Belgarath

An excellent point! Thanks.

That's not my understanding, albeit I'm not remotely passing myself off as any kind of expert! But a simple internet search turns up many sources, that seem reputable, that say an employer may contribute, tax-free, to an employee's HSA EITHER through a 125 plan OR outside a 125 plan. The trade-off is that if not through a cafeteria plan, the employer contributions must satisfy "comparability" rules. If I'm wrong on this, can you provide any citation/guidance? Thanks!

Let's say you have a 401(k)/PS, Safe Harbor nonelective, with 1 month eligibiity for deferrals, (entry date first of month coinciding with/next following satisfaction of eligibiity requirement), and 6 months/500 hours (entry date 1st of month coinciding with/next following satisfaction of eligibility for both SH and PS. Plan IS top heavy. PS plan is cross tested, everyone in own group.Compensation for SH/PS is from date of participation. So some people enter on 10/1/2019 for deferral purposes, and therefore get top heavy. Therefore, unless the "plan" of employer nonelective contributions is disaggregated, they are also required to get gateway. However, they have not yet satisfied the eligibility requirement to be able to receive a PS. Now let's suppose the "plan" does disaggregate the OEE's, so they are no longer required to get gateway. Two items to see if my memory serves: If the employer does NOT disaggregate the OEE's, how would one allocate a Gateway - seems like this situation would preclude cross testing. Agree/disagree? And IF the employer disaggregates the OEE's, and WANTS to contribute a PS for them, I recall the employer would be able to allocate to those OEE's who have otherwise actually satisfied plan eligibility, (say they were hired in February of 2019, for example) but gateway would still not be required, and this disaggregated "plan" could still not cross test. Agree/disagree? Sometimes I think I remember things, and find out I don't... P.S. - I suppose I should have considered document provisions that allow a gateway allocation to participants who are not otherwise eligible to receive it. Some documents even allow it to people who are disaggregated if it is necessary to satisfy 401(a)(4). That wasn't available in the really "old" documents, and I think the "old" documents are what I'm remembering.

Agree with Larry, although we've made exceptions on rare occasions, and generally have deeply regretted it!

Have you checked the "boilerplate" top heavy section on the body of the document in (a)? If a pre-approved plan, it will almost certainly address this in some manner, although it may just refer to the adoption agreement! However, without knowing anything about the specific language involved, my initial thought is that as long as (b) provides the requisite TH for all participants in either plan, it should be sufficient, assuming there is any wiggle room in the (a) document language. Even if there isn't, the Plan Administrator is responsible for interpreting the document, and perhaps their interpretation would be elastic enough to avoid a second TH contribution in the (a) plan. Good luck.

Interesting. That hasn't been my experience, nor was it at my prior large employer. Maybe we are talking about different "plans" or situations? I'm referring to a situation where the employer provides free group term life up to $50,000. Then, voluntarily, you can elect to PURCHASE, and pay the premium for, additional group term life up to a maximum of, say, 5x your salary. So if you make $60,000, you can purchase up to $300,000 in group term life. The premium is deducted from your paycheck on an after-tax basis. Are you saying that in this situation, where you are paying the premium with after-tax dollars already, that you are also taxed on imputed income? If so, could you please provide some citation or proof that this is the appropriate treatment?

Agree with Cuse, but if by "offer" you happen to mean that the employee can purchase the insurance (say, up to twice their salary, or sum such thing) and they pay the premium, then no, there's no imputed income.

After looking into this a bit, it seems to me that the easier route is for them to simply make the employer contributions to the HSA's, without the fuss of setting up a cafeteria plan. They have to satisfy the "comparability" rules, but that doesn't appear to be an issue. Am I missing anything?

Thanks Peter. I wonder if Ilene (or anyone else) could comment on the following excerpt. Every practitioner I know busts their tails to respond promptly to IRS inquiries. Is this IRS "enforcement" going to be a hard-line? What if the practitioner/client/accountant/attorney is on vacation - etc., etc., etc... I'd love to hear what the IRS REALLY means by "stricter compliance" in all of this? What is "abuse" of the current VCP procedures in the judgment of the IRS? The simple statement below doesn't really provide much guidance or comfort/assurance that "reasonable" delays/extensions will be accommodated. "IRS TE/GE attendees at the TE/GE Council acknowledged that there is some intention to require stricter compliance by practitioners who have been abusing the current VCP procedures. For example, the IRS is intending to enforce the 21-day period it provides for practitioners to respond to requests for additional information."

H?ave never really seen much on these, probably because I haven't looked. I'll have to do some reading! But suppose you had a business with only 5 employees, who only wants to make employer contributions to an HSA for their employees, because they (employer) can't afford to pay insurance premiums. All employee/insureds are under a HDHP. Is there any reason why a "SIMPLE" 125 plan wouldn't be appropriate here? Employees would presumably also contribute to the HSA.

