Doc Ument

There appears to be a comment flag on my earlier post suggesting that what I wrote was confusing. Perhaps that is because I was trying to say that solo plans are, by definition, confusing because the word "solo" does not appear to have a uniform definition across the industry. From my view, the word "solo" implies only that the plan was optimized (whatever that means) in some way for small businesses (whatever that means). Such optimization is not required by law but is discretionary for the vendor. For example, one vendor has a DC plan available for every employer, and also has an alternative DC plan available for any company that wishes to use the alternate DC plan, and the latter product is something that vendor calls a "solo" plan. The only reason that vendor is offering such a plan is because there is a market that generally coincides with the employers that historically would have used a Keogh plan. Rather than forcing such employers use the equivalent of a Keogh plan, they might have a plan that is much more flexible that a Keogh plan. Each vendor will take its own course, which is why I cannot answer questions about solo plans (as if all solo plans are the same) unless I can examine the document (since not all solo plans are the same). The word "solo" in and by itself offers no clue as to the nature or the extent of how a plan product has been optimized for the targeted market, but tells me only that such optimization has occurred. It therefore would not surprise me if I saw a solo product that didn't cover common law employees or that prohibited MEPs, but I would not make such assumptions. For example, I believe that you will not find any reference to solo plans in any regulation or IRS Revenue Procedure (not even the Revenue Procedure for Cycle 3 and which governed the approval of Cycle 3 solo plan products). I am relying on Window's file search function in making this statement (I have a folder that contains 90% of the guidance issued on retirement plans over the past decade). In contrast, when I do a general search for "solo plan" on the internet, I find many listings on the first page, each presenting a different definition, such as "an account for the self-employed," or "a single-participant plan," etc. My favorite one states: "A solo 401(k) is a traditional 401(k) plan, except that it covers only one employee -- the sole proprietor of a business -- and, at most, a spouse." Notwithstanding such definitions, the solo plan with which I am the most familiar covers every employee other than certain statutorily excluded employees (e.g., nonresident aliens with no US income), with no option to exclude any more. If I used only the internet for my answers, then the solo plan I am using shouldn't be called a solo plan. The IRS didn't mind, though, because they (apparently) don't have a definition of solo when reviewing solo plans. Consequently, when I see an inquiry that poses an inquiry regarding solo plans as if all solo plans are alike, I am usually quick to indicate to that person that I am unable to generalize anything about the nature of the solo plan without knowing a great deal more about THAT particular solo plan. For example, knowing that the plan is a solo plan does not inform me of the nature or scope of that vendor's optimization for the targeted market. The fact that it is a solo plan does not tell me what consequences will arise if another employer adopts the plan or if any adopting employer hires an employee. Even if that vendor's software does not automate any additional employer participation agreements, it is conceivable that the BPD not only allows additional adopting employers, but also allows MEPs. That means I must scour each specific "solo" document to answer questions about the nature and degree of its being "solo," such as its restrictions on additional participating employers and on which employees may be or must be covered (or cannot be covered).

I agree that ADPSH contributions are preferable to relying on QNECs and QMACs, but I hesitate to say that the latter are obsolete. While not every plan product will necessarily contain a preapproved option for making non-corrective QNECs and QMACs, there's nothing to preclude a document architect from presenting drafters with such sources that can be funded ahead of time, i.e., regardless of the outcome of an ADP test. For example, some plans allow prevailing wage contributions to be treated as QNECs, which would allow such contributions to either be used to help pass an ADP test or be used as an offset to an ADPSH nonelective contribution. Such a plan might prefer having an ADP test rather than an ADPSH obligation. In addition, there might be a use for treating ADPSH contributions as a QNEC or QMAC for a plan year that the employer discontinues its ADPSH. There was a time when any ADPSH contributions that had already made for such a year could be automatically treated as being a QNEC or QMAC for that purpose. However, now that plans with QACAs can have deferred vesting on the QACA contribution, employers may be unable to use such contributions to help pass the ADP test unless an amendment is adopted to fully vest the QACA source, i.e., turning the deferred-vesting QACA contribution into a QNEC or QMAC. The biggest problem I see with QMACs and QNECs (and occasionally non-QACA ADPSH contributions) is with people thinking that full vesting is sufficient for calling a contribution a QNEC or a QMAC. I don't know why, but I see many instances of people forgetting about the distribution restrictions that are required for all these contributions. I like the idea of having a plan product that gives me a QNEC or QMAC source that has all the requirements built-in, if only to facilitate quick amendments with reliance when the facts and circumstances change.

