Tom Poje

good grief. people must think I am old enough to know these things. I have never even seen one of these animals. anyway, someone is working on restating a 401(k) plan sponsored by the govt which has been granfathered forever. does such plan require ADP language or were they exempt from testing as well (since govt plans usually have all type of special rules, I have no clue) thanks!

Rosemary: it used to be when you imputed disparity, ees with SSRA age 67(and comp < TWB) would have .65 added to their E-Bar. now the system adds .654. I know, maybe the extra .004 is considered insignificant, but sometimes even that much can cause a plan to pass/fail. I am not denying that the regs say to use .654 if an ees NRA is 65 and 1 month. the regs are clear on that. however, consider one ee born on 1/1/00 and another born on 1/2/00 them the retirement 'age' for one is 1/1/65 and the other is 2/1/65. now I have a non uniform retirment age. and once I have a non uniform age, I have to use a common testing age of 65. (not 65 and 1 month) (I actually disagree further from the standpoint most documents define a NRA (65) and a NRD (1st of the month following). I don't want to use NRD, I want to use NRA. Again, I could be wrong in my reading of the regs and you are to use NRD, but then I am stuck with the non uniform age rule and I am back to having to use 65 as described above. I have no problem using 65 / 5, mainly because once an employee gets past age 60 or so, the E-Bar is low enough that it isn't going to help much one way or the other.

the problem isn't with the ADP test. the problem is coverage. how many NHCEs are 'eligible to participate' (even though they worked 1 year and are age 21. zippo. so plan fails coverage. I would again ask, why are the people considered HCEs? if it is just comp and not ownership, then the top paid group election might be your only way out. and if you never go above 9 total employees then only the owner might end up being an HCE

fascinating. this might work. or at least it has possibilities. see if I messed up in the logic here.... you indicated the following 1 owner 2 hces ?????????????but are they 5 nhces use top paid group election. (assuming the 2 'hces' are by comp only, this would work) 8 total employees * 20% = 1.6 you are allowed to round up or down. round down to one HCE. owner makes the most $ so he is the hce. oh, but he is not in the plan. so now your hces become nhces . plan will pass! suppose one of the ees makes more than the owner. well then, now I guess for 410(B) you have 1 /6 1 /2 = 33% ratio % nhce % = 75% or safe harbor of 38.75% dang that wont work. well, then the owner has to make more.

This was not raised, so you had to make me look it up. According to the ERISA Outline Book, deferrals are enough to satisfy the 'substantial and recurring' rule pertaining to profit sharing plans, and therefore you would not have to vest 100% upon discontinuence of the profit sharing contribution. The book also notes the IRS has not really ruled on this, it is merely an interpretation of the regs. Interesting, I would have guessed the other way.

for #3, ACP is on all matching contributions. KIP - catch ups could occur on other plan imposed limits - e.g. HCEs limited to 5% deferral rate. also if plan fails ADP test, that is considered a limit, and so anything above that could be considered catch-up. (for the over 50 crowd) in that case, if match is not allowed on catch ups, you would have to forfeit any match related $ as well.

