Mike Preston

I find it hard to believe that the ABPT will be less than 70%. If it is, as I suspect, 70% or higher, then following Tom's excellent example each of the plans should pass 410b. Hope B is a new plan, because otherwise you have a nearly 1,000 participant population where this is first being analyzed? :shakes head:

Show numbers of NHCE's and HCE's in each. Do the combined plans pass the ABT? If so, what is the safe-harbor percentage and the unsafe-harbor percentage? Right now, I'm not accepting your premise that the 401(m) coverage fails. BTW, you don't test non-existent benefits, so 401(m) doesn't involve Plan A and Plan B isn't involved in testing PS.

Cheaper to set up a second PSP. But if no contributions are "intended" the plan will crush itself into disqualified, anyway. Tell plan sponsor that a 3% of pay contribution every third year to protect 6M in qualified money is inexpensive.

Instead try a more targeted approach, such as "The match for 2014 shall not be less than the following: Participant A - A%; Participant B - B%...etc."

If there is a right of first refusal for C's stock that runs in favor of the other individuals or Company#1 and it is not part of a bonefide reciprocal arrangement as described in 1563 you might have to exclude C's shares. Check the bylaws of Company#2. Same thing might even apply to Company#1. Without knowing the specifics I don't think you have quite enough information. Suggest they punt to an ERISA attorney!

I'd run it by an ERISA attorney, but why not an amendment now that conforms the document to the match actually made?

This is an interesting issue. I think the terms of the plan control, so in the absence of a specific exclusion related to those not employed on or after the adoption date I think the person is in the plan for all legitimate purposes (safe harbor, etc.). But one can't defer until the later of effective or adoption date so 401(k) deferrals (and match) are not included in those purposes. But 401(a)(4) certainly would be.

Properly? Seriously, you report what actually happened.

Seriously?

I think one or more of the writeups for EPCRS make it clear that the purpose of the QNEC (originally 50% now 25%) is to make up not only for the opportunity lost relative to the ability to defer, but also the missed earnings, right? So, while the participant can get back to "402(g) max" by electing a higher percentage than what would have been necessary if the election would have been effective for the entire year, he/she can never get back the lost earnings. Here is the cite from 2015-28: "The correction principle underlying the 50% make-up corrective contribution was that corrective contributions should make up for the value of the lost opportunity for an employee to have a portion of his or her compensation accumulate with earnings tax deferred in the future assuming the participant would not have the opportunity to increase elective deferrals in later periods to make up for missed contributions AND EARNINGS that would have accumulated until retirement." You would think that the correction, in this case, wouldn't need a separate earnings component due to the quoted paragraph. But I can see that there is a slight disconnect and so the earnings being referenced are separate issues.

(1) Unless I'm reading it wrong, I think you have a situation where the error is not discovered within 3 months and therefore the correction is a 25% QNEC with respect to the missed deferrals for the entire period of failure. So, yes, there is a missed deferral for 2015 measured from 1/1/2015 through the date that deferrals are enabled. The missed deferral is calculated with reference to the ADP % for the group and is independent of the actual election the participant ends up making. So you end up with the same QNEC whether the participant elects no deferrals or max deferrals. (2) Yes, in the case of a current year missed deferral that requires correction where the plan uses current year testing you can't calculate the missed deferral opportunity until the year is over and, if necessary, the failed ADP test is fixed (since you have to correct everything else before calculating the missed deferral), so you can't make the QNEC until the year is over. (3) Yes, deferrals can't actually begin until the participant has filled out an election, or has been given an election form and decided not to act on it, in which case you should be able to use a cutoff date in the determination of the calculation of the missed deferral opportunity. I think the 45 days should be measured from the same cut-off date, or, if an election is made, the payroll date where such election would first be implemented (even if zero). I'm only about 1/2 way through my notes so it wouldn't surprise me to learn that there is soft guidance supplementing EPCRS that clarifies one or more of your excellent questions.

Self correction via plan amendment is only available in very limited, enumerated circumstances. I don't think yours is listed. Do you?

It just is.

Another questionable result essentially caused by the failure of the plan sponsor to adopt a plan where every participant is in their own allocation group. Strong letter to follow.

I assume the family trusts are grantor trusts, right? If so, the beneficiaries are deemed to own their actuarial interest (translation: much lower than 22.5%), while the Grantor is deemed to own what is not owned by the beneficiaries. I don't think the way you have described double attribution is the way it works. Double attribution is only precluded if both parts of the attribution are because of family attribution. A beneficiary's interest is not because of attribution. Spouse of older children more likely to be HCE/Key than younger kids.

