Tom Poje

Ha. That question brings back memories. A number of years ago someone asked me a question along similar lines. I was trying to explain that you had to limit the compensation used and the individual insisted that a lawyer had designed the plan, he knew what he was doing, and you didn't have to limit the compensation. I pointed out that the 5500 C asks if the compensation was limited to the 401(a)(17) limit . (I said this goes back a few years) The response was that 'this year' the 5500 R was being used and it doesn't ask that question. Can't argue with logic like that!

I think of it this way. The plan could have had 1 year wait, entry date 1st of the plan year or six months after meeting requirements. If the plan had those requirements, what would happen in your testing? Well, all those people you are worried about would never have entered anyway. So, in reality, yes, it seems like you are favoring the HCEs, but on the other hand, you have let in a bunch of NHCEs earlier than you needed to, so you get the advantage of testing them separately. a reward for being more generous than you needed to be on your eligibility.

I vote for ignoring plan's entry dates and using the maximum possible exclusion, which is 1st day of plan year or 6 months after completing 1 yr service or age 21. At ASPA conference 2002 the following question was asked of the IRS. (one has to remember that answer do not necessarily represent an official position of the IRS, merely the opinions held by the agents at the conference.) Q #32 A 401(k) plan povides for immediate eligibility and uses otherwise excludable rule to perform its ADP testing. In determining which employees belong in a particular group must the employer use the plan's entry dates or may it use the maximum entry dates permissible under Code section 410(a)(4)? The employer COULD use the maximum entry dates. (emphasis mine) so even in their eyes it is not a clear issue.

A mechanic who worked out of his home had a dog named Mace. Mace had a bad habit of eating all the grass in the mechanic's lawn, so the mechanic had to keep Mace inside. The grass eventually became overgrown. One day the mechanic was working on a car in his backyard and dropped his wrench losing it in the tall grass. He couldn't find it for the life of him, so he decided to call it a day. That night, Mace escaped from the house and ate all the grass in the backyard. The next morning the mechanic went outside and saw his wrench glinting in the sunlight. Realizing what had happened he looked up to the heavens and proclaimed... "A grazing Mace, how sweet the hound, that saved a wrench for me!"

so the conclusion would be(?) Yes, QNECs are included in determining whether the gateway minimum is satisfied. (see also Federal Register preamble which says the gateway is determined under 1.401(a)(4)-2© and 'a plan is permitted to cross test once it passes a gateway minimum) QNECs are used in the avg ben % test, though of course, no imputing disparity. When cross testing you must run two rate group tests. One with the QNEC and one without QNEC. (See example as cited by DFerrare) In the example that started this discussion, the plan fails rate group testing without the QNEC and would need to take corrective action. What is interesting, I suppose, is that when testing without the QNEC, an ees actual allocation rate did not satisfy the gateway minimum. But that doesn't matter, since the regs simply say once you pass the gateway minimum you can now cross test.

ok, suppose an ee receives a 1% QNEC only. he is not eligible for the profit sharing, but the plan is cross tested. and HCEs received 12%. now, do you bump him up to the gateway minimum. it looks like he received a nonelective contribution. if so, how much? if you say 3% then you are saying you include the QNEC in the gateway. If you say 4%, then it is as if he didn't receive a nonelective since you gave him a full 1/3 of the HCE. Of course, the biggest problem is that the gateway minimum rules was written after all the others.

If I didn't have the plans set up as one plan and in divisions as was suggested, I might first think about exporting the tests to excel and then combining then. Famous last words, 'it should work'. Or simply take the results from each plan and combine them on a separate sheet. e.g. Plan 1 HCE total 32.50% for 5 HCEs Plan 2 HCE total 23.3% for 4 HCEs combined 55.80% for 9 HCEs = whatever % etc. if you had a 2 plans that were merged you would have to do the same type of thing for testing anyway

I base my comments on ERISA Outline Book, in particular the examples on 9.18 and 9.19.(2003 edition) I read this to say the reason for testing the nonelective bot with the QNEC and without the QNEC is to determine whether the plan satisfies a safe harbor only formula (with and/or without the QNEC). Or, put another way, to determine if you would even need to test under 401(a)(4) the final example speaks of a scenario where the plans would not be safe harbor and therefore must test using the rate group method. This would seem to fit the case of a cross tested plan. I could of course be way off on this one.

my understanding would be yes. QNECs are still NECs (nonelectives) with a Q (Qualified). That means that they are simply 100% vested, can be used in ADP or ACP test, subject to deferral distribution rules. But they are still nonelectives. And as such, can be used to satisfy gateway minimum since the gateway is satisfied by nonelectives. In cross testing you can not impute disparity on the QNEC, however.

I have to verify if the report still works. I haven't run the report since upgrading to from 7 to 8. maybe after I try running it I can post it.

no, ee is an HCE, but what I was saying is instead of using half year comp in testing, use full year, thus ee drops from 20% to 10%. see also ERISA Outline Book 11.66 (2003 edition) Section 414(s) comp may be limited to period ee is eligible but it is not mandatory. (But then treat NHCEs the same) again, the document decides. if it says use date of participation then too bad.

what does your document say? for example the old Corbel document would say 414(s) comp is from date of participation the new language says any definition that satisfies 414(s) 414(s) is testing comp not allocation comp. there is no requirement to use comp from date of participation as long as you are consistent with everybody. You can even vary from year to year, but if you are using prior year testing you still have to be consistent. Again, it depends on your document

the only thing that matters is if an NHCE receives a nonelective contribution (profit sharing, forfeiture, QNEC or SNEC) if so, and the plan is cross tested, then they must be bumped up to the gateway minimum. It does not matter whether they are active or terminated. if they get a little, then they get more. The exception, of course, would be a plan with broadly available bands, such as an age weighted plan.

note: the correction for the 2003 PA manual can be found at: http://www.aspa.org/archivepages/education...pa1a-errata.htm so, the error has been corrected!

