AndyH

You got it, Rick. Go forth and teach restructuring.

Thanks for your comments, Mike. I presume that by "free pass" you are referring to the multiple formula exemption under 1.401(a)(4)(2)(vi)? I'm having trouble determining if that would help or not. Obviously that contemplated a minimum formula of 3% of pay, but would that logically extend to 3% less a company match? Would that be uniform? And even if so, it would seem that there still is an issue for a Key employee who gets a top heavy minimum greater than any Non-Key? p.s. In looking at this more, I think there is a definite a(4) issue whenever a plan uses the match towards top heavy and an HCE gets a top heavy discretionary contribution, for example in the case where the HCE does not defer. And if there is no other discretionary contribution, the test will most likely fail. Anybody disagree?

Yes you did, Blinky. Only kidding. Guess it's pick on Blinky day. I've had those.

If a match is used to partially satisfy the top heavy minimum in a non-safe harbor 401(k) plan, is the additional employer contribution to get to 3% subject to testing under 401(a)(4)? How would the non-match top heavy minimum be treated if the discretionary contribution were subject to general testing for other reasons, e.g. cross tested formula? For example, assume a plan's discretionary formula must be general tested, and it happens to be top heavy and one person gets a match of 2%, but is ineligible for the discretionary contribution. Presumably the employer must contribute an additional 1% to satisfy top heavy testing. Would this 1% be treated as a discretionary contribution for purposes of the NCT portion of the general test. Presumably yes. Second example: The plan has no discretionary contribution other than a top heavy minimum Assume an HCE gets a 2% match which must be supplemented by a 1% discretionary contribution. Is this subject to general testing under 401(a)(4)? because it is non-uniform?

Right. Rick, you're jumping ahead of yourself. First look at the plan before restructuring. Does it pass 410(B) as a whole? It it's eligibility criteria reasonable? That's the 410(B) part. Then step 2, which is 401(a)(4), is where you might consider restructuring, and then, yes, the reasonableness might be an issue requiring each restructured component plan to pass the 70% ratio/percentage test.

The 3% requirement never existed on account of cross testing. That is bad information.

Must a Non-ERISA 403(B) plan have an SPD? A document? (I think the answer is no in each case) Plan is with TIA/CREF. They prepared a "specimen" SPD which is stamped "For review by legal counsel" They also send the employer checklists about amending for GUST and EGTRRA. What should an employer proving employees with non-ERISA 403(B) do with these? Would making decisions and revising/issuing an SPD cross the line into ERISA status? Opinions please.

Let me guess, now, a medical practice? Those bosses are always unhappy when employees get contributions! Yeah, I admit it; the word "dispensed" had something to do with my guess.

The most economical correction might be to provide sufficient $$$ to get all the rate groups to 70% to avoid the ABPT.

No, the restructuring is a testing method which does not need to be in the plan document. But make sure you aren't working with a plan document that DOES tell you how to test. Some bad ones do, even though they are not required to, so they can "cramp your style".

Yes, you need to define separate classes, unless your plan document has some sort of "catchall" provision that will automatically provide that NHCEs get the gateway if nothing else, or you could do this by 1.401(a)(4)(11)(g) corrective amendment. Yes, you are right about the sensitive demographics, but as long as you have 2 HCEs and you are treating one like a regular employee, this plan structure can often last. I wouldn't put a last day or 1,000 hour requirement on this, though. And no, I can think of no other tricks for the spouse. Good luck!

Rick, your second paragraph sounds like you are on the right track. Your optimum design would be to max out the Doctor, give the youngest NHCE what is needed (8% sounds reasonable), give the second NHCE only the gateway (5%), and the spouse either 5% or exclude her by class. This is the way the numbers work best, but of course your plan document must allow this. And, to reiterate, there is no need to exclude the spouse; limiting her to the same contribution rate as the older NHCE will pass testing by using restructuring as detailed earlier.

Somehow we are not connecting. You simply give each NHCE at least 5%, then test as component plans like you planned to do. But test one on a benefits basis, and one on a contributions basis (this is not cross testing). You should be able to avoid ABPT, and therefore the deferrals don't enter the picture. Anybody care to help me out here? I don't seem to be getting the idea across. Maybe it's the contributions based testing: It's simple: Spouse 3% NHCE 5% One rate group covering 1/2 NHCEs and 1/2 HCEs. 100% coverage. ABPT not required if plan A has the same result by using cross testing (which this component plan is not doing).