It might be that the commentators thus far know more about the marketplace than I do. As a technician, however, I have some observations that might potentially be useful, or perhaps they are entirely out of date. I second the motion that CuseFan has made about inquiring about "other participating employers." Allow me to provide more detail that I suspect is behind that suggestion. More than once, I have seen people mistake a contemporary "solo" plan with what used to be called a "Keogh" plan. There's a reason why document vendors have stopped referring to plan being "owner-employee" plans or "Keogh" plans, and that is because there is no longer a legal need or any legally motivated desire for having separate plans for only a certain type of business or a certain type of "employee." However, from my limited perspective, that's not to say that some vendors haven't chosen to continue using ancient (pre-EGTRRA) legal divisions for "Keogh" plans and given the product, thus designed, a new name of "solo." You could have two very different documents calling themselves "solo" plans. Until you know the nature of the vendor's solo document, i.e., the philosophy that underlies its construction, you cannot be certain whether it makes any difference whether it is a single employer plan or a MEP. Some solo plans might preclude a MEP, other solo plans might preclude additional employers even if related, and some solo plans might allow for all of those possibilities. Some solo plans might preclude having any common law employees, and some solo plans might not. That is because there does not appear to me to be any "solo" qualification rules except for the rules that the document vendor has designed into their product, and if they have done so, then you are required to abide by the document even if the document has restrictions that are quite unnecessary. It is quite possible that a vendor has continued to have restrictions that are no longer necessary. Perhaps it was deliberate. To summarize thus far, "solo" is not a term that has a definite meaning to me unless everyone is convinced that every document vendor attaches the same meaning to the word "solo." For example, the <only> difference I see in the particular solo product that I use is that the solo AA is a subset of the "regular" AA's options. There's been no deletion of any provisions regarding common law employees nor any restriction on the nature of additional participating employers. There is no difference in language between the "solo" BPD and the "non-solo" BPD. That means that nothing bad happens (under this "solo" plan) if the employer ends up hiring a common law employee or having a related or unrelated employer adopt the plan as an additional participating employer, other than the indirect costs associated with having a plan document with pre-selected AA options that cost more money (in operation) than having a more efficient plan design that is built using the "regular" AA. For example, maybe the solo plan provides only full and immediate vesting, whereas the regular AA has options for deferred vesting. Or perhaps the solo product lacks annuity options whereas the regular AA does. I made up those two examples, as I don't have the documents in front of me. I merely am illustrating that I know that the primary reason for my vendor's separate "solo" product is so that advisors can offer simpler looking documents to certain employers (or their advisors), i.e., those who don't need or don't want a zillion AA options, and consequently the advisor knows that the record keeper can make presumptions about every "solo" plan's provisions (i.e., presumptions that cannot be made about the "regular" plan's provisions from that same vendor). The advisor can then design uniform administrative procedures for all its solo plans and give such projects to staff having less expertise than the staff who administer more sophisticated plans. Perhaps an advisor limits their practice to employers that fit on a particular vendor's solo product. However, it is possible that you have a plan that is designed from the ground up to be limited to only a specific type of employer, or that prohibits having any common law employees from becoming participants. That is why I recommend that you determine exactly what "solo" means to your document vendor. (You might be able to get all the answers by examining the document's language.) As you suggest, it is conceivable that the plan is designed to accommodate only single-employer plans. If the plan does not permit a MEP, then don't forget to also consider the possibility of an affiliated service group of any type of businesses, regardless of their format or their manner of taxation. For example, even if Mom only makes and bakes the pizza using her corporation (or "company") taxed as a proprietorship and Dad delivers the pizza using his corporation (or "company") taxed as a proprietorship, then although they might not be a controlled group (due to the spousal rules), they might be an affiliated service group, especially if all of Mom's pizza gets delivered by Dad, and Dad distributes only Mom's pizzas. To summarize, maybe it does matter and maybe it doesn't matter what type of company is adopting the plan or how each entity is taxed. (It may or may not matter if there are common law employees.) Unless we agree that all document vendors mean the same thing by the word "solo," then such matters depend upon that particular document vendor's definition of "solo."

I have no opinion on the administration of the plan. From an IRS perspective, though, I recommend reading Revenue Ruling 96-47 (i.e., potentially a “significant detriment” as applied to only terminated participants), and from a DOL perspective, I recommend reading the NAPA article referenced below. That article cites DOL guidance and concluded that “a plan may charge administrative expenses to terminated participants, while not charging active participants, provided the method is not a breach of fiduciary responsibility, and the expenses are proper, reasonable and done in a nondiscriminatory manner.” https://www.napa-net.org/news-info/daily-news/can-plan-charge-fees-terminated-participants-not-active-ones#:~:text=The%20DOL%20and%20IRS%20have,done%20in%20a%20nondiscriminatory%20manner

Your question strongly implies that you have found the hold-out language (or option), but I point out that not every document contains that rule or offers it as an option. If it's there, ask the document provider. If it not there, you cannot use that rule.

