Mike Preston

There is a lot of material there. I probably won't have time to read them until after 4/15. Thank you for forwarding the material.

It is a sad world when the answer to questions such as this one are not obvious. But the fact is that some IRS folks have, indeed, put forth the theory that the appropriate measure of the prior benefit is $10,000, not $6,000. They are wrong, of course. And no amount of hand-waving and gesticulating about the "accrued benefit" being $10,000 makes any sense at all. If the benefit paid was $6,000 then the benefit to offset against the current limitation is that $6,000 benefit. Why? Because if the $10,000 was the true "accrued benefit" then the plan would not be a qualified plan if it were to pay something less. Something about an "impermissible forfeiture" comes to mind. However, one never knows if the next "crazy" idea (such as this) will gain traction with some faction at the IRS or the Treasury and we will find ourselves having to respond to a formal challenge along these lines. What a waste.

It can royally mess up your nondiscrimination testing. Example: presume you have a young, daughter of the owner, age 23, with two years of service. Once that person gets the 3% contribution, a new comparability design may be hopelessly fouled.

I could be wrong on this, but my understanding is that the underlying rates that this section is referencing did, indeed, expire at the end of 05. This makes is somewhat irrelevant as to whether this particular section expired. Because with it or without it, the result is the same. Weird, huh?

Remember that the 3% non-elective is reduced by matching contributions, if the document provides. Yes, the Company has to make the contribution; there are no other options short of determining that the plan is not really TH (sometimes there are differences between practitioners). The top-heavy test would not be recalculated. However, this scares me a bit. Except in the case of a new plan, the date that is used to determine whether the plan is topheavy is the last day of the prior calendar year. Hence, you would generally look at the balances at 12/31/2004 to determine whether your plan was top-heavy for 2005. If you are even contemplating making a change to the TH test based on refunds for 2005, you are a year off on your thinking, maybe. Refunds of deferrals (ADP) or matching contributions (ACP) do not impact the test. If your TPA is not much help you need a new TPA.

Beltane is presuming a calendar year plan. If so, I agree. If not, while W-2's are not amended, sometimes the taxation is with respect to the earlier calendar year.

You want an old law answer or a new law answer? [What is your anticipated implementation date of the new rules.] You want an answer consistent with the document or a general response? [What are the actuarial equivalence, suspension of benefit, and actuarial increase provisions?] I have stayed out of all DB minimum discussions because the darn issue is a complex one that usually requires enough time to be billable. And as we have more people beating on us for 3/15 and 4/15 calculations than we can shake a stick at, there just isn't time to do justice to the issue, in general.

One way to avoid it ever being your problem is to strongly recommend, in writing, that they have it reviewed by counsel and, at the same time, lay out the dire consequences for them if the get it wrong.

Yes.

OK, I'll bite. What do you have in mind here, Tom?

Of course. You need to pass both 410(b) and 401(a)(4). Assume for a moment that each plan does, in fact, pass 410(b) on its own. You then can, if you want, test each plan separately for compliance with 401(a)(4).

It is not uncommon for a single individual to get both a 1099 and a W-2 from the same company, year after year. If the company's counsel says it works, I'd not be inclined to fight it. But, it is legitimate only if it is .... and this will sound silly.... legitimate. And if it is not, the employer's 401(k) plan is probably at risk. Hence, the onus is on the employer to "get this right." In other words, I can't see why an employer would go along with this unless it was.... you knew this was coming, right?..... legitimate.

Indeed.I don’t know how it would interpreted any other way by the Service.

Hint: maybe even with restructuring! Bingo! Believe it. Embrace it. Celebrate it.

You are reading too much into my response. Forget the fact that this plan has a questionable grouping as to service. Assume that there was a PS contribution being allocated to all folks in the plan with 1 YOS. Assume it was 3% of pay for some folks and 10% for other folks. ("folks" = both NHCE's and HCE's) Clearly does not satisfy gateway. Passes 410(b) with flying colors (everybody benefits). Can we crosstest the ABPT? I will demonstrate compliance with 401(a)(4) later, so ignore that for now. Yes, say I. What say you? What did your message say?

Both to the extent that they imply usage of crosstesting in the ABPT presents a gateway issue of any kind. It does not.

