David Schultz

Also note that if you allow a specific employee to enter the plan early by naming them (or anything to that effect), you must pass coverage testing using the ratio percentage test as the plan does not satisfy the reasonable classification component of the average benefit test.

A triple stacked match is comprised of: An ADP safe harbor match (e.g., 100% of first 4% deferred) A fixed match (e.g., 100% of deferrals up to 6% of comp) A discretionary match (e.g., 55% of deferrals up to 6% of comp) The discretionary match is set/adjusted each year to ensure that the target owner/HCE gets to the 415 limit - as the above formula would do. All contributions are designed satisfy the ADP and ACP safe harbor. This is a true pay-to-play formula - you get 255% on the first 4% deferred, and just under 200% on 6% deferrals. Just make sure to document that eligible participants were informed of their right to defer and the formula.

The term-of-art here is "cutback" and the anti-cutback rule (Code §411(d)(6)) prohibits the retroactive reduction of an accrued benefit - and it applies to HCEs every bit as much as it applies to NHCEs. While the employer can decrease yet-to-be-declared discretionary contributions for 2024 and beyond, going back and changing an already deposited 2023 contribution seems like a clear-cut improper cutback to me. For my part, I would not be comfortable doing that. That is, I agree with Mr.@Bill Presson but used more words to say it.

Just curious, what is the accountant trying to accomplish (or, perhaps the questions should be why are they trying to accomplish that)? If the CPA expects that no 500-999 HOS participant is going to defer, isn't having them treated as LTPT (excludable from nondiscrimination testing and top-heavy minimum contributions) better than amending the plan to bring them in as traditional participants (where they drag down the test results and may have to receive an employer contribution). It seems to me that the desire to avoid LTPT status is likely to cost the employer more in the long run. I don't see the benefit...

I may be the origin of the above response you received. The process to enable electronic filing of the Form 5330 has been difficult to say the least and at present there is only one vendor who is capable of doing so. This is a problem vexing our whole industry and industry groups from ASPPA to SPARK (and others) are all approaching the IRS for relief. At the ASPPA Spring Virtual Conference last month, Kelsey Mayo (ASPPA's Gov't Affairs counsel) mentioned that she raised the issue with the IRS and the IRS's response was that their website (link referenced above) includes language adequate to claim relief. Unfortunately, thus far, the IRS has been hesitant to provide a more formal statement of relief, IMO but that language does provide a path for relief. Paper file and retain documentation as to why it was necessary. I will also note that the electronic filing requirement is not entirely new. Prior to this year, the requirement applied to taxpayers filing at least 250 forms per year. In light of what counts as a form, this requirement has applied to many plan sponsors for years, but - to my knowledge - no one was filing the 5330 electronically, and the IRS has taken no action against paper filers. What I do strongly recommend is that plans pay any excise tax due on a timely basis - but, of course, I must note that I am not providing legal advice.

A plan does not lose safe harbor status. It has a failure that needs to be corrected. Follow the terms of the plan document. Assuming you have a pre-approved plan document, and this plan was designed to be an ADP/ACP safe harbor, the document certainly says that the plan will satisfy ADP/ACP testing by satisfying the safe harbor requirements. It must still do that - in a nondiscriminatory way. I bet somewhere in the plan document (buried in the BPD section on safe harbor contributions) it also says that the SH contribution must be based on a nondiscriminatory (not just reasonable) definition of comp. If the plan didn't do so, it has an operational failure that needs to be corrected by doing what it is supposed to do (no -11(g) amendment as this isn't a nondiscrimination or coverage failure, it was an operational failure). The plan should recalculate the match based on a nondiscriminatory definition of comp. (and I would do an annual/true-up allocation since the correct periodic match wasn't deposited by the end of the following quarter). That'll be in the BPD as well.