I see two murky issues here, the “mistake of fact” issue and the “forfeiture suspension” issue. I urge you to do as much research as possible into the examples of why the IRS has allowed a return of contributions due to a “mistake in fact.” In this case, I do not think (based on my research years ago) that forgetting a fact is a mistake of fact. A mistake of fact would be more in the nature of allocating an amount in reliance upon a W-2 that subsequently was determined to be inaccurate (i.e., there was a correction to the W-2 after the time of the contribution). In this case that you present, however, the employer had all the facts available and could have and should have ascertained whether or not the true-up had already been deposited. The mistake that was made was the failure to confirm that the deposit had already been made. I think the IRS would view the earlier deposit as being a correct fact, and the employer’s only mistake was to forget that correct fact, which the IRS would call a mistake of plan administration. If forgetting facts justified asset reversions, I think many more employers would be forgetting many more inconvenient facts. I have heard stories of unhappy audits involving the defense of having made a “mistake of fact.” Because this topic is so murky, many plan documents do not attempt to define what is meant by a mistake of fact. When that is so, I suspect it is because the provider believes that the definition of a mistake of fact is that “the IRS knows a mistake of fact when it sees one.” I recommend that counsel be obtained before an employer relies on any plan’s vague “mistake of fact” language. If you see plan language covering this situation, then sure, you can go ahead. I don’t think you will find such language. Starting with the PPA cycle, the IRS has not allowed provisions in preapproved plans for “forfeiture suspension” accounts. The IRS caused quite a ruckus several years ago when they published an article in their retirement plan practitioner newsletter “reminding” the community that forfeitures cannot be carried forward to future plan years in any discretionary manner for any extended period. It is my understanding that the deadline (by which forfeitures must be disposed of) must now be stated in the document, and so far as I know, the most liberal standard currently permitted is that forfeitures that arise (i.e., are declared) in one year must be disposed of by the end of the subsequent plan year. (I acknowledge that a plan might have suspension accounts, but I suspect that a preapproved plan does not have a forfeiture suspension provision that is longer than I have described.) In fact, that newsletter article, as I recall, was written so as to discourage even the short delay that I have described in this paragraph, but I know of preapproved plan documents that use the short delay described in this paragraph. I had a link to the IRS newsletter but it no longer works. The approximate time of publication was 2015. A forfeiture is “disposed of” when the employer uses forfeitures for a purpose stated in the plan, e.g., paying expenses, reallocating such amounts, or reducing contributions, etc. Some plans provide flexibility in this regard, others do not. Some plans have ‘ordering” of methods of disposal, and others do not. Bear in mind there might also be a delay between the time that a termination of employment occurs and the time the resulting forfeitable amount is declared. That period might be as long as a 5-year break in service, depending upon the terms of that document. The time (if any) that a plan provides for disposing of a (declared) forfeiture starts to elapse only after the time period, if any, that elapses between the forfeitable event and the declaration of the forfeiture (which will vary from plan to plan in accordance with the language of the document and, usually, the employer’s elected provisions in that regard. It is also possible that the plan document does NOT contain the provision that the employer can wait until the end of the year following the year that the forfeiture is treated as arising, in which case the employer should (conservatively) dispose of the forfeiture in the same year that the forfeiture is declared, in which case the subsequent valuation report for the year (normally not done until the year has ended) should reflect the disposal of the forfeitures declared during the year covered by the valuation report. It might well be that the IRS doesn’t find that short delay for disposing of forfeitures to be objectionable when there is no language in the plan supporting that practice, but I cannot say for sure that that is the case. I acknowledge that the IRS does not necessarily catch everything that it says it will enforce uniformly for a particular Cycle. As a result, some preapproved plans end up having provisions that were not allowed by the IRS for other provider’s documents. For that reason, if you find preapproved plan language that is more liberal than what I have described, I suspect such language was an oversight by the IRS, and if so, then you should feel free to follow the terms of the document so long as you are confident that the document currently has reliance. You might want to take a look at the subsequent Cycle’s document’s language in this context when the time comes to see if the language changes, as the IRS often notices things in a new cycle that it realizes it should have already noticed in a previous cycle.

I agree that the ADP safe harbor match gives a free pass with regard to ADP testing, but I believe that (1) the free pass does not extend to the ACP requirements, and (2) that there is more to the ACP safe harbor than just complying with the percentage limits and avoiding any allocation conditions. My reading of the regulations is that you cannot have an ACP safe harbor without putting that provision on the AA (for a preapproved plan). That is why it is on the adoption agreement. There needs to be plan language adopted that sets forth the promise of maintaining the ACP safe harbor as the plan's specified method for complying with the ACP requirements, just as there must be plan language adopted for the ADP safe harbor that sets forth the promise of maintaining the ADP safe harbor as the plan's specified method for complying with the ADP requirements, which is why an amendment is required to get out of either an ADP SH or an ACP SH midyear. I say that because... Regulation 1.401(m)-3(f)(1) states in pertinent part: (emphasis added) “(1) General rule. Except as provided in this paragraph (f) or in paragraph (g) of this section, a plan will fail to satisfy the requirements of section 401(m)(11), section 401(m)(12), and this section unless plan provisions that satisfy the rules of this section are adopted before the first day of that plan year and remain in effect for an entire 12-month plan year.” This rule’s language is identical to the ADP safe harbor regulation’s language. I believe it means you are committing, via a timely AA election, to (1) maintaining both the ACP safe harbor and an ADP safe harbor (of any type) for the entire plan year (since you need an APDH to have an ACPSH), (2) limiting any discretionary match to 4%, (3) not recognizing deferrals in excess of 6% for any match, (4) not increasing the rate of match as deferrals increase, (5) not giving a better match to any HCE than what any NHCE would get at the same level of deferral, which I believe to be the basis for plans stating that no allocation conditions can be attached to the additional match. In other words, you can completely avoid allocation conditions on the additional match and still violate the last item I’ve listed in some other way, such as by having an HCE get more than an NHCE inadvertently because you have multiple matching contribution formulas. And I believe you can't overlook item #1) and say you have an ACP SH by observing items #2 through #5. In addition, not that anyone in this tread is claiming otherwise, I believe Regulation 1.401(m)-2 states that ADP safe harbor contributions are generally subject to the ACP test. However, most plans don’t need to do that ACP test for an ADP SH match because there is a special ACP testing rule that exempts the first 4% of compensation provided under the ADP safe harbor match (3.5% in the case of a QACA) from needing to conduct the ACP test. For example, if the plan has an ADP non-QACA safe harbor match and there is no ACP safe harbor, and if the ADP safe harbor match is 100% of the first 4% of compensation, then that ADP safe harbor match would not require an ACP test. In contrast, an ADP safe harbor formula of 100% of the first 5% would require an ACP test (in the absence of an ACP safe harbor). When an ACP test is required, I believe that administrator can choose to test either the full amount of the ADP safe harbor match or just the portion that exceeds the stated threshold (that rule is probably found elsewhere in the same regulation). The second sentence of the ACP test provisions of §1.401(m)-2(c)(2)(iv) states: (emphasis added) “In addition, a plan that satisfies the ADP safe harbor requirements of § 1.401(k)-3 for a plan year using qualified matching contributions but does not satisfy the ACP safe harbor requirements of section 401(m)(11) or 401(m)(12) for such plan year is permitted to apply this section by excluding matching contributions with respect to all eligible employees that do not exceed 4 percent (3 1/2 percent in the case of a plan that satisfies the ADP safe harbor under section 401(k)(13)) of each employee's compensation.” So unless the employer has committed to using an ACP safe harbor on the adoption agreement and that provision was both adopted and in effect at the start of the plan year, and all the other conditions are observed, I believe that (1) not only is the additional match always subject to the ACP test (no matter how many of the other conditions you satisfy), but also (2) maybe the ADP safe harbor contribution itself also needs to be ACP tested (depending on how generous it is).

