Tom Poje

I think you are stuck. even though the deposit of match is not until 'now', since the formula is fixed, you could have run the ADP/ACP test before the 2 1/2 month deadline and known the plan would fail. that wouldreally have been strange, because you would 'fix' the problem before the current match is made. I guess no gains on the match. and it would be impossible to do that if it was the first year of the plan. Of course all this assumes that unlike the clients I have, all your clients provide you data early in the year. I suppose under 1.401(m)-2(a)(4)(iii) © if you waited until after the end of the current year to make the deposit (more than 12 months after plan year end) then there is no ACP test and thus no excess aggregate contribution. you then test the match under a(4) [see 1.401(m)-2(a)(5)]. of course, the deduction then shifts to the following year. strange rules. I am not sure about your comments on "matching contributions that will cause the ACP failure are based purely upon the already reduced deferrals" If I am matching $ for $ and the person defers $1000 then the match would be $1000 according to the terms of the document, even if the person had to take $50 for excess contributions - or at least how I see it. Rate of match determined after corrections for failed test. 1.401(a)(4)-4(e)(3)(iii)(G). this says to determine any rate of match after applying the rules of 1.401(k)-2(b)(1)(i) and 1.401(m)-2(b)(1)(i). now, one of the rules under 1.401(k)..... is that you would forfeit the related match. so which comes first, the chicken or the egg. Do you run the DAP test first, calculate the correction and the forfeiture and then run the ACP test, or can you run the ACP test first, fix that, and then run the ADP test. the results are different, and I believe the IRS has informally indicated either method is possible.

I would have said 1.401(m)-2(b)(4)(i) but I find it hard to pinpoint the exact reg cite at times myself. (Though since its a match, it would have to be (m) rather than (k) Since you only indicated ACP failure, are you implying the ADP test passes, and if so, you don't have anything you can 'shift' to the ACP test to help? yes, it is kind of screwy that you have to put in a contribution that will cause you to fail the test and then have to pay a penalty.

the reg cite is (1.401(m)-2(a)(5)(iv))

you are permitted to either test the ACP on all match or just match exceeding 4% (if I remember the regs correctly.)

no. in fact you can make the effective date of the plan itself 1/1/2009, and the effective date for deferrals 10/1/2009.

currently the max deferral is 16,500 and max comp is 245000, so that is 6.73% on a weekly payroll you would have 16,500 / 52 = 317.30 catch up 5500 / 52 = 105.77 so someone could elect to defer 423.07 weekly in 'anticipation' of having an amount treated as catch up. however, if they quit before plan year end those amounts might never be treated as a catch up because they never reached the 16,500. since to max out on deferrals would take 6.73% - and that is with max comp, I don't see how the 6% issue would ever come into play - unless you are also excluding some type of comp. I suppose you might be trying to avoid situations in which someone deferred 0 part of the year and then maxed deferrals , but you could always match on a payroll basis and avoid that.

the Code 415(d) Cost of living adjustments describes how the limits are calculated. you take the 3 CPI-U factors for the quarter July - Aug for a given year and divide by the the base period factors from July -Aug 2001 which were 177.5, 177.5, and 178.3 = 533.3 there is nothing in here that says they can't decrease that I am aware of it. (these rules apply for the other limits as well (e.g. compensation, 402(g) etc) the website for finding the factors is http://data.bls.gov/cgi-bin/surveymost?cu click on the first choice the latest factor is released roughly the 20th of each month so the most current factors (May, June July are 213.856, 215.693 and 215.351 = 644.9 644.9 / 533.3 * (160,000) = 193,488 since adjustments are made in increments of 5000, this implies the limit for next year would be 190,000 which would be a drop from 195,000. 402(g)(4) says the base period is July-Aug 2005 (and a base amount of 15,000), so you would have to make some adjustments to see the effect for the deferral limit. 9e.g. the base afactor is 590.6 instead of 533.3)

you are probably thinking of Class Exemption PTE 2002-51 Briefly (and I am probably missing something) deferrals were deposited not more than 180 calendar days from date the amounts received by employer Can only use once every three years Have met all the conditions of the VFC program (including a notice to participants) and received a no action letter If have an exemption there is NO excise tax if no excise tax, then of course no 5330 if the excise tax was less than $100, then you don't need to give the participants the notice provided the $ was paid to the plan instead. you still had to go through the motions of filing the 5330, but it goes to the DOL instead.