I think the fundamental question is: Is there ANY language in your plan documents that authorizes the Plan Administrator (or any other party) to freeze a participant's account based only on knowledge that a DRO is being discussed? In most states, the filing of a divorce petition results in an automatic stay preventing the parties from unilaterally taking/transferring marital assets. If the participant does something improper - prior to the plan being aware of an actual DRO - then the court can deal with the participant's improper actions. It isn't the plan's place to intervene; the court can do that. The plan's duty is to follow its terms and provide benefits to participants, not to protect either party in a divorce proceeding. My belief is that such freezes are an operational failure (not acting within the plan terms) and potentially a fiduciary breach. I'd tread carefully (or preferably not at all).

I absolutely learned more than a few things from Mike Preston. He was kind and intelligent, but not in the least bit pretentious. This is sad news indeed...

I think we all like your rants, @Bill Presson, but I believe Peter is concerned about the attitude of - to paraphrase Shakespeare - the first thing we do, let's blame all the lawyers. (Sadly, that is an improvement on the original version). 😉

If your concern is the rule that restricts the adoption of an alternative defined contribution plan (Treas. Reg. §1.401(k)-1(d)(4)), that rule only apply to plans successor plans adopted by the same employer. In the facts you provide, I assume New Company is a new company, so that isn't an issue. If your question is whether you can start a new 401(k) plan in April 2024 that is effective on the first day of 2024 - sort of. You can start the plan now, the non-deferral portion can be retroactive to the beginning of the year, but deferrals can not be made retroactively - so you want to make sure the deferral provisions are effective when the plan is actually implemented.

You cannot retroactively adopt a safe harbor matching plan. Despite the fact that the non-safe harbor plan has a discretionary match, the IRS (fairly enough) does not like the idea of allowing plans to retroactively adopt the safe harbor (benefits) with a matching formula after the start of the year since, the participants didn't know that they had a guaranteed, fully-vested match available at the start of the year. If I understand correctly, you have a situation where a company was acquired and now the employer/related employer group has two plans: a safe harbor 401(k) and a non-safe harbor 401(k). That, in-and-of-itself, is not an issue. You can have plans that cover two separate groups, and those plans do not have to have identical features. One can be safe harbor and the other not. One is subject to ADP/ACP testing, the other not. The discretionary match can always mirror the safe harbor match formula. The employer could probably even make the match as a QNEC (or amend the plan to provide for full vesting) to mirror the safe harbor - if that is the objective. There rules have a few complications but aren't a major red flag. Do you have a specific concern regarding the different benefits/nondiscrimination testing?

This is indeed important detail. I triple down on my suggestion that the Plan Administrator hire counsel. Bad advice was given, bad advice was taken, and bad advice led to unfortunate consequences - for the granddaughter, for the son, for the Plan Administrator, for the plan sponsor, and perhaps for the 3rd party advisor. We cannot resolve those issues here...

Potentially. Again, I am going to recommend legal counsel. There are a number of issues here. But speaking generally, if the PA made an incorrect distribution, giving a beneficiary's money to the wrong party, then there is likely a fiduciary breach. Assuming there is a realistic belief that the PA could be liable for a breach, the PA can make the plan whole with a restorative contribution which can be used to pay the correct beneficiary. Separately, recovering the excess distribution from the other beneficiary can be pursued. This is not legal advice. The fact pattern here is interesting, but there is a lot of missing information that could reflect responsibility and/or liability by a number of parties for a number of reasons - or even that no one is due anything. To steal the line of a frequent poster here: free advice is worth what you pay for it. Unless this issue is over a trivial amount, I strongly recommend the plan consult with an ERISA attorney.

This was created by SECURE 2.0 Act 317 and was effective with the enactment of S2.0 (12/29/22). It modifies Code §401(b)(2) to add:

IMO, this is a scenario where the plan needs to hire legal counsel. It sounds like someone said that the plan terms provided for the default beneficiary to be the participant's issue (children) "per capita", rather than "per stirpes". (The comment about grandchildren not being ineligible to receive distributions is odd.) Why was this direction given? Assuming the plan did require distribution per stirpes, then the Plan Administrator erred in making the distribution of the entire balance to the son and it will have to make the granddaughter whole. The plan can likely attempt to collect the excess payment back from the son, but that may not be an easy process (i.e., the son may refuse) so the plan may need to review whether it wants (and the cost/benefit relationship makes sense) to make demands/file suit/etc. Keep in mind that the Plan Administrator is generally the employer. The Plan Administrators hires service providers, such as the "401k company," to perform functions, but the service providers generally avoid serving as a fiduciary and just take direction. The questions I'd ask are: Who is/are the proper beneficiary? And if it is not the son (only), how was the distribution made incorrectly in the first place? Most plan documents have clear language regarding default beneficiaries. Who gave the incorrect advice and who followed it? The answers to those questions will help to determine who should be responsible for fixing the issue.