with svc pack 8 at 6.0 a change was made to the nondiscrim calculation involving permitted disparity. first, I will say two things. I have been wrong before, and I will be wrong again, I am merely stating why I disagree. 2nd, I didn't see any indication of this change in the svc pack 8 notes, but I could easily have missed it. It caught me by surprise, but I didn't read every single thing in the notes. anyway, with svc pack 8, if you impute disparity, the factors are now adjusted for months as well. Reg 1.401(l)-3(e)(3) clearly states that this is required when using an age other than SSRA, so there seems to be a basis for this argument. however, 1.401(a)(4)-12 definitions says (1) If the plan provides the same uniform normal retirement age for all employees, the employee's testing age is the employee's normal retirement age under the plan. now, for example using the Corbel document (though I am sure other document are similar) I have Normal retirement age = 65, normal retirement date = 1st of month coincident or next following. so now the system no longer tests on normal retirement age, but rather normal retirement date. and that is the crux of the matter. but, for the sake of argument, lets suppose that I am to use normal retirement date as my testing age rather than my normal retirement age.In that case, I have someone whose mormal retirement is 65, another whose normal retirement will be 65 and 1 month. I no linger have a uniform reirement age. therefore, I must use 1.401(a)(4)-12 testing age (3) if the plan does not provide a uniform retirement NRA, the employees testing age is 65. I am back to testing age 65 rather than 65 and 1 month as a testing age. and 1.401(a)(4)-7©(4)(iii)(B) says the .75 percent adjustment, pursuant to 1.401(l)-3(e) using as the age at which benefits commence the lesser of age 65 or the employees testing age. Note this is the lesser of 65 or the testing age, so using 65 and 1 month makes little sense to me. actually, in this case, if one does not have a uniform retirement age, then one could test using SSRA, as permitted under the SPBJA changes (though one has to be careful there because of other testing concerns) if you agree with the above argument, then you will have to code your plans 'date of event' rather than 'first of the month following'. while this might not quite coincide with your document, it is what I have to do for now. In a DC plan this date is almost insignificant anyway.

I have only been vaguely following this thread, but lets see... 2 plans exist, one a DB and one a DC (401(k)) which to make things more complicated had a QNEC. The general test is basically the 410(B) test except each HCE is treated as a separate plan. There is an option whereby you do not aggregate the plans, but if that option was chosen you can not cross test either plan, and in this discussion, it appears that was not a choice. plans were aggregated, and it sounds like at least one HCE failed the ratio % test, so the average benefits test is being used. this consists of two parts. 1. average benefits % test - for this everything is aggregated. 2. rate group test - this should only include the db and profit sharing contribution. this test is an amounts test. the defrrals and match would not be included because their amounts test consist of the ADP and ACP test, and you don't have to test them a second time. The QNEC throws a wrinkle into things (especially if it went to HCEs as well) For purposes of satisfying rate groups, the test must be satisfied when the profit sharing contributions are tested alone AND ALSO when the QNECs are added with the profit sharing contribution. (See ERISA Outline Book 9.38 -2001 edition) In this case, I would also include the DB 'cuz I have to aggregate the db and dc plan.

Andy- we are using Relius Only run a couple of age weighted plans, and I just finsihed one in which I actually have someone past age 65. The system used the same factor as if they were age 65, which is fine because that is the way the document is written. Based on that, I would assume if the document called for a different method, I would have to adjust the number of 'points' the system calculated. as for the E-Bar calculation, it used the current age on the participant. Because the current age produces a smaller APR, the E-Bar was slightly larger than everyone else, which, as we mentioned would normally be considered discriminatory. and whatever that reg cite says I am ok, cuz I can do this.(if I used the APR for age 65 then the E-Bar would have been the same,) as for imputing disparity on ees past 65, Iwould have to check and see what the system is doing. I can only recall one or two cross tested plans I handle with an employee past age 65, and by that time I don't need to impute disparity.

ERISA Outline Book (9.84) (2001 edition) 'if employee is beyond testing age, employee's testing age is his current year (unless requirements of 1.401(a)(4)-3 (f)(3)(i) relating to accruals after normal retirement are not satisfied) of course that assumes the plan has a uniform retirement age. Suppose a plan doesn't have a uniform retirement age. then you would have to use age 65 or SSRA. ..... software we use also tests on current age, but the latest version uses normal retirement DATE if you impute disparity, rather than normal retirement AGE. Thus, for example, the rank and file increase by .654 rather than .65.e.g. normal retirement age = 65, normal retirement date = 1st of month following. Reg cite 1.410(l)-3(e)(3) is cited, but I have looked at this and this appears to determining what disparity applies to formulas (and I guess, in particular, safe harbor formulas) not testing. A few years ago I posed the very question to Larry Deutsch (of course, he could have always changed his answer by now) "If this is a DC plan, then under the definition of testing age, the testing age is the normal retirement age under the plan, and the normal retirement date has no impact."

