you would want the consent in advance of this person/entity, and how much can happen in 20 years, I concluded that I understood why you so rarely see such a provision. Actually, I don't know if I've ever seen one.

But Lou S., say the participant is over 59-1/2 and had started his Roth contributions account in the K-plan 6 years ago. If you're treating the $100 in your example as taxable, you're not really treating it as coming out of the Roth, I think. That's my point.

Agree with Larry, but would add that the issue underpinning this is whether the employees have continued opportunity to vest. If the withdrawing employer continues to have plan for these employees as spinoff, there is nothing in the transaction that would cause these employees to have a termination of employment. So they will continue to vest as continue to work, so no PT.

Lou S, maybe we are misunderstanding each other. When in your earlier post you said, "Process from the Roth," I thought you meant treat the corrective distribution as a distribution from the Roth account on the 1099-R. I agree (if that's what you're saying) that you would report it on the 1099-R as a refund of an excess after-tax (contribution not taxable, earnings fully taxable), and ignore the fact that the money had taken a temporary detour through the participant's Roth contributions account. Is that what you're saying?

David, the 1.408A regs are for Roth IRAs. This is a K-plan and the amount was an in-plan Roth "rollover"/conversion, not a real distribution and rollover.

Right, Lou, it is irrevocable, but we would not be letting employee revoke it, but rather would be saying to employee, "Hey, you know that money in your after-tax that you thought you could convert to Roth? You didn't really have it to convert because it failed ACP, so we're backing it and its earnings out of your Roth and putting it back into your after-tax, temporarily, and then we are distributing with same tax effects as if we had not let you convert it and it had failed ACP so was distributed as such."

Off hand I am not aware of guidance on point and have not looked, but suppose the individual already had his/her Roth account in the plan for 5 years and was over 59-1/2. Wouldn't this result in an unwarranted tax benefit? My guess is the IRS would say you have to unwind the Roth conversion to the extent of the after-tax contributions that fail ACP.

I'm pretty sure based on the text that the prohibition in Section 1.401(k)-1(d)(3)(iii)(C) is intended to apply in all cases, i.e., whether using safe harbor or facts and circumstances, but it could be clearer. You could argue that you can exclude anyone you want, for as long as you want, as long as the plan passes coverage, but again, 1.401(k)-1(d)(3)(iii)(C) sounds to me like a requirement that you stop the 6-month holdout. I mean, the first sentence says you can have other conditions, but then it says that after 12/31/2019 you "may not" apply a suspension.

State court matter. Court can probably enter the order without his consent, but more legal fees. He may also be in contempt. But you need lawyer to talk to state court.

IRS has never defined "retired" for this purpose, so facts and circumstances. You say he has not retired, so maybe no RMD. Would need to know more facts and circumstances.

For self-correction of significant errors need to correct by end of second plan year after year of error. First year of error here ended 8/31/2016, so if plan document required contribution by then (e.g., specified that match would be made on each payroll, or by end of plan year), seems too late for self-correction for first year. But if plan arguably would have allowed contributions for year ending 8/31/2016 to be made by, say, 12/31/2016 and still be timely, may be within EPCRS period for self-correction, I think.

Need more information, I think. So Company A never had employees, and still doesn't? Company B had employees, but transferred them all to Company C when Company C was formed? Would need to make sure that A and B are not brother-sisters under 414(b) and/or (c), such that C could be in a controlled group with A and B. Why doesn't C just adopt B's plan and B unadopt?

I think I would side with your original position on this one, Kansas401(k). This is not a situation where a participant who wants access to funds has a planned fire/rehire. It sounds like he or she had a real separation, entitling him or her to distributions. The plan document stops those distributions once he or she "returns to employment." The employee got a job offer to go back to work 3 weeks hence, and accepted. That does not mean he or she has "returned to employment." Just means he or she is very likely to return to employment in 3 weeks, but a lot can happen in 3 weeks.

Mike put his finger on it. The exclusive benefit rule is the issue. I think this is a common arrangement, and probably often the parties don't require that the buyer adopt seller's plan for the "leased" employees, but in the transactions we have advised we always require that because of the risk that the "leased" employees are really common law employers of buyer during the "leasing period" you would have an exclusive benefit violation unless buyer adopts. Where you are selling assets of a division, it's not a problem, because the seller's plan is not going to be terminated, but where it's the whole company, having the buyer adopt seller's plan to get a failsafe on the exclusive benefit issue would mean you then could not terminate. Of course, the buyer could always adopt seller's plan and keep it or merge it into a buyer plan. A companion issue if you want to leave the "leased" employees on the seller's health plan during the transition period, and it's uninsured, is the "MEWA" issue. I think most are comfortable that if it's for a period short enough to fall within the M-1 reporting exception, they just blow it off.

The underpinning in regs for Tom Poje's point is 1.401(m)-1(a)(ii). The second, eoy match is an additional level of match, so the group that gets it would need to be nondiscriminatory under 401(a)(4), so either pass ratio percentage or average benefits. Then ACP would be applied to all of the matching contributions.

I'll be interested in seeing what others have to say, but I think you may need to request a PLR from IRS to know answer to this one. 1.401(a)(9)-2, Q&A-2(a) says the RBD under the retirement alternative is when the employee "retires from employment with the employer maintaining the plan." So the premise is that there is a single employer, i.e., "the" employer. I think you could argue it either way, both technically and on policy grounds.

This is getting more complicated than I anticipated and I think points other than my original question may have been introduced. For the sake of simplicity and trying to bring this to closure, what I am saying is that the Code and regs are clear that you are TH or not for "a year" (say, Y1) based on whether more than 60% of value as of end of prior year (say, Y0) belongs to keys. The regs are clear that you also determine who are your keys for "the year" (i.e., again, for Y1), based on participant attributes during the prior year (i.e., during Y0). The Code is ambiguous on the latter point, but the regs and, as it turns out, legislative history and practicality, trump the Code's ambiguity on that point.

I think you're all right. I was just puzzled by the fact that the practical rule, which is stated clearly in !Q&A T-12 of the regs, is only barely supported by the statutory language, if that. But taking the statutory language at face value (i.e., determine TH based on end of prior year balances, but determine keys based on facts in current plan year) would lead to many impracticalities, basically because determining whether you are top-heavy or not for a year would become a dynamic task to be performed daily during the year, and top-heavy status could also be, arguably, dynamic during the year. It just doesn't work. Good to know that the Conference Report caught the Congressional staff's failure of expression and backfilled for it. If I get time to take a look at the preamble to the regs, there may be some admission there regarding pushing past the bare text of 416(I)(1) to get to a practical result as well. Thanks for everyone's comments.