Tom Poje

the answer is Yes or No depending on how things are handled. 1. the annual safe harbor notice should indicate that there is an age 21/1 year of service requirement. (And don't get confused on the 1 year svc - that is only for eligibility, once you have worked that, you are eligible for the safe harbor regardless of hours) 2. you now run 2 tests. those that met statutory requirements and those that didn't. but wait, under the safe harbor rules, those that met the requirements automatically pass. but wait, the otherwise exlcudable also pass as there are no HCEs. so technically no test needs to be performed. now, what happens if you have an HCE that falls into this category. An 18 year old that makes $100,000 (Ha) or an owner's kid. Remember, the regs from a few years ago allow you to treat these employees as having met the requirements for ADP/ACP testing purposes only. (But NOt for 410(B) - it is a strange scenario!)

it sounds like you have the following: company matches during the year, on a per payroll basis, for example 50 cents on the dollar, up to 5%. Employee A did not defer the first half of the year, but then deferred 10% in the second half of the year. Obviously, this would require a true up at the end of the year, so that the average over the whole year averages out correctly. Employee B deferes during the year, and receives the match. However, he either fails to work the required hours, or terminates thereby making him 'ineligble' for the match, if that is what the document says. This is a recipie for disaster. Not only must $ be removed fron his account but gains/losses must be calculated. And if its individual investments the headaches grows. We ''gently, kindly, and lovingly" tell clients to 'not do this'. Austin is correct, you may have problems if this was not clearly communicated on participant statements etc. If you are going to match every payroll period, I would have all match $ put in a pooled account, and at the end of the year alloacte it to all participants. After all, technically, at least if there is a last day employment requirement, no one has earned it

there is no requirement to send a list of salaries. I have heard of some strange suggestions before, but never that. you are required to privide a statement of a participant's balance to all terminees. (Strange to say, this is not actually required of active participants unless they request them, but why you wouldn't is beyond me) you are also required to provide the SAR to all (whoever completes the 5500 should provide that to you) In a nut shell, this statement says: Hi! there are $$$$$$$$$$$$ in the plan. there were $$$$ in gains and -$$$$ losses, etc. The participant can then look at his individual statement and see that his $ is paltry compared to the big scheme of things. I think most people simply throw the SAR away as it doesn't make a big difference one way or the other. As for tracking those people down - well, there is not much you can do if you do not have there address. But that is another issue, how do you pay people out if you don't have their address....

and since the person turned 70 1/2 in calendar year 2002, the minimum distribution is due 4/1/2003, though some prefer to pay by 12/31/2002 to avoid 2 distributions in the same year. The latest regulations, at least based the data I read last week say: for 2002, one may follow the 1987 proposed regs, the 2001 proposed regs, or the final regs!!!!!! If using the final regs, you have to amend the plan (of course) identifying the date chosen to follow the final rules. The final regs allow you to use the new uniform lifetime table (e.g. age 70 = 27.4, age 71 = 26.5 for the factor in determining the amount of distribution) http://www.cigna.com/professional/pdf/2002...Regulations.pdf

see IRS notice 2000-3 question 10 Is a plan required to provide safe harbor matching or nonelective contributions to participants who have not yet attained age 21 and completed a year of service if the plan uses one of the safe harbor nethods? Answer: (My abbreviated form) NO! the catch is that you must apply 410(B) otherwise excludable testing. that should not be a problem as it is doubtful there would be any HCEs in the one group. Anyone with less than a year can not be an HCE due to comp, so the group would always pass testing unless you had a 5% owner in the group.

Owning more than 1% and making over 150,000 makes one key. By definition, a 5% owner is someone who makes more than 5%, so if you owned exactly 5% and made less than 150,000 you would not be key, unless you were an officer making 130,000

Ray- no one came out and said why ACP problems arise, but: 1.usually only HCEs put away after-tax dollars. these must be included in the test 2. since everyone has the ability to put away after tax $, everyone must be included in the ACP test. (e.g. if your plan has a last day requirement for the match, anyone who terminates is excluded from the ACP test) you could end up with a bunch more zeroes on the ACP test with after-tax contributions.

one other possibility a lot of people miss. you indicated plan had high turnover. by chance were any of these HCEs with less than 500 hours? remember, it is optional whether to include/exclude all terminees with less than 500 hours. many softwares simply assume it is to your benefit to exlude such people. but if there are hces in the group it will help to include all terminees in the testing

set the 415 limit to $300 rather than $40,000, and run 100% of comp formula

at one of the ASPA conferences the IRS voice an opinino (and an opinion only) that you do not include such people. Of course, the question was about HCEs, and if you include them that would only help the test, so that might have had bearing on their answer

your choice. see 1.402(g)-1(e)(6) and/or 1.401(k)-1(f)(5) the example from the ERISA Outline Book is as follows: plan year 12/31/2001 excess contribution = 1000 excess deferral = 600 1. distribute 600 first excess deferral the 400 as excess contribution 2. distribute 1000 as excess contribution. the excess deferral is considered corrected at this point. however, for tax purposes the plan must report 600 as an excess deferral, and the remainder as excess contribution.