nothing wrong with it, but... and this one would be hard to prove one way or the other BRF, effectively available. certainly its current available, but if the NHCEs can't afford to defer more than 3% than one might argue it is not effectively available. howevr, if that happened, you probably wouldn't be able to pass ACP testing anyway.

so, if I understand things correctly: I write my document to say I am going to use prior year testing. I should be able to put into the document language that says cap limit on hces of 2 points plus prior year NHCEs. at this point I have a plan document limitation rather than an administartive limit, and therefore I am ok. Maybe I would never do that, but just want to make sure if logically one could do that. thanks Mike!

ah, Mr. Mike: would a plan limit on HCEs the lesser of 2 times or 2 points plus last years NHCE rate be definitely determinable? obviously at the beginning of the year that amount might be overshot because it hasn't been 'determined' yet, but at some time it will be. and if hce is over that amount and age 50 would that constitute a legitiment catch up?

I am not so sure on that one. 1.401(a)(4)-12 (definitions) Testing age (1) if plan has uniform retirement age for all employees, then testing age is normal retirement age. thus I could have 65/5 and that is uniform. so an ee hired age 62 has NRA = 67 and one hired 68 has NRA 73. now (4) if an employee is beyond the testing age otherwise determined for the employee under paragraphs (1) through (3) of this definition, the employee's testing age is the employee's current age. so someone who works past NRA would appear to fall under #4. In fact, if that wasn't true there would be no reason for #4. How is it possible to get beyond NRA for testing purposes? well, my brain might be on the fritz, I can't think of an example at the moment. Reg Section 1.401(a)(4)-8(B). says a plan doesn't fail this paragraph merely because allocations are made at the same rate for employees who are older than their tesing (determined w/o regard to the current-age rule in paragraph 4 of the definitions of testing age. consider an age weighted plan that uses the same 'factor' for an ee at age 65 and older. Thus a 65 year old will receive the same allocation as a 68 year old with the same comp. But if you were to look at E-Bars using 'current age, the 68 year old would have a lower E-Bar which appears to be discriminatory. This paragraph says no, it doesn't fail to satisfy this. at least that is how I read it. I am currently working on an age weighted plan that fits this description. what a coincidence.

good question, it doesn't hurt to verify stuff. you did not indicate that the ACP test failed, so I will assume it passed. In that case refund of deferrals causes related match money to be forfeited (not refunded). since it is a forfeiture, anyone entitled to receive forfeitures can receive, including the HCE. it ceases to be a match, its almosy like an extra ps contrib to the plan. (If plan had failed ACP test, then you are allowed to refund ACP first, then refund ADP, so sometimes the match $ goes back to the HCE)

I think, therefore I am a fool.... my guess would be as follows: there is a lot more going on than you may think!learn something new everyday! You said, 'For allocation purposes you have the short year on B, so contributions to B get limited.' Actually, I am not sure if you really have a short year! See 5.15 of ERISA Outline Book 2001 edition. The termination of the plan does NOT end the limitation year for 415 purposes, so a short period is not created merely because the plan has terminated. If allocations are made in the year the plan terminated, the normal section 415 limits applies to the allocation. then it goes on to say that even if the final allocation is 'as of' a particular date, that is merely the allocation date in a 12 month limitation period. Then it goes on to say that what could create a short year is when the final distribution of assets occurs. if on November 30 then you have created 11 month year. so it looks like the issue really depends on when the assets are actually moved/merged. all that aside, my understandong for 415 purposes, you consider the limitation years of all plans involved . Thus you still have 12 months of comp to consider from the period 4/1 - 3/31. and therefore shouldnt have a problem anyway.

