loserson

My reading of that IRS page is that the plan must treat itself as terminated for purposes of vesting, but not that the IRS will treat it as terminated or force it to terminate. The compliance unit identified plans that discontinued contributions and all they did was make them 100% vest, with no mention of forcing them to terminate and without any clear reference to special reporting or a dedicated Form. It said some of them had to amend their 5500s, but with regard to vested participants, not to report as a discontinued plan and not to mark a 5500 as final. So I read that as evidence suggesting the IRS does not, as a practical matter, require termination or even require extra reporting for plans that discontinue contributions. Whereas this page makes it sound more ironclad that the IRS does not treat a plan as fully terminated until the termination date is set, the benefits/liabilities are determined, and the assets are gone. Though for the original question, I think it will probably make sense to just terminate.

Isn't the practical outcome of ceasing employer contributions that after years of neglect you have to 100% vest all participants? Presumably the participant in a solo 401(k) was already fully vested. Do you have to do special reporting when the employer neglects a plan or is it enough that the participants are all fully vested?

The availability of a loan is not a protected benefit. §1.411(d)-4, Q-1(d)

A plan is not fully terminated until all assets are distributed. The IRS requires terminations be formalized, including a date, determinations, and distributions. Termination is not required. You can keep going even if there are no contributions. The plan still has to maintain and monitor the assets, follow notice and reporting rules, and make RMDs when the participant is 70.5. It probably makes sense to terminate. The admin and vendor fees for a plan, even a DC plan, will be higher than the admin fees for a low cost IRA.

I think you are looking for a benefits consultant, actuary or benefits attorney to advise on the things you can do with overfunding, then to estimate the highest value use of the money, then to figure out how much of that use will go to F, then to multiply that by 50%. A consultant might have the most creative ideas, an actuary would have the most grasp of numbers, and an attorney might be more comfortable with such an unusual legal question. If you asked me how to use up the overfunding, after maxing out benefits and buying life insurance, I would say: match 401(k) contributions within the DB, add family members as participants, hire more employees, pay for participant and retiree health coverage, or just sell the company and merge the overfunded DB into the buyer's not-overfunded DB. There might be other ideas, but you want to get your expert, if you get one, to figure out the various values of these and then determine which is most valuable to F. Sale of the company and merger of the plan into another plan is probably the most valuable to F, because a buyer would probably pay close to dollar-for-dollar for those contributions. They would relieve the buyer of the need to make contributions. But the cost of the sale and the cost of the merger might be enough to ruin that plan. And F might not want to sell, so it might be difficult to convince them to use plan merger as the relevant metric if they claim plan merger would never happen. Relatedly, can you just say that the overfunding is functionally a company asset (despite being legally a trust asset) and increase the value of F's company dollar for dollar? Then reallocate the value of the company? The trust assets probably increase the sale price of the company and they allow the company to reward employees without increasing payroll. I know this is too simplistic, but I feel like your best argument is to count the overfunding's nominal value.

This sounds like it is ultimately a divorce negotiation problem. F is refusing to divide a marital asset. It depends in part on whether they are fully divorced, or if they still have some window to negotiate the division of assets. F was willing to 50-50 split for the QDRO, but insists on 100% keeping the overfunding? Is that because the settlement is done and the divorce is final? It sounds like he knows that M has no leverage. But M should weigh the value of the overfunding versus the cost of lawyers. Lots of divorcing spouses will squander two or five or ten times the value of an asset just to deprive their ex of getting the full asset. Is this enough money that it is worth fighting over? Sort of. The court can give M a bigger value of the QDRO to de facto capture some of the value of the overfunding. It depends where in the divorce process M and F are. Nope. A QDRO cannot order a plan to pay a form of benefit that it was not already going to pay. A divorce judge cannot add a new type of benefit to the plan or change funding or distribution practices. QDROs can redirect the beneficiary of existing plan benefits. An ERISA plan has to qualify a DRO and if the order tries to change the form of benefits offered by the plan then the plan can refuse to qualify the DRO. Not a divorce lawyer and no idea on precedent. But my instinct is that you should figure out the most valuable possible use of the overfunding, attribute the appropriate proportion of that amount to F, then treat that as the value of the overfunding to their marital assets. If he insists on getting 100% of the value of the life insurance, then F could sacrifice some of the regular benefit, or some of the non-pension assets. Did they already finalize the division of assets? Did M already file the QDRO? You can go back and re-divide assets but that might be more trouble than it's worth. If the division of assets is not finally decided, then M should try to add the overfunding to the list of assets. At the very least, this would need F's assent, though. So if F is not interested in splitting life insurance, my guess would be F may not be interested in retroactively making M an employee.