Algebra 101 where x = salary y = 415 contrib = .25 x but x = 11,000 + y x = 11,000 + .25 x .75x = 11,000 x = 14,666.67 now, solve for next year's compensation!

interesting scenario my logic would look at it this way: under the regs to determine if the plan's formula is safe harbor, you would treat anyone who receives a top heavy as includable and not benefiting. since no one else receives a non elective contribution, plan 'passes' determination that the formula is safe harbor- no a(4) test needed...but then for purposes of the Schedule T (Coverage) you treat ees who receive top heavy only as includable and benefiting. Now you have one HCE receiving a nonelective and zippo NHCEs, so coverage is failed.

I would think merely stating 'effective date of the 401(k) is 10/1' would be sufficient, since you only have to satisfy the safe harbor from the effective date of the 401(k) arrangement. since the terminees were not there from the effective they would get nothing. It would be sort of like 'use date of participation'

without looking it up, my initial guess would be no, union ees would/could be excluded. a safe harbor 401(k) plan must either give a shnec or a shmac to pass the ADP test. I don't see why there would be a difference if the method chosen is nonelective or match. you are still disaggregating (union/nonunion peple) the ADP test in either case.

you are correct. the same applies to ownership, it must be more than 5%

if HCEs comp > 200,000 then excluding deferrals will help testing as the HCE comp does not change. On the negative side, it penalizes the lowly NHCE from deferring as it results in a smaller contribution. sort of like using match to satisfy top heavy. why bother deferring when your co-worker gets something for nothing by not deferring.

they are excluded from testing. (The reason for the return is not due to test failure - the reason for the return is a violation of something the employee wasn't entitled to at all.)

you can try this one - maybe modify to suit your needs. it is set up for small plans

R Butler: hopefully then I will see you (and others) at the conference! the boss (Lorraine Dorsa) has a booth this year so I should be easy to track down.

the new top heavy rules pertaining to EGTRRA are not to take place until plan years beginning after 12/31/2001. your plan year is before that date, so the old rules still apply, as I understand things.

That sounds okay, my comments were in reference to language contained in a Corbel document (and I am sure others) definition of Year of Service: "...An employee who is credited with the required Hours of service in both the initial computation period and the Plan year that includes the anniversary of the date on which the employee is first performed an hour of service, shall be credited with two Years of Service for purposes of eligibility to particiapte." note that the above does not refer to vesting, only years of service for eligibility.

Watch out how you define terms in your document! be careful if initial computation period is anniversary and then shifts to plan year! e.g. an ee hired 12/1/00 gets 1 year credit from 12/1/00 - 11/30/01, and then would also receive credit from 1/1/2001 to 12/31/2001. thus entering 1/1/2002 after just 13 months!

Lori: an interesting comment, that the associates are showing up in the rate group test. That implies the following possibilities 1. the software is doing it wrong and doesn't offer other possibilities 2. Unintentionaly, the software is not being used properly 3. a combination of the two you did not mention the name of the software, and obviously this is not a forum to get into a discussion on the merits of one software or another - I feel confident enough to think most have their pluses and minuses. I can make comments on the system I currently use, and then probably offend someone along the way, but that is not the intention. I hope the following comments are not taken as such. with Relius, you would either have to code up two plans under the same employer. when performing the average benefits percentage test, you would include both plans in testing. when performing the rate group test, you would only include one plan, and then fill in the grid for non excludable/non benefiting employees. This would give you a correct rate group report. However, at this point (Because you are only looking at one plan, and just threw a bunch of zeroes on the test), the plan might 'fail' the avg ben % test, so the 'overall' report would be in error since you actually passed that test. (It is interesting, on the old DOS system you had a toggle in which you could indicate plan passes average ben % test to get around this problem. Anyway, I note the above, just in case that is the software you are using.

75,000 50,000 78,353 52,235 81,720 54,480 85,485 56,990 90,803 60,535 93,518 62,345 96,368 64,245 99,000 66,000 100,000 66,000 100,000 66,000 80,000 80,000 here is 1987 - 1997 there are two sets of data, because you had a comp limit and a 20% top paid group limit. and if my fading memory serves me correctly you had to llok at both current and prior year. 1997 started the new rules, thank goodness they made things easier.