Luke Bailey

Obviously, it's really bad (I won't get into the details) to backdate documents. So you want to make it clear to the reviewer that nothing like that was attempted. I'm not clear on your facts, but if that is a potential inference, you want to proceed REALLY carefully, and legal counsel may be advisable.

Well, maybe. I mean, I think the IRS's doctrine of perfect adherence to the document could be questioned theoretically, but would be pointless because the IRS (and independent CPAs doing plan audits) will never accept the counter argument, which is that any contract permits deviations, etc., if they are harmless or consented to by the other party. Since plan documents are contracts that are totally within the control of one side of the deal (the employer), I can see why institutionally the IRS and DOL do not want to engage in the argument, other than at the plan audit and EPCRS level, where individual facts and circumstances come into play. The IRS and DOL understandably want as much order, and as little chaos, as possible in the space they have to regulate.

The legal basis is the "definite written program and arrangement" requirement of Treas. reg. 1.401-1(a)(2), which is a pre-ERISA reg that remains in effect. Suffice to say that one could argue that the requirement as stated in the reg (previous sentence) is not completely "definite" itself as to exactly what it requires or means, but the IRS has been very consistent over the decades in saying that it means very close and consistent adherence to a reasonable interpretation of what the document says. In situations such as the one you describe there are sometimes different reasonable constructions of the plan language, and, failing that, sometimes the ability to get a reasonable correction in VCP.

If it's a self-insured plan (which it probably is, based on your terminology), then it presumably is not subject to Missouri insurance law, unless the promoter has in effect built a MEWA, which is its own separate set of issues. Too small to be subject to COBRA, so you may be OK. But this is the superficial legal theory. The actual facts could make things more complicated. Note that going forward the employer may want to look into doing a QSEHRA. Largely avoids the problem of what happens when employee terminates.

The preamble to 409A says that until partnership 409A regs are published, apply the employer-employee and independent contractor provisions in the regs "by analogy." But many aspects of partnerships are not really analogous. If a partner has an arrangement with his/her partnership that will, based on services done in year X, make him or her entitled to a guaranteed payment for year X, and that partner, before the end of year X-1, enters into a binding election to defer the amount that otherwise would be a guaranteed payment in year X, and the deferral is not recognized in his/her capital account, then that should work so that the guaranteed payment would not be on the partner's K-1 for year X. However, there are collateral effects. The other partners will have phantom income in the amount of the guaranteed payment if it is simply added to their K-1's proportionately as increases to their capital accounts. So do they then get tax distributions to hold them harmless from the phantom income? Do you instead just pay it out to the other partners proportionally to their income shares? How you do it needs to be in the partnership agreement. How do you allocate the eventual deduction when the amount is paid, given that partners may come and go before the payment? How do you ensure the chargeback for partners who got the income, but leave before it is paid to the guaranteed payment partner? Very complicated accounting.

Note that in addition to above (all of which I agree with), you need to check whether state tax rules conform to federal, e.g. for SUI.

You need to look at the Section 72 regs. I would expect that if there is a lump sum refund of contributions on death, then the remaining basis is all recovered (i.e., sum of all post-tax contributions, minus amounts previously recovered under Section 72). But I can't quote you the provision. Again, it should be in the Section 72 regs.

At least one of the sides needs to hire a lawyer to dig into facts and put out a reasoned proposal, even if, because of the small amount involved, litigation should be avoided.

The proposed provision violates the $18,500 deferral limit by whatever the match is. The plan design would make the match an elective deferral along with the conventional "elective deferral."

Didn't know there were still annual val plans out there. But if we make that assumption, then given the circumstances of the participant's ownership and share of assets, he/she is likely a fiduciary and perhaps responsible for, and more knowledgeable about, the plan's investments. An interim val is likely in order under the circumstances.

It may depend on whether the plan document gives the power to elect to the employer or the "administrator," although typically the two are the same, i.e., the employer. If it give the power to the employer, then the employer may or may not choose to exercise the discretion, and in reaching its decision is not bound by any fiduciary constraints. If it gives the power to the administrator, as such, the administrator must probably exercise it as a fiduciary and is found to exercise the power in favor of participants, e.g. by retaining maximum amount in plan.

I have used it. It seems pretty comprehensive, at least back to 2006.

I don't think jim241 has yet explained the crucial fact(s) of why the 50% owner's taking a distribution of 44% of the plan's assets will impact the other 56% negatively. Obviously, if the plan is 100% invested in publicly traded stuff, it would not impact the value one iota. We must be dealing with something else, and therein likely lies the tale.

I don't know the answer, but isn't there an argument that for the purpose of determining "safe harbor compensation" (you can follow the trail from 1.401(m)-3(b)(3)(ii) to 1.401(m)-3(b)(3)(i) to 1.401(k)-3(b)(2) to 1.401(k)-6) the plan year (see 1.401(k)-6) means just the 3-month period the plan was in effect? 1.401(m)-3(f), which generally requires a 12-moth plan year but allows an initial year to be as short as 3 months, strongly implies that in such a case the "plan year" is only 3 months long.

The issue comes up for me far more often in M&A transactions, where acquirer will question compliance.

If underfunded, problems, if overfunded face 4980 excise tax, if assets = liabilities, perfection.

card, good point. I was not really thinking about the vesting schedule inside the plan. I think at this point MGOAdmin needs to clarify what the facts are. A lot of good problems were raised, though.