The implication in the first post of Mellish is that there is a carve out. To the extent something is carved out it is done so permanently (barring a revised QDRO trying to muddy the waters). But then there is a reference to the fact that the QDRO calls for initially making said carve out at some point in the future (when P reaches earliest retirement age I would guess) which argues for the fact that there is no existing carve out until P reaches said early retirement age. So, if there is no carve out then there is nothing taken from the P at the point in time that B dies and therefore B loses all rights that might have been determined at some point in the future. So, which is it? A carve out or not?

You are correct, A-org and FSO are arbitrary. You have to run through the analysis twice. In this case I, like K2retire, will end up administering one plan based on a combined census as a single ASG.

As long as you are asking the IRS for a compliance statement, I see no reason why the IRS shouldn't allow it.

PT?

Way back when this issue was of major concern to those who sponsored money purchase plans there was a time when the existing guidance, sparse as it was, pointed towards the need to reverse the payment. I don't know when it happened or what the specific citation was, but somewhere along the line the conventional wisdom changed to a view something like the following: If the distribution was acceptable absent the re-hire, then it can proceed notwithstanding the re-hire. I don't have time to look up the cite, but I'm fairly confident of the conclusion.

As others have pointed out, any language you are relying on to indicate some sort of corrective action is required to pass coverage is irrelevant since with the 3% SHNEC, you pass coverage as far as employer contributions go. As others have also pointed out, you are confusing the term you should be using ("allocation rate") with another term ("rate group"). Unfortunately for you, the term "rate group" has a specific meaning when discussing how testing works and that specific meaning has nothing to do with what you are trying to discuss. So, the simple solution to help clarify things is to simply stop using the term "rate group" when you should be using the term "allocation rate". So, besides a tongue-lashing for using the wrong terminology and another tongue-lashing, IMO, for saddling a client with a prototype document which arbitrarily limits the number of allocation rates available for no good reason other than the firm designing the plan cares not one whit about what is best for the client (strong message to follow), let's assume that after passing coverage you find that you fail 401(a)(4). [bTW, with NHCE B receiving a total ER contribution of 31% of pay I find it hard to believe you would fail 401(a)(4) based on the census you have shared, but I will assume you do just to round out the discussion.] You then must dig into the structure of the document to see whether the language allows a special gateway allocation in addition (or as an override) to the excruciatingly limiting language regarding allocation rates. If it does, you have one allocation rate notwithstanding the disparate monetary allocations. If it doesn't, your view prevails. To reiterate, all of this, of course, is a result of 401(a)(4) testing, not 410(b) (coverage) testing. Your document, in addition to the coverage safe-harbor language, might also have 401(a)(4) safe-harbor language. If it does, you must follow it.

Point whoever it is that is claiming you don't know what you are talking about to: http://www.dol.gov/ebsa/regs/fab2008-4.html When their eyes glaze over tell them that the general requirement for bonding is allowed to be waived by various sections of the regs (see 2580.312-31 for the specific exemption applicable to insurance companies). There is also a nuanced issue that you won't see directly addressed in a regulation, but the above mentioned FAB2008-4 contains a reference in Q14. And that is that a bond is generally required to protect "plan assets" and the exemptions in the regs are applicable only if the contributions to the plan are made from the general assets of the plan sponsor without those amounts first becoming plan assets. There are other regs that essentially say that monies withheld from paychecks (think 401(k) deferrals) morph into plan assets in a very short period of time (at this writing I think the safe-harbor is after 7 working days). Hence, even if a 401(k) plan were to be solely invested in insurance company products and would otherwise qualify for an exemption, the employer would be subject to the bonding requirement should any deferrals be forwarded "late". To my knowledge I've never seen the DOL try to go after an employer finding itself in this position.

Boy, there must be some special reason he wants to operate a business through a trust. It was my understanding that a trust in that circumstance would pay much higher taxes. Guy going to jail?

If the plan was first terminated and then the owners were able to request a distribution earlier than the non-owners there is definitely a benefits, rights and features violation with or without the fee issue. Probably a failure to follow plan terms violation, too. Somebody should tell the owners that the money they no doubt rolled is subject to immediate taxation because the plan could very well be disqualified due to their actions. Then head on over to your friendly neighborhood ERISA lawyer for some guidance on what they should do to fix this. I know I'd suggest that they took distributions of in excess of what they were entitled to and suggest that they follow the EPCRS procedure to recover that excess. The amount of that excess would be determined as their pro-rata share of the expenses charged along with a share, if any, of any trust value decline which took place after their distributions and before the non-owners were paid. But that is just me.

How would you suggest one can check the SS master death registry?