The plan I looked at that allowed trust beneficiaries was a DB with a normal form of benefit that was J&S 50. I don't know if they ever actually paid to a trust a J&S benefit, but the design seemed to expect that they would. It has no special stipulations or limitations on paying JS to a trust. So I think it must be permissible at least in some situations.

Why would you even want to do this? I am confused what they think this achieves. Seems too clever by half. Unsure if they are trying to trick somebody (like the IRS or the plan or their lenders) or if they just want to consolidate everything for investment purposes. Consolidation of assets is the only argument I can think of that makes sense. Have you asked them if they can just set up a bank account in their own name to receive the distributions and then they can auto-transfer the pension distributions into their trust? That'd be easy for you and I cannot see what harm it causes them, unless they are technologically unable to set up the auto transfer (in which case they can call the bank or walk into a physical branch). If they are just consolidating all assets into a trust, they can do the transfers themselves without making you write the checks to the trust. If they are doing this for testamentary reasons, they can establish a pour-over will that moves all their assets into the trust upon their deaths. They don't need to pre-fund the trust if this is just testamentary. Did you look at the plan definition of Beneficiary? The first plan document I looked at after seeing your question specifically says a trust can be a beneficiary, both in the definitions and in the beneficiary designation. This plan document got a DL with that provision in it, so the IRS never complained about it. If your plan doc says trust, or seems to clearly exclude trusts, then you have your answer.

uh oh Are you sure they didn't just do a cashout? If it really was just a de minimis benefit, I think the best route might be to calculate her benefit, pay her out in a lump (if permitted by plan), and amend the filings as necessary. And make sure somebody will send a 1099 next January.

Seems like the first to ask would be the payor. And the lawyers for the payor, and whether they have the details of the settlement. Somebody told the payor who to pay, and either the payor or that somebody backwards down the line must be able to identify why they cut you a check. Could also try the SEC, especially if you can find the settlement details in this list of SEC press releases: https://www.sec.gov/news/pressreleases. Might be easier to google "sec settlement press release" and the payor name, and see what happens. The press release will list a responsible agent at the bottom. You might struggle to contact them but at least it gives you some idea of an actual person who might know what's going on. Any chance you have investment records dating back to 2001 so you can see who invested in what?

My first instinct is this is a conditioning problem but maybe if you work through the regs, then it might be okay because the HSA is not an ERISA plan. I feel weird because this was not exactly a "salary" reduction, because this was not salary. It just feels too cute, is my first impression before digging through the reg. Why do they do it this way? Why not just pick between: (1) giving everybody a salary increase equal to the 401(b)/HSA contribution amount and then let them handle their own elections, and (2) giving everybody a 403(b) NEC equal to the same amount. If it's nonprofit, my guess is that everybody is kinda low paid, so maybe not everybody can afford the HSA contribution if the money goes into the 403(b)? Technically it is better for a nonprofit to pay less FICA, so the HSA contributions are better for the employer than the 403(b) because §125 HSA contributions reduce FICA obligations. But the dollar amounts might not be very big unless it's a lot of contributions. Not sure why they want to claim employer 403(b) contributions, is there a 990 consideration? Are they looking to report themselves as more generous employers?

Check the form of benefit in the SPD. Typical DB plans pay a benefit as one of a few types of life annuity. Typical life-annuity benefits are paid out regularly until one or both annuitants die. This is good if you live a long time, because you might get a check every month for 40 years. This is bad if you die early and you might get monthly checks for a few years or maybe even die before you commence benefits. The DB plan will have a death benefit, including a death benefit if you die before commencing the benefit checks start coming. You already found that and it looks like it only applies to spouse, partner, or minor child. So your uncle was in the worst bucket for being a DB plan participant - died before commencing benefits and had nobody to collect after him. DB plans are akin to old age insurance and since there is nobody covered by the uncle's accrued benefits in the plan, there is apparently nothing to pay. If it were a DC plan, then the money would be his and it would either go to his designated beneficiary, or if he had not named one, then to his heirs or legatees. This rule applies to DC plans, but this rule does not ordinarily apply to DB plans, does it? DB plans pay a benefit defined by the plan document. The benefit is usually a life annuity plus some death benefits. If the uncle's death benefits do not go to anybody, and there are no lives covered by a valid life-annuity to collect the uncle's accrued benefits, then what benefits are supposed to be paid by the plan? It's nothing, right?