401a26

It is actually incorrect for a different reason. One may use the ABPT without being subject to the gateway at all. However, is anybody else concerned that an allocation only to those who satisfy 10 YOS is a defacto eligibility requirement of 10 YOS? Gee, and I thought 2 years was the max.

It very well might be a hole in the Rev. Proc's logic, though. Seems like a great question for the IRS at some point. Of course, we aren't likely to get clarification before this year's witching hour (3/15/06), but I doubt that even if it is a hole in the logic that the Rev. Proc would then stand, unmodified. Once the IRS acknowledged this theoretical hole, they would patch it such that a 412(c )(8) amendment doesn't cause a plan that would otherwise have a remedial amendment period to lose that period. Is that merely wishful thinking? I know there has been a "sea change" at the IRS, but if they were to go this far, I would be incredulous.

I agree, as long as you are saying that the entire population to "look at" is that which is determined solely with respect to employment by B. I guess I would say that you pretend the AB plan covered two groups, the A group and the B group. Looking solely at the B group and going back as far as one needs to satisfy the TH calculations and key employee determination issues, determine whether the B plan is top-heavy for 2006. In the case of employees that worked some time for A and some time for B there are issues that need to be resolved.

Do I understand that what you are suggesting is that you can get around 404(a)(7) by establishing 4 plans where previously only 2 existed? In the case at hand, we already were stretching 401(a)(26) with counts of 25 DB plan participants and 50 DC plan participants. I overlooked that because there may have been some permissive disaggregation which somehow enabled the DB plan to squeek by. But if you now want to break the DB plan down to 2 plans, one with 20 and the other with 5 (where the 5 are not subject to 404(a)(7)) then unless those 5 are all NHCE's who are approaching retirement age you will either fail a26 or have not a whole heck of a lot going for the deduction. Hardly seems worth the pain for the gain.

I would never, ever counsel a client to interpret a plan document to match anything other than all deferrals, catchup or not. That is, if there is the slightest doubt as to how the language could be interpreted. Further, if the document seems clear that matching contributions do not extend to catchups, I would counsel that they should modify their document to extend the matching contributions to catchups. I am very concerned that the IRS will somehow view catchups not being matched as either an attack on the universal availabilty requirement or as a discrimination issue. Take the case of a plan where in a given year the only catchups in the plan are made by an NHCE. Those catchups are not matched. Had they not been catchups they would have been matched. Hasn't the plan violated 401(a)(4) because the rate of match for the NHCE was somehow less than the rate of match that an HCE might have received had they made the same amount of contribution, merely because for the HCE none of the contribution was catchup? Think plan cap of a specific percentage of pay on deferrals, like 75% (the safe harbor in the regs for universal availability with respect to catchups themselves - but not extending to matching contributions, IIRC - but somebody should check me on this point). Think part time NHCE who was full time in a prior year and continues to participate. Think second income earner who wants to defer the max. Why is this person, who is deferring a small amount, in a position where the matching contribution is lower just because some portion of the deferrals are catchup? Yes, I know we can get around this issue by introducing a cap on the match that is percentage of pay related, but some plans don't have such a limit. Don't have time to construct a complete example during contribution calculation season, but wasn't this issue pretty well hashed out moons ago?

I don't think they need to. If the amendment is something that is required for qualification purposes, I don't think 412©(8) saves you. On the other hand, if the application of 412©(8) is solely with respect to minimum funding, which is what I think we are stuck with, I don't see the Rev. Proc. undoing what is clearly in the code.

I think this is a most novel approach to how 404(a)(7) is to be applied. Not that I dislike it. I'm rather fond of it. But that is just me. I dare say the IRS would be whatever the opposite of fond is. In fact, they may go even further. Not that I would blame them. By the way, do you have anything, anywhere which supports your novel approach? I believe that 2.0 * .25 = 5 and therefore $500,000 is your 404(a)(7) limit. To reword your conclusion slightly: The 404(a)(7) limit only applies if there is at least one overlapping participant.

Yes, you can. If you want to provide for an allocation to the active but less than 1,000 hour employee participants, and the normal provisions of the plan do not do so, you must amend. You have the normal -11g period to do so.