Well...technically, a 401(k) feature cannot be retroactively adopted but that is what is being proposed. Again, I don't think it is a big deal in this situation, BUT it is technically a violation of the rules and if even a technical violation can be avoided by simply adding a delayed effective date (especially when it was suggested on the front-end), why wouldn't you? What is the harm in adding the 1/1/24 effective date for the 401(k) features? I believe that thepensionmaven is asking the right questions and is giving the right advice, and he should continue to do so; but where I'd now walk away from many a client who refuses to take sound advice, my initial comment was intended to reflect that, IMO, this isn't much of a concern.

The IRS gave some guidance on this topic in Rev. Rul. 2002-45 (https://www.irs.gov/pub/irs-drop/rr-02-45.pdf). The question comes down to whether the fiduciary reasonably determines that there is reasonable risk of liability for a fiduciary breach as a result of the surrender fee/MVA. From 02-45: I cannot make that determination for you or the plan, but the fiduciary's justification (or lack thereof) for purchasing the SVF with the MVA, the facts that gave rise to the change (i.e., an unanticipated merger), and the participant's opinions regarding the MVA would weigh into that decision.

In general, I agree with your sentiment. However, for an owner-only plan the situation is different. There are no effective availability issues, no operational failures, and no fiduciary concerns, etc. Since the owner is the participant, it really is a just a plan where the participant(s) has elected to not defer, as opposed to a plan where the employer decided to not facilitate, implement, or promote to the employees a plan feature the employer agreed upon when adopting the plan document. In short, I'd suggest that the document include the delayed special effective date for deferrals, but IMO, it is not a hill to die upon.

Unfortunately, that is a textbooks example of why prior year testing for the ACP test is generally a very bad option. What does your plan document say (the relevant language is likely in the BPD)? The pre-approved documents I know provide that the 3% is based on the first plan year in which the plan "provides for" matching contributions (see Treas. Reg. §1.401(m)-2(c)(2)). Effectively, the question is: when was the 401(m) plan adopted? The 3% rule exists because you can't have an ACP or ADP for the year prior to the adoption of the plan. IMO, by adopting a discretionary match, the plan "provided for" a match and created the 401(m) plan in 2022. The fact the employer elected not to make a discretionary match just means that they had a 2022 ACP of 0% - and fell into the trap associated with using prior testing for the ACP test. The plan would likely be much better served by amending to use current year testing for ACP; they can remain on prior year for ADP (but it is too late for 2023).

This would be a facts-and-circumstances situation, but verbal promises by the PA/PS on behalf of the plan are still promises. Could an employer elect a discretionary matching contribution in the plan document, but tell all of the employees in an enrollment meeting that there will be an unlimited 100% match allocated at the end of the PY, then elect not to make any match since the plan provides that the match is "discretionary"? It isn't just the plaintiff's attorneys I'd be concerned about. I suspect the DOL would take an interest if informed by the employee that the plan sponsor is [ ... what's that term? oh, yeah ... ] cutting back on promised (and accrued) benefits. Whether or not benefits were promised verbally is going to depend on who said what to who and how.