The 1099-R is the only way the IRS can police the treatment of the distributed money. Simply sending a letter with the correct treatment lets the participant know, but when it comes time to apply it or enforce it, the participant's tax preparer and the IRS would both look to the 1099-R. It is also likely that the participant might not understand the letter or might not keep the letter with tax documents, especially if it does not arrive in ~January. If the 1099 is wrong, then they might all treat it incorrectly.

If he is not entitled to anything, then that sure sounds like a DB pension, assuming he had no spouse and no designated beneficiary. Not sure what to do with "shares" but that could mean different things. Frozen can mean different things. Sometimes it just means closed to new participants. Sometimes it means no new benefits accruals but only for certain participants. You can ask for the plan SPD. You could ask the plan sponsor to write up a quick explanation of why the estate is not entitled to benefits, and tell them it's for your records to show you investigated the pension. My guess would be it's a DB that was only partially frozen and he is entitled to nothing because he had no spouse and either the plan did not allow other beneficiaries to be designated or it did but he declined to designate one.

Yes, but they share the same 402(g) limit. The TSP is offered to federal employees, including uniformed service members. It is subject to the 402(g) limit. In 2019, the 402(g) limit is $19,000 annual elective deferrals. So somebody who is eligible for federal TSP and also eligible for an employer 401(k) plan can contribute to both. But the total elective deferrals for the year, to all plans, cannot exceed $19,000 in 2019. However, the annual additions limit is separate if the employers are unrelated. So if the 401(k) plan sponsor is not the federal government but a separate employer, then they each have their own annual additions limit. Annual additions include elective deferrals, non-Roth after-tax contributions, employer nonelective contributions and other allocations. In 2019, the annual additions limit is $56,000. That means the individual could defer at most $19,000 pre-tax or Roth deferrals in 2019 into one or both plans, but can make non-Roth after-tax contributions up to the annual additions limit - after accounting for employer nonelective contributions and other allocations.

Agree with the consensus. Read the plan document, but if there is no QDRO, then the second spouse had to consent to waive. If she did not, she displaced the designated beneficiaries upon marriage, by operation of law. Another interesting question would be: what would happen if the second spouse had also divorced? Does the prior designation reassert itself? Does the second ex remain the beneficiary as she was during marriage? Assuming the plan is silent, and that the participant never explicitly designated the spouse as beneficiary, and did not make any election naming her, then my inclination is that the earlier designation becomes effective again. In other words, that it was superseded by his second marriage, but if he is no longer married then it is no longer superseded. Has anyone seen guidance on this point? Not sure if there is better guidance here than Treas Reg §1.401(a)-20, Q/A-25(c), which is fairly vague.

I can't recall ever advising a qualified plan that used the sponsor as trustee, but thinking more about this, the decision to use the employer as trustee would be a fiduciary decision. And if the plan sponsor is bad at being trustee and handling plan assets in the way a typical trustee would be able to, wouldn't this potentially be a fiduciary failure? Have not looked specifically at this issue before, but it seems like maybe pushing the fiduciaries to find a professional trustee is a better step than figuring out if they can barricade unbanked participants from taking distributions. At the very least, recommending a better trustee is CYA for you as their counsel. I also think "my employer won't write me a check for my 401(k)" is the kind of thing that (1) individuals are more likely to complain to EBSA about and (2) EBSA is likely to consider worth a look.

I would say sure, just put it on the intranet or website, give it to any participant who asks for it, and in the GHP SPD tell people where to go on the intranet or website to find it. Include URL if possible. It does not technically follow the ERISA/EBSA or IRC/IRS notice requirements, since it's title 45 (HHS/Public Welfare), not 26 (Treasury/IRC) or 29 (Labor/ERISA). But 164.520(c)(3) allows electronic notice to people who consent, which is similar to the rule in DOL Technical Release 2011-03. Someone else might know if HHS has released guidance on consent, but I do not specifically recall anything. The main difference is EBSA and IRS allow you to presume employees received it if the employer gives them regular access to an internet-capable device as part of their work duties - but 164.520 does not explicitly allow you to presume employees consented. But 164.520 is also very vague about consent, whereas the DOL is very detailed about getting consent from non-actives. If you are worried about the disconnect, you can think of ways employees could give more explicit consent. Like including electronic distribution consent in the open enrollment or including explicit HIPAA consent in the regular ERISA-notice electronic distribution consent. My concern would be more whether you can show good faith to reasonably notify them and that they had ample opportunities to access it. If participants get the SPDs and ERISA notices that way, I am less concerned about whether they might need to technically consent to receive the HIPAA notice electronically. I think OCR is more likely to tag you for bad drafting of the documents or failure to distribute them as diligently as the SPD, rather than failure to get sufficient consent for electronic distribution.