So just to be clear on what's being proposed, there would be a binding legal agreement that would say, if you're still here in 5 years (a) I will start a plan, and (b) I will contribute an amount to it. The plan will have a 5-year vesting schedule. It will not be in effect until the end of the 5-year period and will excluded pre-participation service (and there is no predecessor plan to block the exclusion of pre-participation service). If you are not here 5 years from now, you forfeit everything. Is that it?

Not sure there is a completely clear answer, since there are no final regulations on point, but if you take the position that this is not an ASG, I think the Service would be hard-pressed to show you were wrong. First, 414(m)(2)(A), which was enacted long ago before LLC's were popular (or maybe even existing as a form or organization in most states) refers only to partnerships and corporations, and you're good on both of those, i.e. you are not a partnership (and have not elected to be taxed as a partnership), and if they want to say you're a corporation, you're good there as well because not a professional service corp (the reg is only proposed, of course, but I believe you have reliance as your question implies). You can also argue that the check the box reg (Reg § 301.7701-3(a)) says you're a corporation for all federal tax purposes, and 414(m)(2)(A) is a federal tax rule.

And B is still in existence and is an adopting employer of the merged plan?

I think that's OK too, and I guess the main effect is that for purposes of calculating any match and nonelective, the employer can, if it wants, take into account compensation for the entire plan year (1/1 - 12/31).

Agree with all of the above. Re the issue of SPD not helping the case as to what the employees understood, if the SPD is at least ambiguous and you have substantial other evidence to show that the employees understood the benefit to be what the employer thought it was (e.g., annual accrued benefit statements for a DB plan that reflect the employer's intended pension formula, or employee presentation materials for a 401(k) that accurately describe the employer's matching formula, etc.), you have a good chance, in my experience, of getting a retro amendment in VCP.

Given the way the plan has apparently been drafted, I don't know if there is a good answer, or even a conservative one. If the plan segregated the benefits/accounts accrued under each employer and said that you got benefit 1 on separation from employer 1 and benefit 2 on separation from employer 2, you'd really have 2 plans, and so the conventional definition of separation from service would work, serially and separately for separation from each of the two employers, with respect only the benefit with respect to that employer. But if the plan has a single benefit or account that accrues and/or vests in a blended fashion across service for both employers, you'd probably be better off requiring separation from service from both. Maybe not clearly permissible, as pointed out above in several posts, but certainly better than the alternative of saying that you can terminate with employer 1 and immediately go to work for employer 2 and get a distribution of your existing benefit accrued under both employer 1 and employer 2. That is clearly a nonstarter. XTitan raises tangentially an interesting issue that may or may not have an answer buried in the reg itself. (I took a quick look and didn't see one). Suppose you conscientiously choose the 20 - 50% rule, determine you have a valid reason, write it into your plan, etc., and later the percentage drops to 19%. Do you have to segregate the benefit accrued during the period before the drop to 19% from the benefit accrued after that date, and apply the definition of service recipient differently for the two periods? I think that 1.409A-1(h)(1)(I)'s use of the term, "With respect to a deferral of compensation under the plan," probably implies that it is specific to each deferral, so such a bifurcation would be required.

Hmmmh... If they have both adopted the same plan, there may at least be an argument that you can, or maybe have to, define separation as separation from both. You're going to have to work through the definitions carefully. Under 1.409A-1(c), do you really have one plan, or are their two plans housed in the same document, 1 for each employer? That could depend to some extent on how benefit accruals, vesting, etc. are determined. And once you've made that determination, how does it affect your view of who is/are the service recipient(s) under 1.409A -1(g), which could in turn affect how you view "separation from service" under 1.409A-(h). I don't recall that the regs really contemplated this sort of situation, so there may not be a clear answer, but my guess is that if one concludes there is really one plan (as opposed to two different plans in parallel universes just housed in the same document), then you may be OK. You might want to call the IRS off the record. I think this is a tough one. Answer may not be clear.

Ok, I agree this is much more complicated. I think I was assuming that both companies had adopted the plan, and were saying that you had to separate from both to have a separation from service. Rereading the question, it does not seem that is the case. Calexbraska has not told us what the plan actually says as to what would trigger a distribution , e.g., separation from both, so we're guessing. But If the companies are not even 20% related, and only one has adopted the plan, I agree you can't require separation from service with both companies as the triggering event.The IRS could analyze that as having a good distribution event (i.e., separation from service with the plan sponsor), but the payment date with respect to the distribution event is not fixed, i.e., the payment date is whenever the individual chooses to separate from the tangentially related employer. Maybe the plan says something different, so I am just guessing, obviously, at what the problem might be. If the plan can be interpreted as saying that what would otherwise be a separation from service with company 1 (the plan sponsor) is not a separation if the employee "transfers" (however defined) to company 2, and the import is not that separation from company 2 is a distribution event, but that the separation from service distribution event goes away until the employee returns to service with company 1, and otherwise goes away completely so that distribution can only be made upon the attainment of a date certain, death, etc., maybe that would be OK. I would want to comb through the regs to see if anything there can be interpreted for or against such a provision. Again, without knowing the plan terms we're just guessing. Anyway, for participants who have not yet had a separation, check out Notice 2010-6. It's fairly liberal, Calexbraska, in permitting corrections for problematic plan provisions. You may want to consider using it here.