I often don’t see these posts until days later, but since no one has responded to you and you are still following this, I will give you my two cents. (That’s another way of saying that if you respond, it might be a while before I see your response.) While it is true that many employers make a discretionary match every payroll, even they do not know how much the contribution is going to be if they are not using each payroll period as the matching computation period, since there will be a potential true-up owed after the end of the year. Many other employers won’t put any money in. Many participants won’t share in earlier contributions if they leave during the year and there is a requirement that they be employed on the last day of the year to get an allocation. As I understand it, document providers wanted a preserve having discretionary matching contribution allocations after the IRS said they were taking that away. So the providers proposed to the IRS that the IRS allow such discretionary allocations in the same way that nonelective contributions are made “definitely determinable” for a cross-tested plan, i.e., by putting the formula (or the arbitrary amounts) in writing. The IRS said “okay, but since this is a match, you need to tell the participants how much the employer decided to allocate for the plan year.” So many employers don’t know how much they are going to contribute until, say, the tax deadline after the end of the plan year. Under the regulations (and some documents), matching contributions can be as late as the end of the year subsequent to the deferrals for purposes of the ADP/ACP test. So in order to have a simple universal deadline, the IRS and at least one Provider agreed upon a notification to participants within 60 days of when the employer declared its last matching deposit for a particular year, at which point that became a standard that presumably ever provider now needs to meet. For PPA documents, participants didn’t’ get any notification of how much was allocated to them for the plan year as a discretionary nonelective contribution or as a matching contribution. So they’ve been making deferral decisions knowing only that there might be a match. Their SPD might say that the amount of their match will be a discretionary amount. This will be a step up. To put it another way, this is not a law where Congress thought about the best approach for informing participants as to make to make the best elective deferral decisions for a coming year. It is an IRS policy to make employers “definite” as to which deposits over the past dozen or more months they are associating with a particular year’s discretionary match, and having done that, how much of that overall amount is going to each participant (e.g., a formula, a list, whatever). Or, the employer can choose to make the allocation formula "definite" in the document, and then no notification is required.

Sorry for the delay, I wasn't on the forum yesterday. What you describe certainly looks good to me, but out of an abundance of caution, I would maybe do an electronic search for "prevailing" on the AA and make sure there is not any other election that seems pertinent. I am perhaps thinking of a document I saw where there was an election to use prevailing wage as an offset to any other nonelective contribution (which would include a QNEC or an ADPSH NEC) but there was a separate election to treat the PWC as a QNEC, and I advised a drafter to check that other box before applying the PWC toward a QNEC (or ADPSH NEC). Of course, that plan might have had something in the BPD that covered that generic "offset" election, but it just seemed like the best advice to check two AA options rather than just the "offset" one. If there is no such box for the plan you have, you should be good to go. I think it's pretty clear that the plan you describe is not the one I am thinking of so I am probably being unnecessarily thorough.

QNECS are subject to both 100% vesting and the distribution restrictions of Regulation 1.401(k)-1(d). To amplify Bri's response, the PW contributions also would need to be subject to the distribution restrictions in order to be treated as a QNEC (or as an ADP safe harbor contribution). That is why there is typically a specific election when drafting a document to treat such contributions as a QNEC (which would generally include treating them as an ADP safe harbor contribution).

