Luke Bailey

I don't think so, Bri, because for 401(a)(4) you aggregate plans of the employer, including a controlled or affiliated service group employer. You also aggregate, of course, DC plans of an employer or controlled or affiliated service group employer for 415, but Treas. Reg. 1.415(f)-1(f)(1) has a special rule for 403(b) that treats the participant owner of the 403(b) contract as the employer with respect to it, but that is only for 415.

Sorry to hear about your loss, Tax Cowboy.

JackS, it's got its own law. Go here: https://www.law.cornell.edu/uscode/text/5/part-III/subpart-G/chapter-84/subchapter-III Section 8440 of above says you treat like a 401(a) plan.

Nate S, I think you're missing my point, which is that it could work, easily.

Well, sometimes. Broken clock right twice a day, that sort of thing.

Jakyasar, I am unaware of any PBGC guidance that addresses how to treat an LLC for purposes of ERISA section 4021(d), which defines "substantial owner." ERISA section 4021(d) deals only with sole proprietorships, partnerships, and corporations. If the PBGC treated the LLC as a disregarded entity, then the question becomes are you and your spouse, in the example, substantial owners of the S corp, and of course you would be because you own 100% and that ownership would be attributed to your spouse as well. If the PBGC did not treat the LLC as a disregarded entity, but rather as a partnership or corporation, then I think you would get the same result because Company A's ownership of Company B would be attributed to you under the rules of IRC sec. 1563(e), as referenced in ERISA sec. 4021(d). The above makes intuitive sense, right, since in substance you (and by attribution your spouse) own 100% of the company you're working for?

Ananda, since we're dealing with EPCRS, which is a discretionary IRS program, it's how the IRS views it that matters. Did you find any specific elaboration in Rev. Proc. 2021-30 of this issue that would support your view ?

I haven't checked this, acm_acm, but if it comes up in the future I'll check it out. Good to have in mind. Thanks.

My point is it could. Try to use a fake credit card # and see how far that gets you.

Need better description of facts. For example, if IRS position is that plan is qualified, what is the basis for claiming tax owed? The six-year vs. three-year statute is based on whether the amount of the deficiency is more than 25% of gross income. The plan's trust should have a 3-year statute if 5500 filed.

From Section 4.01(b) of Rev. Proc. 2021-30: "A Plan Document Failure consisting of the initial failure to adopt a Qualified Plan, or the failure to adopt a written § 403(b) Plan timely in accordance with §1.403(b)-3(b)(3) and Notice 2009-3, 2009-2 I.R.B. 250, is treated as a Plan Document Failure that is not eligible to be corrected under SCP." So VCP, or in your case, Ananda, Audit CAP.

Riley, I've been asking this for 40 years. You might look up articles on E-Verify and why it's never really been used. A really interesting story.

Mr. Bagwell, there does not seem to be any specific guidance on this issue. The fact that the partner's self-employment income is negative, so they will have no SECA, would certainly seem to indicate that there is no Section 415 compensation, so that the contribution would violate Section 415 and potentially disqualify the plan. On the other hand, perhaps an aggressive argument could be made that not grossing up the partner's SE income hypothetically by the amount of the excluded PPP loan forgiveness violates Section 276(a)(1) of the COVID-related Tax Relief Act of 2020, as referenced in Rev. Rul. 2021-20, that "no deduction shall be denied, no tax attribute shall be reduced, and no basis increase shall be denied, by reason of the exclusion from gross income [of the PPP loan forgiveness]." Code Section 108, which deals with debt forgiveness, lists certain tax attributes that do get reduced by run-of-the mill debt forgiveness (which the just-quoted PPP special rule overrides), and my guess is that that is what Congress was thinking of in Section 276(a)(1) of the above-mentioned act, but I guess "tax attribute" as used in the act could arguably have a more generalized meaning, and arguably Section 415 compensation is a "tax attribute," and therefore in calculating it you would add back the amount of the PPP loan forgiveness. The IRS should probably address this. Would seem to me like Section 415 compensation is not a "tax attribute," but rather the Section 108 attributes are the only ones in play here. Anyway, the CPA is probably just addressing the fact that the partner has the distributable cash necessary to make the contribution. You might ask the CPA for backup. I suspect that none of the software vendors are going to change their SE and Section 415 calculations to gross up for PPP loan forgiveness, but you might want to check with the provider of your admin software. Bottom line, I tend to agree with you.