Have you ever seen a document that provided/allowed such a contribution to be allocated only the HCE's? Even if the citation doesn't convince you, I find it hard to imagine that most (if any) IRS pre-approved documents, which is what are nearly always used these days, would allow this. So even if you are convinced it is otherwise allowable, check your document language very carefully. We've been asked the same question before by a group of doctors, and our answer was a resounding rejection. Also, if the first citation is unconvincing to you, so that you move past it to the next step, then an allocation DATE is something I'd argue is an "other right or feature." Although the regulations don't specifically list this (unless I skimmed over it - entirely possible possible!) under 1.401(a)(4)-4(e)(iii) examples, it does say "Other rights and features include, but are not limited to)" That's my basic take on it all, and I'm "stickin' to it." I sure as heck would not want to try to defend such an allocation procedure if questioned on audit, and I wouldn't waste any time looking for a way to try to get around it. I guess what I'm really getting at is I think it is easy (and correct) to not allow it, and I think it is difficult, if not incorrect and impossible) to reasonably defend it. Sorry if this seems blunt - I'm a little Grumpy this morning. Or Sleepy, or one of those Dwarfs...

Someone else in the office is using that volume of the EOB at the moment, so I can't quote it, but I should think that 1.401(a)(4)-1(c)(8) would prohibit this. Also fails the smell test...

Maybe I'm misreading your post, but I don't see why they can't amend to prior year testing for the 2020 plan year? It sounds like the plan is less than 5 years old, and has always been current year tested - is that correct? If so the plan satisfies the "5 year rule" and they could amend to prior year testing for 2020.

For folks using it - opinions?

Gracias.

Cross-tested plan. Compensation for allocation purposes excludes certain items such that 414(s) testing is required. Plan requires last day/1,000 hours to receive employer contribution. Pan is NOT top heavy, nor is it a safe harbor. Question is this: Terminated participants with more than 500 hours, so not excludable - when doing the rate group testing, for purposes of the 70% ratio test for the rate groups, do these people have to be included for purposes of 414(s) testing? Seems like I recall that they are not included in the 414(s) test for this purpose, as they are not benefiting. In other words, are they automatically included in the 414(s) test if you go to the average benefits test, or only if 1 or more rate groups fail the ratio test, and you must move on the the ABPT? The plan in question passes the 414(s) with flying colors regardless, so this is more a question to avoid a future brain cramp. Thanks.

Hi Kevin - I think we'll just agree to disagree on this issue. As far as the EOB goes, I think the sections are perfectly consistent. The plan can offer one or the other, or both. (Just as an aside, this is a purely academic exercise for me, as our plans do allow a split election.) And I don't think it is a tortured reading of the regulation, but of course I could be wrong on that. I'd love to see the actual PLAN language in OP's plan to see what, if anything, it actually says on this. I have a suspicion that the Plan may not specifically address this anyway. It's too early on a Monday morning for me to delve into the BRF issue - the coffee hasn't kicked in yet! (But I'm very much inclined to not worry about it.)

Hi Kevin - ok, so based on your EOB citation above, this I just pulled out the EOB to take a look - 2018 version. Chapter 11, Section XV, Part A.1, 1.c.1 very clearly states that the plan is not required to divide the election between pre-tax and Roth. FWIW, I just checked the 2016 EOB, which is the oldest version we have handy, and it is the same as what I cited above. So I'm still a non-believer at this point. Or to perhaps be more precise, the plan isn't required to allow an employee to divide the election between pre-tax and Roth...

Hi Kevin - not sure I agree with that interpretation. I think the phrase "to the extent permitted under the plan" allows an employer to require all deferrals to be one or the other, as long as you are permitted to change. So although nearly all plans permit a split election, I don't think they are REQUIRED to.

All right, where's Ned Ryerson? ?

C.B. - I think you hit the nail on the head. Although it LOOKS funny (on the Relius report) to have the bottom of the rate band used for the HC, (and not for the NHC) this is only for purposes of the 70% ratio test for the rate group, and all the people in that band are included. For purposes of the ABPT, then "real" benefit percentages are used. So I think it is all ok.

Based solely upon what you have stated, and assuming it is all completely correct... In a general way, this (sadly) isn't all that uncommon. The employer can go through the IRS "VCP" program, which will involve make-up contributions for the "missed deferral opportunity," make-up matching contributions, and interest. If done and calculated and submitted to the IRS properly, there's no reason why the IRS wouldn't accept this correction. Many (most, all?) of the Third Party Administrators on these boards have done corrections for similar situations, and received IRS approval. Caveat - this is just general information, and please do not view it as being tax, legal, or definitive advice of any sort.

I would merely observe that although it was handled incorrectly by the Plan, (assuming this was a direct rollover), the PLAN has nevertheless satisfied its requirement to distribute the RMD for the year. (At least, that's how I understand it...) It is just that the RMD amount was incorrectly sent as a direct rollover. And almost certainly reported incorrectly as well. A screw-up no matter how you look at it.

2018.

This really is a question for Relius, but perhaps some of you have dealt with it already. Suppose you have a midpoint EBAR (for HCE's) of 8.740. So any EBAR's within 5% above or below this midpoint are considered to have an EBAR of 8.740. This gives you a range of 9.177 to 8.303. In the Relius testing, it uses the LOW of 8.303 for the EBAR in the testing. So a HCE with an unbanded EBAR of 9.1, for example, is in the testing at 8.303. Is this correct? I thought that the testing would have to use use the midpoint of 8.740.