I agree with CuseFan for both the notice and the amendment. I speculate that there is language in the plan document reflecting both of the following regulatory deadlines. Regulation 1.401(k)-3(e)(1) states the general rules that a safe harbor method must be in place (adopted) before the first day of the plan year (i.e. 12/31/20 for a 1/1/2021 safe harbor). The only exception to that general rule appears to be only for a plan that is “activating” a nonelective safe harbor method midyear, as follows. Regulation 1.401(k)-3(f)(1) states that the deadline for both the amendment and the supplemental notice regarding that amendment to change from the ADP test method (using current-year data) to the nonelective safe harbor method is 30 days prior to the end of the current plan year. Thus, subsection (e) applies if the plan that was already a safe harbor plan on the first day of the current plan year, and subsection (f) applies if the plan meets the requirements for converting into a nonelective safe harbor midyear and chooses to do so, in which case there is an earlier amendment deadline.

TPG: ESOP Guy says the unallocated shares need to reside in an ESOP. That renders most of what I have to say irrelevant. But pretend you really could merge the ESOP into a K plan, or that you allocate all the shares beforehand, or that you want to be prepared for the next time this situation comes up (which is why I am jumping in). I am not convinced you’ve begun to think about the other hassles that could be presented by your proposed merger, so let me talk you out of considering a merger for another reason besides the presence of unallocated shares. For anyone else reading this, this post deals only with a few general document-related issues I see as being potentially relevant. Those of you having no interest in documents are advised to bypass this post. My post arises from my experience with plan documents in general, not from having expertise with ESOPs. I am not an ESOP expert. The reason why I started this project was out of curiosity, and having finished it, I figured I would post it. My comments are offered only because all too often I have found that many transactions occur without first considering (1) the impact of the transaction on the document(s) and (2) the impact of the document(s) on the transaction. TPG, let me remind you that you used the word "case study." (Here is a case study.) The Documents If the K plan is not a preapproved plan, bear in mind that if the plan already has a determination letter, then the employer cannot obtain another determination letter except upon plan termination (IRS Revenue Procedure 2016-37). I would think twice before having a K plan with merged ESOP funds that cannot obtain a current determination letter covering the merger, even if both plans have pre-merger reliance. I would also think twice about having a merger like this be the subject of an initial determination letter request, since the entire history of both plans will be under the IRS microscope in that case. If the surviving K plan is not going to be a preapproved plan, there is nothing of interest in what follows. If the K plan is going to be a preapproved plan, you have a better potential path toward reliance, but I recommend that you proceed with an abundance of caution. You are likely to have 411(d)(6) protected benefits in the ESOP that are not compatible with any preapproved non-ESOP plan document, such as unallocated suspense accounts, the treatment and distribution of dividends, separate accounting and vesting for tax credit contributions, synthetic equity (probably not in this case, but just saying), and/or provisions relating to the forfeiture or distribution of company stock, etc. My concern is not related to the fact that ESOPs could not use preapproved documents prior to Cycle 3. My concern is that even if a provider has a Cycle 3 preapproved ESOP, that firm’s preapproved non-ESOP 401k plan most likely is not designed to accommodate any provision that is unique to an ESOP. [Background: My reading of Revenue Procedure 2017-41 is that the IRS did not allow providers to submit an “all-purpose” AA for a K plan’s Cycle 3 approval. Although that Procedure provides that providers could have provided an AA that has combined its K plan, PS-only plan, and non-target-MP plan features into a single AA, a provider could not combine ESOP features into that same AA. It appears to me that the cited Procedure required any plan with ESOP provisions to be the subject of a separate submission by the document provider. I suspect that the IRS would therefore have declined to review any non-ESOP K plan that contained any ESOP-unique provision, such as provisions containing details regarding a merger of an ESOP into the K plan. I am familiar with three Cycle 3 preapproved K documents, and none of them have ESOP-like provisions.] If the K plan is still in its PPA incarnation, and you conclude that modifications to that PPA preapproved document are necessary to accommodate protected benefits arising in the ESOP, then I think you should presume that you would potentially transform the preapproved K plan into an individually designed plan. I say that because no ESOP feature was permitted under the preapproved program for the PPA remedial amendment period. For that reason, I recommend that the K plan be restated to a Cycle 3 plan document prior to the merger, as then you could argue that modifications (if any) to any preapproved plan on account of an ESOP protected benefit is at least a modification that is within the subject matter that is now covered by the preapproved plan program. Even if you would lose reliance (in the absence of an IRS 5307 submission), you wouldn’t necessarily drop out of entitlement to six-year restatement cycles. And if you decided to submit the modifications to the IRS, you could start by using Form 5307 (though I would be prepared for a request from the IRS to re-submit using a Form 5300). That is why I stated earlier that there appears to be a potential path to formal and full reliance via a preapproved K document. You would not be asking for a determination letter for an ESOP, but for a modified preapproved Cycle 3 401K plan to protect 411(d)(6) benefits upon plan merger, and the fact that the modifications are ESOP-related is no longer going to throw you out of the preapproved plan program. You might also want to ask the provider of the K preapproved plan whether that provider thinks that their non-ESOP Cycle 3 K plan can be used as-is or modified for this purpose. However, I highly recommend that you provide a list of the specific ESOP-related protected benefits that you have identified as needing continued protection in the non-ESOP K plan as part of your inquiry to the document provider, especially if the provider does not provide a Cycle 3 ESOP (which would be a general indicator of how much experience they have with ESOPs). You want the provider to be clear as possible about what exactly you have in mind. Making such contact with the provider might be one of the first tasks to assign to the expert ESOP counsel that has been recommended by prior commentators in this thread.

BG5150, yes I meant keys / non-keys (thank you for the catch).