I suggest (a) saying in your plan document that you will have a policy, that may change from time to time, (b) having a written policy that says you give folks a choice, with a deadline, and (c) if they don't meet deadline, proportional to the breakdown between the two types of contributions in the account. The problem is that you can't really predict what someone would want. The Roth or more valuable, so you might want to distribute the pre-tax first. But as Belegarath notes, the participant might have been counting on the tax deduction, otherwise they would have done all Roth. Depends on the individual. And anyway, most plans that allow Roth Contributions will allow in-plan rollovers, so if the don't like that Roth were distributed, they can convert some of the pre-tax that's left to Roth, just in the later year.

I think this is an instance where common sense does not prevail. Section 416 uses the 318 attribution rules. See IRC sec. 416(i)(1)(B)(I) and Treas. Reg. 1.416-1, Q&A T-16. IRC sec. 318(a)(4) says to treat an option as equivalent to ownership. See also Treas. Reg. sec. 1.318-1. My recollection is that there is at least one PLR that says you don't count an option until it is vested, i.e., has become exercisable.

To the best of my knowledge the issue is not covered in the 401(k) regs. The closest I have come to finding "guidance" on the issue is in the IRS's List of Required Modifications (LRMs), from which the below for Excess Contributions. Virtually identical language in LRMs regarding excess deferrals: "Statement of Requirement: Excess Contributions for a Plan Year, plus any income and minus any loss allocable thereto, must be distributed no later than 12 months after such Plan Year. If both Pre-tax Elective Deferrals and Roth Elective Deferrals were made for the year, the Plan may provide that distribution of Excess Contributions will consist of a participant's Pre-tax Elective Deferrals, Roth Elective Deferrals or a combination of both, to the extent such type of Elective Deferrals was made for the year."

C.B. Zeller, no problem for separate DC plans, but DB would run into 401(a)(26) if controlled group, right?

Thanks for explaining all of this so clearly, Tom.

Thanks, C.B. Zeller.

I would also suggest including a short narrative explaining why the 1099-R is being corrected and that because the excess did not qualify for rollover it should be rolled back out to avoid penalties.

I could be wrong, but I just checked and I can't find it in either ERISA (Sec. 103 is where I would expect it) or the Code (Sec. 6058). Treas. Reg. sec. 301.6058-1(a)(4) says to be filed by deadline in form. Maybe I'm wrong and someone else will have the cite.

MarkS, are you saying everything's OK except that the availability of QNECs as source of hardship distribution was not communicated to employees? If that's the case I would argue it's not a Code/IRS violation, but an ERISA disclosure violation for which the correction is to communicate the change ASAP to all the plan's participants.

Whether someone has terminated from employment or not is a question of facts and circumstances. If he is no longer being paid by the company for his services, his activities may be supervisory in his role as a shareholder.

I agree that this may be a decent workaround. Just pointing out that no approach is going to conform to any IRS guidance, and could be questioned in an exam.

I just reviewed a proposal today from a recordkeeper I had not previously encountered but that seems to have a well-thought-out fee structure that does per capita. $120 per participant annually. It's a small plan for a relatively new law firm. $120 per head will come out to a little less than 18 basis points if it were pro rata. It seems fair to me for all but the smallest accounts, e.g. for a $10,000 account it's about 1%, which is high but not terrible. And if the employer matches decently, seems fair. I mean, why should, e.g., a clerical employee who's been with the firm for 10 years and has $100k in their account subsidize a young lawyer who is just starting out and has a small account? Obviously, that's a cherry-picked example, but it would be one possible case. Agree that it is more likely the partners would end up subsidizing everyone else, to differing degrees, but it would make more sense if that's the case for the partnership to just pay the fee itself rather than subsidizing it through the plan. I agree this can be looked at in different ways, but it seems to me that the DOL's guidance is correct, i.e., that it is a potentially fair way to allocate fees. It also is less likely to lull an employer into "fee creep" as the asset base gets larger over time. The fundamental question is whether a recordkeeper's costs are driven more by number of participants or asset size, and it seems to me they're more likely driven by number of participants. I guess the employer could also contribute, discretionarily, an additional $120, in this example, to each non-highly compensated employee's account, or to all NHCE accounts < some number, e.g. $10,000. Of course, if recordkeeping fees are calculated on a per-participant basis, but paid from shared mutual fund servicing fees, then they are borne by the participants proportionally to assets, not really per capita.