1. Restructuring does not affect the type of money being tested, only the bodies. 2. That is not correct. Anybody (NHCE) in the NCT test as a whole (whether or not in a separate component plan) is subject to the gateway. B must get the 5% also if the Doctor gets more than 9%. Said another way, restructuring cannot be used to avoid the gateway applying to someone.

I am assuming that the spouse and second NHCE must be included in the test. If so, it sounded like testing them all together will result in one or more rate groups less than 70%. That would require you to proceed to the average benefits percentage test which you might fail due to the spouse's deferral. But if you test separately, as you suggest, testing one on a benefit and the other on contributions, both component plans may have ratiio percentages of 100%, in which case you do not have to proceed to the average benefits percentage test, so the deferrals would not be an issue. There are two issues unanswered here, however. How do you satisfy the gateway requirement if one of the NHCE gets only 3%. Are you limiting the Doctor to 9%? Second, on what basis are you excluding the spouse and the second NHCE from the discretionary contribution? If the eligibility provisions are not "reasonable", then the ABPT is not even available to you.

Then, no, you cannot do that. But what you can do is test component Plan B on a contributions basis for the NCT. The deferrals are never part of the NCT test. If the ("cross tested" profit sharing)and the 3% SHNEC allocation as a percent of pay is the same or less for the spouse as for the NHCE in her component plan, you will have 100% coverage 1/2 / 1/2 in that rate group and the average benefits percentage test (where the deferrals must be included) will not be necessary. You must always include nonelective contributions in the NCT test and never include deferrals or matches. And you must always include all of the above in the average benefits percentage test. Restructuring into component plans cannot alter this reality. But it can help you pass NCT and perhaps avoid the average benefits percentage test.

1. Yes, unless. 2. No, unless. Unless is unless he could be excluded from the test on the basis that he was an "otherwise excludable employee", i.e. under 21, never worked Year of Service (including 1,000 hours), in plan because it has eligibility more liberal than allowed. Such people could be tested separately, and if there are no such HCEs, they automatically pass.

Your mechanics are correct if you only have a 3% SHNEC and deferrals, but you actually would not even have to test it if everybody gets 3%! If there is a supplemental profit sharing contribution then it must be included in each NCT. If the spouse gets the same as the NHCEs and the EBAR of the youngest NHCE exceeds that of the Doctor then your scheme works as well. Is this what you had in mind? You didn't describe anything else subject to "cross testing". It's not completely clear what your scenario is. If you are thinking that component testing can allow you to exclude certain contributions, then that is NOT ok.

You can do as you suggest, or the reverse if you wanted to. You just need to do it the same way for all allocations in each test. So, both tests could be done on allocations or benefits basis, or one could be done one way and the other the other way. You are right about the impact of deferrals. They can have a huge effect, and often times the result makes little sense.

Except, as I understand it, the plan administrator cannot foreclose on it without the occurence of a distributable event, because that would be tantamount to a prohibited in-service distribution. This is why loans in DB plans are rare. The contingency rules are messy.

Thank you, Everett, got it. I'm not sure this is completely clear, but I do agree that it makes sense.

Offset issue-anything new, any cites or anything in print? I just looked at a takeover MP/PS situation where a MP plan consisting exclusively of prevailing wages is, according to the terms of the profit sharing plan, used as a direct offset to profit sharing allocations. There are some people with small non-prevailing wages who would, but for the offset, receive a profit sharing contribution. According to Everett's comments, this is not acceptable. Is there anything in print that I can point to? Does anyone agree or disagree?

There are no prototype cross tested plans, although pseudo-prototypes are often used and submitted as volume submitters or individually designed.

Each of your rate groups must pass coverage, either by ratio percentage (70%+) or Average benefits. The rate group test would included only employer non-elective, generally speaking, but this would include a 3% SHNEC as well. If any of your NC rate groups are under 70%, you must proceed to average benefits, and in that test you include all contributions that could be permissively aggregated, plus k, m, and esop allocations.

Half the solution is identifying the question. Could you clarify please. How can a safe harbor nonelective contribution have disparity between NHCEs and HCEs, unless of course HCEs do not get the contribution, in which case, what is being tested? Sorry, but your last two sentences don't make sense to me. Maybe the question is: If the rate group test is passed, do you need to do average benefits? If that is the question, the answer is yes unless all of your rate groups are at least 70%. And if you must do average benefits and you fail, you fail the entire 401(a)(4) general test.