There are other possibilities: 1 - This is a boilerplate provision designed to make sure that non-keys get the TH allocation regardless of hours (i.e., a 416 fail-safe provision, possibly required for IRS approval), and SOME OTHER provision of the plan turns on/off TH contributions to key employees (such as an AA option). 2 - This is saying that the allocation to TH contribution to HCEs is conditional upon meeting an HoS requirement that is stated somewhere else (e.g., AA). 3 - Both of the above.

On the subject of fees for participants who either cannot be (or choose not to be) cashed-out, the IRS doesn't like to see a "significant detriment" on their ability to leave their money in the plan. See Revenue Ruling 96-47 and Revenue Ruling 2004-10. (That doesn't mean you can't charge reasonable fees.) There might be language in the plan document reflecting this guidance that you would need to observe if the plan's language is more conservative than what you think the law provides. The document might also fail to contain a mandatory cash-out at age 62/NRA even though a plan can be drafted to contain a mandatory cash-out at age 62/NRA. I have seen documents that continue to require consent even after the time described by IRC 401(a)(14), i.e., participants may choose to keep their money in the plan for life. For such a document, you would want to establish what preapproved cash-out provisions are available and which are not (as a matter of product design) before signing on the dotted line.

I drafted this before Kevin C's most recent post, so there might be inadvertent overlap. A reference to making a retroactive amendment to the beginning of the PY might arise from a mis-recall of a statement in Notice 2016-16. If that Notice was applicable to this situation, it would be worth looking at that condition, which requires that certain amendments that would otherwise be prohibited by that Notice can be "saved" if accompanied by a change in the computation period to use the entire PY as the computation period, retroactive to the first day of the PY. However, I don't think that provision applies because I don't think the Notice applies. Notice 2016-16 prohibits increasing a match midyear and certain other changes, but generally does not address a decrease in the ADP SH formula because that is covered by the regulation. So, the proposed decrease would be a reduction in the promised safe harbor contribution and would require a 30-day advance notice of the amendment reducing the match, and replacing the safe harbor method for the PY with a current-year ADP test requirement for the entire year. The match could be used in that test as a QMAC. (CARES has not changed any safe harbor rule in either the Notice or the regulation.) So YES you can "make the amendment," but you LOSE safe harbor status if you do so. That's a business decision. If this was not a safe harbor match, I believe such an amendment could be done, but a true-up would be owed for the portion of the PY during which the plan-year computation period was in effect (i.e., the true-up is a protected benefit). Someone who front-loaded their deferrals might still be unhappy, as it would not be a full-year true-up, but I don't see anything prohibiting such a reduction other than the anti-cutback rule, and that rule is satisfied IMHO if the true-up is made for just the portion of the PY that elapses through the later of the adoption date or the effective date of an amendment of this nature.

What I am about to say might be totally irrelevant to this conversation, but I feel that I can’t go wrong by mentioning it. Not being an actuary, I do not follow all the details to which “AFTAP” applies, but I do follow what needs (or doesn’t need) to be stated in plan documents (especially IRC 436), and I have been advised that there is a Section in CARES that affects the determination of the AFTAP for both the year ending in - and the year beginning in - 2020. Specifically, if I read it correctly, Section 3608(b) of CARES relieves employers from the need to recalculate the AFTAP, and perhaps that would be useful for the objective(s) being pursued in this thread. I have not seen much commentary on that Section of CARES. If this is irrelevant, please forgive the intrusion.

Michael: You do not have a SIMPLE plan unless the plan year is the "whole calendar year," so you don't have to worry about the SIMPLE rules for terminating the plan. Even if the plan did use the whole calendar year as the plan year as required by Regulation 1.401(k)-4(g), that regulation specifically states that a SIMPLE plan can be terminated only as of December 31. I suspect the document you are using requires the plan year to be the calendar year. Here is the regulation: "(g) Plan year. The plan year of a SIMPLE 401(k) plan must be the whole calendar year. Thus, in general, a SIMPLE 401(k) plan can be established only on January 1 and can be terminated only on December 31. However, in the case of an employer that did not previously maintain a SIMPLE 401(k) plan, the establishment date can be as late as October 1 (or later in the case of an employer that comes into existence after October 1 and establishes the SIMPLE 401(k) plan as soon as administratively feasible after the employer comes into existence). "

I fail to see the relevance of how it came to pass that a particular plan was assigned to my in-box. Maybe I agreed to take over a group of plans from another practitioner and there is a terminated plan within that group (whether or not I realize that at the time I agree)? Or, perhaps, I know this employer, and the employer is unhappy with the speed at which their current advisor or recordkeeper is progressing with the distribution of all assets, and prefers that I take over that process. Or, perhaps I am the new partner, and I am taking a spot check of an associate's work, and ask to have that particular account assigned to me, and it happens to be a plan with assets in it and a terminate date of one year ago. This was an off-the-cuff example in what was intended to be an informal and nice acknowledgement of some criticism that I earned, entered at the end of a long day. I decline to engage in this topic any further, so if I've left any loose ends open that defy reality, so be it. Thanks to all of you who responded on the subject matter of this thread.