Ah, I see. Can you handle by making trustee indemnify the plan for its failure? Or doing an amendment to the trust agreement to make them indemnify? Or switching to another trustee? Edit - Just saw your subsequent comment that sponsor is trustee

I think the IRS position is you cannot adopt a plan after the last day of the first plan year, but I cannot recall if that comes from Regs or somewhere else. Under the new SCP for plan doc failures, you can maybe do SCP under EPCRS. If we assume this is a scrivener's error (i.e. a clerical failure to update the doc date to reflect the 2017 date everybody intended to treat the plan as adopted) then I think SCP is fine. But if you assume that the plan was never adopted because it was for a previous year, then SCP is not available for failure to timely adopt a written plan. I think SCP probably makes sense if nobody treated it as existing before 2017.

I didn't want to get into 105(h) because it's thorny. But the issue having rightly been raised, I briefly looked and it seems like mdm09 is probably right to be concerned that the superintendent would be an HCI and that this structure would impermissibly discriminate in his favor. Treas Reg §1.105-11(c)(3)(iii):

I guess I am interested why a plan would want to do this. If it's a cost thing, can't they just increase their fee for a distribution check?

I get the skepticism of their motives and trustworthiness, but if their questionnaire is revealing what you believe to be bad practices, maybe take the opportunity to revisit plan practices? The IRS has been doing this survey stuff all across the board for all types of taxes and taxpayers. It's why most IRS instructions include the number of hours they estimate the form will take you to complete. This survey is how they guess that and they are asking for your help updating their guesses. "Who has the data?" is directly relevant to how long it takes to complete a form. If they wanted to ask you questions to identify bad practices that might trigger an audit, they would put those questions on the regular reporting form and ask it of everybody. And in any case, I am not sure why the IRS would care about bad recordkeeping practices so much they would audit you - unless they can levy a tax or penalty, or maybe initiate a criminal case, I do not think they especially care. EBSA might care about bad practices, since they see themselves as protecting participants and catching infractions, but they generally cannot levy penalties or taxes. But the IRS is about that money. But if it's voluntary and you do not want to gift them your time to complete their survey data, I guess chuck their letters in the trash.

Having trouble understanding the play here. Why are they doing this? Is this just a cost question? Does the plan not already charge an admin fee for distributions? I mean, maybe just charge enough to cover reasonable costs of a distribution check? I am concerned at the idea that they would be getting "consent" to an IRA by locking away people's distributions unless they agree to it. We might be talking about people unable (or unwilling) to get bank accounts, so their options are get an IRA or else no distribution. I think the consent in that case could be a little hollow. And if your defense is "well they could just get a check from the IRA" then why is the plan making them do this? Why go through the annoying process of a rollover just to get a check? Treas Reg 1.411(a)-11(c)(2) has distribution consent requirements. You might take a look there and think about any implications of deeming folks to have consented to an IRA rollover under various scenarios. Treas Reg 1.401(a)(31)-1, Q&A-9, says that you must allow partial direct payment to distributees alongside partial rollover but does not address direct payment method. If there is no reasonable cause for this rule, then it seems like you are obstructing people from taking distributions, which for sure sounds like fiduciary duty territory. IRAs have gotten really good lately, and a smart investor can get really cheap and good index funds. And people are way better off not getting taxed, and they are way more likely to leave a rollover in a tax-advantaged account. So it's not crazy in principle. But I am concerned that functionally this rule will keep people from their distributions. And if the employer is saying "non-IRA bank account ACH is okay, just no direct checks" then we lose the paternalism argument entirely.

You can amend the plan to allow for the CBA folks and leave the rest to be tested separately, yep. Treas Reg §1.401(m)-1(b)(2): See also Treas Reg §1.401(m)-1(b)(4)(iv) on disaggregation for portions of plans that are collectively bargained.