David Schultz doesn't like cameras too much and Brad is much better looking than he is... Agreed; I know that Kelsey agrees as well. I am just cautious into who's mouth I put my feet. And, yes, Kelsey knows her stuff - no doubt about that. I couldn't agree more that the complexities and "newness" of the issues here, along with a lack of guidance and a lot of articles that provide high-level analysis, can be dangerous. Though as the author of some of those articles (which I would argue went into more depth than most but still didn't uncover every possible issue), I will note that it is an iterative/takes-a-village process to walk down all of the paths and determine all of the potential operational issues for something so new and different (indeed, my articles have specifically called out that the devil is in the details). I've spoken with Kelsey, Robert, Derrin, Ilene, and many others about these topics in emails, phone calls, at dinner, and over drinks (I really need to get a life), and we are all still finding new issues and complications daily.

I disagree (and used this very scenario in a recently published Journal of Pension Benefits article on the LTPT legislation). To clarify my earlier comment: The pre-2023 service exclusion was added by S2.0 to the parallel ERISA provision (which is new and was added to extend LTPT coverage to 403(b) plans) - this is the S2.0 provision you quote above, but it only applies to the ERISA provision and 403(b) plans. The Code (401(k)) provision, was amended by S2.0 to change the requirement from 3 to 2 years, but it did not change the rule that only service periods starting before 1/1/21 are excluded for purposes of that provision (again, this only applies to 401(k) plans, pre-2023 service periods are excluded for 403(b) plans). I am confident that I am not alone in this interpretation. (I am pretty certain Ilene Ferenczy, Derrin Watson, and Robert Richter all concur with this interpretation, although I only speak for myself). I can say that Relius and Omni are being designed to determine eligibility in this manner. And I do not read that Sayfarth article to disagree with my interpretation. A fundamental principal of service crediting is "all service counts" (credit to Robert Richter). The only time prior service does not count is if there is a rule that excludes it (such as a break-in-service rule or a law, such as SECURE Act §125(b)) which provides that pre-2021 service is excluded for purposes of the LTPT rules that apply to 401(k) plans. The Seyfarth article is saying that in 2025 the service requirement is reduced from 3 to 2 consecutive YOS. But I see nothing in the law, regs, or Seyfarth's article that says pre-2023 service is excluded for purposes of applying the 2 consecutive YOS requirement. I do not wish to be argumentative, but I think this is an important point to ensure everyone understands correctly (and I agree, this is a PITA result): When the S2.0 change from 3 to 2 consecutive YOS kicks in, my interpretation is that anyone with 500 - 999 HOS in 2 consecutive 12-month periods that begin on/after 1/1/21 will be eligible to enter as a LTPT (assuming they were at least age 21 when they completed that 2nd consecutive YOS). Those consecutive periods can include 2021 and/or 2022.

I get where you are coming from but there are pros and cons. LTPT certainly does add more moving parts and complexities, and with those come risks. But making everyone eligible immediately, or after completing 500 HOS, has it's downsides too (testing. top-heavy, and employer contributions applying to the otherwise excludable employees who could be excluded from all of that if treated as LTPT). You and your clients will need to weigh those pros and cons and decide what's best for you. Personally, I am a fan of anything that gets more participants and money into plans!

@austin3515, I am not sure that this is true. SECURE 2.0 Act §125(a)(1) modifies ERISA §202 to add the LTPT requirements (making them an enforceable right and extending the rules to 403(b) plans. The new ERISA §202(c)(1)(B)(i) adds the 2-consecutive year rule and the new §202(c)(4) provides that periods prior to 1/1/23 may be disregarded for the purposes of that section. However, SECURE 2.0 §125(a)(2) also modifies Code §401(k)(2)(D)(ii) to reduce the eligibility requirement from 3 to 2 consecutive YOS, but it does not modify the rule in SECURE (1.0) Act §112(b) that excludes service prior to 1/1/21 for purposes of Code §401(k)(2)(D). That is, if an employee (at least age 21) works 750 HOS in 2021 and 750 HOS in 2022, but less than 500 HOS in 2023 and 2024, that employee would be eligible to enter the plan on 1/1/25 based on the requirements in the Code, when the requirement is reduced to 2 YOS. You are reading the 2023 limit for the ERISA provision and applying it to the CODE provision, but unfortunately, IMO, that is incorrect.

Why in Heaven's Name would you do that? Because participants are good investors or the costs go down?