Thank you all. Per a number of your suggestions, I will revise my advice (to people asking about the termination process) to include a statement to the effect that practitioners who terminate many plans believe that the standard stated by the Revenue Ruling is generally of concern only when there has been a delay in distributing all assets that is in "bad faith." ...Meaning, that a reasonable attempt to comply with the Ruling generally suffices. Having said that, if an advisor discovers that their new takeover plan still has assets in the trust fund despite the plan having terminated a year ago, I think it would be prudent to make sure that the presence of such assets is not the result of the employer not yet having taken any action in preparation for the distribution of assets, and, where that is the case, the advisor should make it clear that the employer needs to start taking decisive actions in getting assets out of the plan in order to demonstrate good faith compliance. On the other hand, if the actions taken by the employer or previous advisor to date have been at least as "fast" as would be "typical for the industry, then the "one year" rule should not be a cause for anxiety.

I am not trying to invalidate anyone’s experience or challenge anyone's beliefs.I appreciate becoming better informed that in your collective experience this rule generally is not a big deal. However, not long ago, I asked a nationally known consultant if he was aware of plans having been disqualified on account of distributions made on account of plan termination prior to a plan termination having been nullified pursuant to RR 89-87, and the response was “absolutely yes." I have not personally witnessed a plan disqualification because I do not assist with the termination process, I merely give general overviews of the process. When I am asked about this rule, I provide the words in the Ruling and urge the employer to seek counsel if there is any doubt as to the process and its timing. A true one-year safe harbor provision in the context of distributing assets would not say what that Ruling says, the Ruling would say instead that a complete distribution of assets within one year will be deemed to satisfy the facts-and-circumstances evaluation. If the word on the stop sign is “stop,” and you are going 1mph, you can get a ticket. I know because I got a ticket, as well as "points" on my license and a moving violation on my insurance. I still remember that event when I catch myself rolling through a stop sign decades later. Perhaps most practitioners have not had problems because they have provided good advice throughout the termination process, whereas employers in riskier situations and without good advice would have survived the termination audit had the employer abided by the letter of that Ruling. My only point is that it seems to me that the IRS absolutely has the ammunition available to cause pain if it finds that the employer has held on to the money for a reason that fails the IRS smell test. I see no purpose in the IRS including a "one-year" provision in that Ruling except to suggest that the IRS standard of review changes at that point, i.e., the IRS will become more reluctant to give employers the benefit of a doubt when evaluating the facts and circumstances. Pretend that 10 years have passed since the date of plan termination, the owner still has their money in the plan while everyone else has been paid out, and there has been no maintenance on the plan document during that period. Problem?

It may be true that many, many plans take more than one year to liquidate, but the questions is, how many of those plans get audited, and what is the fate of those that do get audited? Recall that at one time everyone thought you could choose to disregard the DOL regulation saying deferrals need to be deposited "as soon as possible" and many, may employers chose to wait (for a good reason, for a bad reason, or for no reason at all) until the 15th day of the month following the date of the deferral until the plan was audited and the penalties were imposed, and the word got around that the DOL was saying that "as soon as possible" really does mean "as soon as possible" and that the stated event of the 15th day of the following month was, essentially, an unsafe harbor that was more or less the DOL's way of saying that any plan that waited that long to make a deposit was not depositing deferrals as soon as possible. The DOL has now provided a safe harbor period of seven business days as its way of saying "this period is presumed to be as soon as possible in today's world." For a terminating plan, the one-year rule is in Revenue Ruling 89-87, and like the DOL's original guidance, the time period is characterized as an "outer limit" and not as a "safe harbor." Here's a small sample of the language in that Ruling: "Whether a distribution [of all assets] is made as soon as administratively feasible is to be determined under all the facts and circumstances of the given case but, generally, a distribution [of all assets] which is not completed within one year following the date of plan termination specified by the employer will be presumed not to have been made as soon as administratively feasible." The word "generally" doesn't say you "generally" have one year to distribute, the word "generally" means you generally fail to satisfy the timeliness requirement if the process takes that long. Whether or not you go past one year, the IRS can determine that you did not distribute assets as soon as administratively feasible and they can therefore nullify the plan's termination. If you go past one year, you will need to have very good documentation supporting facts and circumstances justifying a period in excess of one year because in that case the IRS starts negotiations with the proposition that you have presumptively failed to act timely. That is why I tell employers not to terminate a plan unless they are SURE they will (1) distribute all assets as soon as possible, and (2) be finished doing so within one year. If either proposition is uncertain, then they should wait to terminate until they are confident that they can meet both conditions. If a termination is nullified, it is true that some plans might not have any adverse consequences, but, for example, consider that any distribution that was made from the plan SOLELY on account of plan termination is no longer a "good" distribution because the distributable event of "plan termination" has been deemed by the IRS as never having occurred. Another example is in the context of keeping a document up to date: While you get to ignore changes in law taking effect after the date of plan termination, that remains true only for so long as the date of plan termination continues to be recognized as such by the IRS, i.e., you satisfy RR 89-87. If upon audit the IRS determines that the plan's termination never occurred, then the plan might instantly acquire missing amendments and restatements that suddenly and retroactively pop-up because the plan has suddenly and retroactively become an ongoing plan. Any delinquent pop-up amendments or restatements must be dealt with under EPCRS.