I looked at the PA-1A (2002) I have - it matches what you say, except on the table for plan year end it also says 'estimate' regardless, this is sure to be confusing. maybe it was corrected if there is a 2003 edition out there????

see 1.410(b)-(6) it lists all the excludables. terminees are listed under (f) {I know that 'cuz I have to cover it at the ASPA talk} note that you don't have an hours restriction for deferrals, so terminees would not be excluded from the 401(k) portion of 410(b) testing I suspect you have no last day requirement for match (or 0 hours) , therefore you would not exclude the terms from testing since they would receive a match if they deferred. Remeber, for coverage you have 3 tests 401k 401m nonelective you have to pass all 3. I suspect your attorney wants to know because the propoasal was as follows: More than one division, but only want one division to have a plan. Therefore, new document required, a nonstandardized one- as pensions in paradise indicated. therefore, need to know the population so testing can be looked at. hmmm. maybe that is giving the attorney too much credit???? generally 401k is 100%, but in a controlled group scenarion , if only one group benefits it wouldn't be. you might pass the nonelective (profit sharing portion) but the plan might have been nasty in regards to the match and so you would still have to test. A standardized plan is a no-brainer. All nonhighlies benefit all 3 tests. But you pay a price for that option

I'm not giving any clues as to the defined benefit. and the song doesn't contain the lines "Will you still need me, when I'm 64" though I had thought about that as a pension song.

min partic - DC do not have to do exclude from testing - those fail age /svc, union ees, non resident aliens optional to exclude terms < 500 hrs (only if they do not benefit (e.g. you can't exclude them from 401(k) testing) you have to be consistent - you can't include HCE terminess and exclude NHCE terminees. coverage (410(b) doesn't care how mush, er, much you get, just wheter you get with a standardized document you have to pass coverage, so it sounds like you are ok. ......... yes, it sounds like attorney is doing coverage. if you are talking about testing, it seems to me you wouldn't want to include any terminees. assuming a calendar year, everyone has accrued 500 hours (or so assume) and can't be excluded. however, if you are looking at data to get an idea for next year, it wouldn't make sense to include any terminees.

no preview. a few years ago I did a version of Old MacDonald. I might repeat that one as an extra. The original version of Yesterday, way back before the Beatles became popular, back when they were working in the pension industry (you didn't know that?)- well anyway, they recorded a song encouraging people to defer. Hmmm, I think they did one about defined benefit plans, though most of that song has been lost. I 'discovered' one verse of it - I'll keep looking, maybe I'll discover the rest of it.

Imputing disparity is an option, you never have to use it. However, if you have more than one plan, you can only adjust the E-Bars on one plan only. By the way, try the following. One plan, intergated at 5.7%. lets suppose everyone receives 6.5% + integartion at TWB. Try calculating the E-Bars with disparity using the allocation method. Everyone should end up with the same E-Bar. Ultimately that is why an integrated plan is considered safe harbor, if you ran the test imputing disparity it passes.

if the contribution was required (DB or money purchase,etc) then you would include the receivable. In the case of profit sharing contributions, generally the amount of 'receivable' is not included. At the ASPA conference last year, the IRS indicated it did not have a problem if the amount was included. Comments from such meetings do not necessarily reflect the actual position of the Treasury. This particular comment was a big change from the way things have been done in the past. Most, if not all people, have been told never to include such amounts. There are various old thread that address this issue.

ah ha. my first response was someone is crazy, but then ..... If my feeble mind recalls correctly, the annual addition limit is an end of year value. (As opposed to compensation limit which is beginning of year value) If you have a non calendar year plan that begins in 2003 and ends in 2004, then you would use the limit for 2004. And based on the current facts and circumstances, that amount will be 41,000. Perhaps that is what is being referred to - non calendar plan years.. I had posted a spreadsheet somewhere about a month ago that showed how to calculate the value.

gee Andy, I don't think anybody has ever referred to me as a 'regular'. I've heard other terms, such as 'oddball' and 'wierd', etc. Oh, and hopefully I will be in 'disguise' for the talk, never could do anything normal!

This could depend on your plan document. If the situation becomes a controlled group make sure the document doesn't 'suck the other members into the plan. I believe that would happen if the plan is a standardized document. Regardless, assuming a document that doesn't bring the other companies in.... As for testing coverage if a controlled group arises and rears its ugly head, the employees of the 'other' companies not participating would be treated as includable and not benefiting for purposes of coverage. Unless you have some HCEs in this group the plan would probably fail coverage. ADP test would be unaffected

1.410(b)-7(e) says for the avg ben % test you aggregate all plans that COULD have been permissively aggregated. Thus all paln are normally aggregated for this test, and this test alone. Therefore, if you can pass the nondiscrim classification test unaggregated, then you do not not have a DB/DC combo, and thus no DB/DC combo gateway applies since that only applies to aggregated plans. Thus, I think you are ok even if you test the Avg Ben % test on an unaggregated basis as well. I think that rule only applies in regards to that particular test. I suppose an easier question would be, does the plan pass avg ben % test if you combine the plans? Remember, combining them for this test does not necessarily mean you are aggregating the plans for other testing purposes.