AndyH

Here's how I understand it. A QNEC could be to all participants or only to selected participants. If it is to all participants, and allocated in the same manner as a ps contribution, there is no difference. It is uniform and must be tested only if there are other employer contributions that must be general tested. If it is a separate contribution to select participants, it must pass 401(a)(4), and if it goes to any HCE, it must be general tested because it is not uniform and does not meet the "no HCE benefits" exemption. A SHNEC is by definition uniform. It therefore passes a(4), but the regulations say that you may use it for 401(a)(4) testing purposes. So, in your plan, Harvey, if a uniform percent of the ps contribution is treated as a QNEC, there is no difference between this and a ps contribution for testing purposes. Only if it is non-uniform does it differ. Does this make any sense?

Yes, Richard, you seem to have it right. Each component plan must pass r/p by 70%, i.e. just a normal coverage test-all nonexcludables are in the test. Then, each rate group needs to pass the midpoint. But, in determining each rate group, I understand that the other component plan is treated as non excludable and non-benefitting, but Merlin is saying you may ignore the other component group as if it did not exist. On that point, I am uncertain. That is not my understanding, but I could be wrong. I have been told that Merlin's approach is incorrect. After reading his cite, however, I am now unsure. I have in the past done it both ways and have not found a large difference in the result because each component plan has to be 70% + to begin with, and if they are both near 70%, the results will be nearly identical under either approach.

Yes, sounds like they missed the boat. But, it is true that the ps was permissively aggregated. It's just that the k component should have been backed out of the test. A couple of years ago I saw the reverse situation in a takeover plan. The actuary who designed the plan submitted a test as part of an FDL application for a DC/DB combo including DB, PS&K. The k deferrals were incorrectly included in the NCT component of the a(4) test, and the IRS reviewer caught it and required that the test be corrected! The actuary agreed that he had made a mistake. P.S. By QNEC, do you mean safe harbor nonelective? If so, that is in the test absolutely because it's actually a ps contribution with certain characteristics.

Well, now I'm not sure. Merlin's way is what I used to think, then I was told otherwise, and I'm pretty sure I've seen textbook examples using both components in the ratio calculations. Perhaps the key to Merlin's cite is "may"? Maybe it can be done either way? Let's see how others interpret this.

Richard, lots of questions! I've had to focus on all of this stuff due to a takeover DB/DC combo, so I'll tell you what I've learned: 1. You can put any of the people in the test in either component plan. This can be done randomly or selectively and can be changed year to year. So, yes, you can put the excluded HCE in either plan. 2. With respect to 410(B), if your component selection is arbitrary, the NCT conditions (objective business reason/reasonable classification) standard is not met. Therefore you must use the ratio percentage test and each component plan must have a r/p of 70%, counting all (non excludable) NHCEs and HCEs in the denominator and only those benefitting in that component in the numerator. Hope this helps. If I missed a question, please advise. Sounds like you suspected these answers yourself.

No, then cannot be permissively aggregated. And with regard to QNECS, you must test by both including and excluding them.

Tom, so in the case of an age weighed PS plan, it sounds like your system would calculate a lower contribution for someone age 66 than age 65, which is what ours does (ASC), to our displeasure. And, it sounds to me like your system does not handle permitted disparity correctly for someone past 65. It seems to me that if someone past 65 gets a lower contribution, then the increased pd factor should be used, but if the person gets the same contribution as a 65 year old, then the age 65 pd factor would be appropriate.

Yes, Tom, you've hit it on the head exactly. Thanks to you and the others for the comments; all were helpful. I've got an existing takeover PS plan where the owner is 69 and wants to look at redesign. NRA is 65 and he's been the plan for many years so 65+5 P doesn't apply. So, NRA is 65, attained age is 69. My interpretation is that under the regs you could treat the person as either 69 or 65. If you treat him as 65, you're ok, because of the sentence Tom quoted. If you treat him as 69, he gets a lower contribution than a 65 year old-potential age discrimination issue (for non-owners). But, if you then test the age 65 allocation using age 69 assumptions, you will fail, so testing age needs to be treated as allocation age. So, I was curious what others, and other systems, do. Our packaged system uses the post 65 factors; we override them to use age 65 factors. I'm trying to see if that is still correct and appropriate. I was curious what others, and their systems, do. Plus, this presents similar issues for a new comparability plan. Any other views or comments on this would be welcome. Thanks again.

No, you don't have any discrimination-type issues because each plan satisfies the ratio/percentage coverage test of 410(B) by itself. Each plan would need to cover no more than 70% of the NHCEs, perhaps less if you used the Average Benefits test. I think the 15% limit is determined by taking the compensation of any participant eligible for either plan.

Thanks for the response, Belgarath. That makes sense, except what doesn't make sense to me is the IRS wants the plans amended for GUST first. And, we don't have approval on some of the volume submitter plan documents, so we can't amend many plans for GUST (properly). Many plans won't be amended till late 2002 or even 2003. Yet, EGTRRA is in effect, so if you follow the IRS' guidance you've created a cutback?

If somebody works past NRA in a cross tested plan, what testing age are people using for the annuity rate conversion, NRA or attained age?

If it is determined that a profit sharing plan does not incorporate the 401(a)(17) comp limit by reference, and the plan has no last day requirement, and it is amended mid-year to incorporate EGTRRA's $200,000 comp limit, is this a cutback for other employees? If so, is the IRS contemplating any relief under 411(d)(6)? This question is discussed in more detail on Corbel's website here: http://www.corbel.com/news/pensionupdatesd...tail.asp?ID=157 Opinions?