John correctly states the 401k rule regarding switching the "year" of testing. However, I believe that such an amendment is, nonetheless, a discretionary amendment that is subject to the IRS policy that any such change must be adopted within the year of its implementation (regardless of the subject matter), which in this case, I believe would be the testing year that has already ended on 12/31/18. To the extent that you agree that a change in ADP method is a discretionary amendment that should have been adopted by the end of 2018, I think that you would in some cases still be able to switch the testing year to the "current-year" method as part of a "Regulation 1.401(a)(4)-11(g)" corrective retroactive amendment, but one of the conditions for adopting an 11(g) amendment (regardless of its content) is that you need to adopt that amendment by the 11(g) deadline, which if I recall correctly is the 15th day of the 10th month after the plan year of correction. So I caution that it might be too late to adopt a discretionary or corrective amendment for calendar year 2018 in November of 2019 (except, perhaps, under EPCRS). You can certainly change to the current-year test method for 2019 by adopting an amendment by 12/31/19 .

The fact that a distribution is made to anyone at any time for any reason does not trigger an RMD. RMDs are triggered only by plan language implementing the 401(a)(9) regulations, and the only factors in determining the first calendar year for which an RMD must be paid are (1) age, (2) ownership status (in some cases, depending upon plan language), (3) employment status (in some cases, depending upon plan language), and death (when applicable). After you know the first calendar year for which an RMD must be distributed, then and only then do other issues come into play with regard to whom it must be paid and how much must be paid, and, in the case of death, the period over which all RMDs will be made to a beneficiary (for example, whether the beneficiary is subject to the 5-year method or the lifetime method, and whether or not the spousal rules are to be used when the lifetime method applies).You did not provide all this information. (I am not likely to engage in further analysis, this entire entry is just a roadmap). If 2019 is the first calendar year that an RMD is required (even if not due until 4/1/20), death does not change that result. When death is the event causing a calendar year to be the first year that an RMD is payable, then the first RMD will be paid or become payable to the participant, not the beneficiary. It will typically be treated as taxable income to the participant's under the final tax return for the participant (income with respect to a decedent). The participant's tax advisor will make that determination. For the first calendar year that any participant is subject to the RMD requirements (assume 2019 in this case, though not enough facts were presented to guarantee that result), the first money out of the plan for that year must be treated as an RMD until the RMD requirement is fulfilled for that calendar year (2019). That would be true even for a living participant. For example, Q&A #7 of Regulation 1.402(c)-2 states that any amount required to be distributed as an RMD for an RMD "distribution calendar year" (2019 in your example) is to be treated as an RMD. That means that even if the participant had not died, the first money out in 2019 would be treated as an RMD for the participant for 2019 (it would not, for example, have been eligible for rollover treatment by the participant). Death does not change that result. Assuming that 2019 is the first calendar year for which an RMD is due, then any amount still owed on the 2019 RMD at the time of death (perhaps all of the RMD) is still payable to the participant for 2019 (i.e., reported as income on the participant's tax return for the year of death). Any additional funds payable in 2019 would be paid to the beneficiary in accordance with the plan's distributions on account of a participant's death, but that does not mean there is a 2020 spousal RMD required (though that might be true). The plan document should state that beneficiaries are not subject to the death RMD provisions until the first calendar year following the year of the participant's death. To know what happens for years subsequent to the year of death, you need to evaluate the plan's language as to which Death-RMD distribution method is applicable (the 5-year rule or the spousal lifetime method). In some cases, the document might give the beneficiary the choice of method to use to determine when Death-RMDs must be paid and what the amount will be. In other words, a spouse does not necessarily receive lifetime distributions using the spousal-beneficiary RMD lifetime method. That could occur, for example, if the plan specifies that all Death-RMDs are to be determined using the 5-year rule. Under that 5-year rule, no funds should remain in the plan after that 5th year. To summarize, the amount payable to the participant for 2019 as an RMD is calculated in accordance with the plan's language for RMDs to participants (up through the year of death). The amount payable for subsequent years is determined in accordance with Death-RMD plan provisions and are payable to the designated beneficiary. If the entire amount is paid to the beneficiary in 2019 (other than the 2019 RMD payable to the participant), then the only RMD for 2019 is the RMD payable to the participant for the year of the participant's death (2019). This all assumes that 2019 is the first calendar year for which an RMD is required. For example, if a non-owner participant turned age 70 1/2 in 2015 and retired in 2017, and the plan uses the post-SBJPA required beginning date for all participants for all funds, then 2017 would have been the first calendar year for which an RMD was required, and the RMD rules upon death would not apply for a participant who dies in 2019, since RMDs would already have become payable prior to the calendar year of death. in that case, you would then use the plan's rules for determining RMDs for calendar years following participants' death occurring after RMDs have already commenced being paid to participants.

Mr. Bagwell, don't forget, though, that the 5% gateway contribution is based on 415 comp (not 414(s) compensation) for the plan year (not the limitation year), but most plans give you the ability to use only post-entry 415 compensation. Such an option is specifically authorized by the gateway regulation, and is, to my knowledge, the sole context when anyone gets to use post-entry 415 compensation (...a delightful fact to know and to amaze your friends and family with during the trivia portion of your next cocktail party). So, you need to include bonuses, etc. in determining the 5% gateway contribution. In contrast, you can leave 414(s)-compliant PS compensation as-is if you are passing the ratio (three-to-one) gateway requirement.