interesting. you are going to end up with a bunch of NHCEs between 3 and 12 months who will be includabe and not benefiting. the ADP test can exercise the revised rule where you treat these nhce as 'otherwise excludabe' and treat similarly employed hces as includable. however, coverage doesn't have that rule, so you will have a bunch of nhces treated as includable and not benefitng and a bunch of hces as includable and benefiting. well, maybe...remember, HCE definition is 80,000 (indexed) in prior year, so you shouldn't have many hces who would enter the plan after 3 months - not enough comp in prior year. the only hces who would be a problem would be owners. but of course this is in the future. It sounds like you are looking at establishing a new plan, and therefore you could have a bunch of hces the first year. Without seeing numbers, I think after the first year you wouldn't have a problem because you could test otherwise excludable option, and after the first year you probably wont create any more hces with only 3 months of service. I suppose the initial plan year you could include all nhces with 3 months and then switch to 12 months, but good grief. why make pensions so tough.

that is what the regs say, they get ignored. (I am not sure I would call that a 'problem', it should help the test)

yes, you are correct. the new rules are for plan years beginning after 12/31/2001, so in the case of a new plan 1/1/2001 - 12/31/2001 to determine if you need top heavy for 2001 you use old rules. then you turn around and use new rules to see if a top heavy is needed in 2002 as well. and as a result you could end up with 'former' key ees.

doesn't appear to be any problems with the svc packs themselves. svc pack 8 has a top heavy distribution program that is suppossed to be run (or at least should be run), so read your notes that come with the program. its buried in there. we have encountered one issue. had an owner that quit in 2000. he is still an owner with a balance, system still treats him as key, though he has not performed svc for a year. I guess it is one of those quirks - system must be looking at ownership first, rather than term date.

they count in the year made, not deposited. as for your comments on daily, a company like Manulife holds off reporting the last quarter of data until the December deferrals come in, so they get counted in the correct year- assuming of course they are timely deposited. I don't know what other companies do...most of the plans we process like this are through Manulife

I will assume your plan has immediate eligibility to defer. Therefore, you have the following possibilities: 1. Include all on one test 2. Perform 2 tests - one consisting of all hces and only those nhces who completed 1 yr svc and are age 21. the other test consists of only nhces (otherwise excludable) so you automatically pass 3. perform 2 tests. All ees who completed 1 yr svc and age 21, and all other participants who didnt. slightly different than #2 because it might include HCEs. Whatever you do for the ADP test, you must also do for the 410(B) test, except option 2 doesnt exist under 410(B)

these are my nondiscrimination notes. If I get 'rented' out to talk at a user group I have used these - if Relius gives a talk on nondiscrim, they use the same basic notes with some modifications (they borrowed the original ones) I have not updated them recently, but the basics are correct. as for the spreadsheet, please contact Fred Payne - he is on the Benefits Link. He wrote the sheet, I think it does a great job - he has it even checking for the gateway - (of course it doesn't do everything) but I don't feel its within my right to post his spreadsheet.

Attend one of your user groups meetings /and/or suggest that as a topic. other than that, the responsibility still falls on your shoulders as well. options that I can think of: using 'otherwise excludable' (statutory exclusions) - if you use for ADP test you must, I repeat MUST use for coverage and vice-versa. nondiscrim testing - SSRA as testing age. I would NOT use this option, unless you have a document that specifically calls for it and you have a determination letter. Testing age is supposed to be plans NRA. group accrual rates - I would only use cautiously. while the regs permit it, the regs also require that it does not significantly favor HCEs. In other words, the way the system currently handles things is weak. you really need to know what you are doing in this area. impute disparity - no problem, ok to use, but be careful if part of your contribution consists of safe-harbor nonelectives. you can not impute disparity on these. don't ask me how to run these. well, actually, someone else on the Benefits Link board created an excel sheet to handles these, and his spreadsheet appears to be working fine. safe harbor/unsafe harbor - if plan is top-heavy, you have to run unsafe harbor for purposes of coverage and the values you put on the schedule T. then you have to run safe harbor to see if plan still passes coverage. if it doesn't, you have to run a(4) testing to see if plan passes.

I think that one is pretty much up to you. I would think if its a 401(k) plan you would want to include the ADP report anyway. If plan is cross tested, most people can't make heads or tails out of the reports anyway, so those reports probably only confuse people. since we do back office work on cross tested, we generally send the nondiscrim reports because our client has hired us for that specific purpose.