I'm not certain of this, it's just my read of it and others should comment. 1.410(B)-(5)(d)(5)(iii) seems to say you must be consistent among employees and plans, but then (e)(2) seems to say that you can separately determine the employee benefit percentages using different methods (as itemized in that section) provided that the differences don't result in a significantly higher average benefits percentage. So, that's a start to this discussion. Good question. P.S. maybe I read too much into the question. Clearly you can do 1 or 2; my question is whether you could do the annual method for the dc and the accrued to date for the db, or vice versa.

Takeover cross tested profit sharing plan under IRS audit. Discovered in 2002 that a person was left out of the profit sharing contribution incorrectly for 1999. Plan is not a safe harbor; general test not done (being done now). Company will contribute amount that employee should have received. Should the omitted contribution be included in the general test for 1999? I'm inclined to say no, unless IRS sanctions such treatment as part of audit closing agreement, because it's too late to be considered an annual addition for that year, and too late to be considered a 401(a)(4)(11)(g) corrective amendment. But, what if not under audit? Need to go through VCO? Other opinions? Generally, when do non-required contributions have to be made to be included in general test?

From The ERISA Outline Book (1998 version in front of me) Page 9.16 "In fact, the plan must be able to satisfy 401(a)(4) when QNCs are combined with other nonelective contributions, and also when the other nonelective contributions are tested separately from the QNCs. See Treas. Reg. 1.401(a)(4)-1(B)(2), 1.401(k)-1(B)(5) and 1.401(m)-1(B)(5)" I just happened to research this the other day.

It sounds to me like you are correct, and that Harry O's comments seem to support that as well, provided that each plan provides uniform BRF's within each plan.

Plan comp can be used for the 3/1 gateway only if it satisfies 414(s). I see plans that use comp that doesn't satisfy 414(s) but they're tested on 414(s) for 401(a)(4) and ADP, so the gateway issue is another thing to watch for in such cases. Also, I think that a plan that has a QNEC must be tested both with and without the QNEC, but if it goes only to non-HCEs, then the only test that matters is with QNECs out since the other would be better. Also, I think employee aftertax is out, and aren't ESOP additions out as well?

I'm not certain of the answer, but since you haven't had a response, the following may be helpful: 1.401(a)(4)(1)©(6) seems to say former employees are tested separately. Then, you have 1.401(a)(4)-10 which deals exclusively with nondiscrimination with respect to former employees.

If all of us who have read this thread as of right now chip in 8.4 cents each, we'll be able to buy the survey!

Tom, this is the accrued to date, or account balance method, where the ebar is based uon the projection of the balance/years benefitting/average compensation, and under this method, his average comp and years benefitting don't change, he retains (almost) the same ebar as last year's test reflected, and since he is young, that is a high EBAR. And, true, bringing all of them in might not have a net positive effect; but we've got 2 HCEs near the top and the NHCEs who used to have higher EBARS are gone, so we're exploring options. And, regarding failsafe, almost all of our plans had 401(a)(4) cross testing failsafe language approved. With the GUST restatements, however, they're requiring us to remove it. The language increases the youngest NHCE below a failing group incrementally until the EBAR exceeds that of the HCE in the failing rate group, then goes to the next youngest NHCE, etc.

Thanks for the cite, Merlin. That will do it. And your idea about a corrective amendment is a good one, except that the plan already has failsafe language. But, we're trying to avoid using that because it happens to produce undesirable results in this particular case.

Very interesting ideas. I wonder how one would put such a thing in a document. I frequently see flat dollar minimums in discussions, which in theory is great, but again how does that fit in a document? I asked Joan Gucciardi that this October and she said it's most common application of a flat dollar minimum would be in a corrective amendment, but obviously that doesn't work for nonvested terminations.

It's only the employer, so the answer is 3%.

We're testing a cross tested plan using the accrued to date method. The test fails on this and other (e.g. annual method) bases. The plan requires 1000 hours and employment on the last day to receive a contribution. There's a participant who would have an high EBAR and help the test (because of the accrued to date method), but he didn't receive a contribution because he terminated with under 501 hours, so he's statutorily excludable. Can we include such people (in the test)? What are the rules for including (in the test) statutorily excludable people who are in fact excluded? If we brought him in, does this mean we have to bring in people who didn't meet the plan's age and service requirements as 0%s?

What I was trying to say that it can be argued that he had the following choices: 1. Elect a lump sum and satisfy the bonding/escrow requirements. 2. Defer payment until he can elect an unrestricted lump sum 3. Choose another option under the plan. If he started collecting lifetime payments, is that not a life annuity election? If not, on what basis under the terms of the plan did he receive payment? If no basis, is that failing to follow the terms of the plan? Some have suggested adding an additional optional form of payment, some type of modified cash refund, or as an alternative adding an option which remains in force only until such time as the restriction no longer applies. But, then, what happens if the person dies in the interim? Presumably, it would depend upon the temporary option elected, i.e. if life annuity, no future payment. Doing this would be valid under the terms of the plan and would not be a changed election. If such a temporary election were made, I would think that the subsequent lump sum would be the actuarial equivalent of whatever "temporary" form of benefit had been elected, i.e. (annuity rate at age of payment) x (initial pension). This is a little off the wall, but what other alternatives exist?