you can of course exclude anyone by name if you want (as long as you pass coverage) priests who are members of a religious order, and have taken a vow of poverty at one time were excludedable from 403(b) plans, but as I understand things even this is no longer true Universal Availability The following groups of employees, who previously could be excluded, may no longer be excluded from making employee salary deferrals to their employer's 403(b) plan: Union Employees -Visiting Professors -Individuals who make a one-time election to participate in a governmental plan -Employees affiliated with a religious order who take a vow of poverty.

the example #4 of 1.411(d)-3(a)(4) is a merger between a 5 yr cliff schedule and a 3/20. (this reg was written before the revised nonelective vesting schedule now at 2/20, but what the heck.) it would seem from the example you could end up with 2 vesting schedules - or at least that is the way I read it. "...a method of avoiding a schedule 411(d)(6) violation with respect to account balances attributable to benefits accrued as of the applicable amendment date...." last sentence " ... for those account balances and earnings." thus the example ends up with a vesting of 3/20, 4/40, and 5/100% [a combo of the two] - your situation is the same, except you have a 2/20 schedule and 3 yr cliff. you end up not quite at old $ use old schedule and new $ use new. old $ use a mofidifed old schedule

I believe the regs require the notice be provided so that people can make an informed decision. 30 - 90 days is considered 'reasonable'. can you do less than 30 days? well, you lose the guarantee of 'reasonable', but I suppose you could provide a notice less than 30 days. probably becomes a facts and circumstance issue at that point in the game.

we are moving the office to a new loacation, so the books are packed away at the moment, but I also seem to recall buried in the regs if matches are deposited more than 12 months after plan year end they are tested as if they were nonelectives rather than matches (certainly its true for deferral, but I thought the same rule applied to match) fun stuff. Probably 2007 and 2008 can be self corrected under SCP becasue you are within the 2 year time frame, but you are asking me to recall stuff from my brain which is risky business.

I'd generally say yes since safe harbors only need go to those eligible to defer. unless the document was so poorly written (unintentionally of course) that would require otherwise

tripped across the following in the ERISA Outline Book 2006 edition page 1.14 2.a "....if the physician is an independent contractor of the hospital, the physicians should be able to maintain a plan with respect to that independant practice" this is from the section of deinitions on Affiliated Service Groups.

I take it that means you don't have enough 'ineligibles' that you can also give a QNEC to in order to pass testing (of course, under the transition rule there is no reason for griping because the government gave them a window and they simply didn't take advantage of amending until it was too late)

are you saying you gave all the NHCEs a match? ugh, if I remember you cant even use the avg ben % test (nondiscrimation classification test) if you have fail safe language.

I don't think so. 1.410(b)-7(d)(5) says you can't aggregate plans unlee they have the same plan year (not 'same plan year end')

all limits (except 415) are 'begining of year values' so: plan year runs from 2/1/08 - 1/31/09 is that plan year top heavy? I need to look at the balances as of 1/31/08. This would also be the plan year 2/1/07 - 1/31/08. is the individual an officer making more than 145,000 in that period? if yes (and assuming the number of officers is not capped) then the indivdual is key ad the balance as of 1/31/08 would be used to determine if the plan is top heavy for the 2/1/08-1/31/09 period.

before 'going under' a prior TPA calculated 2007 cross tested contribution. during this time the company with the plan switched ownership. the prior owners paid the contribution to the new owner under the assumption it would be paid to the plan. thus it ended up in an escrow account and wasn't deposited until December 2008. so, is it simply late and therefore, even though for the 2007 year you have to use use 2008 comp to determine the 415 limit?

I don't think there is anything 'per se' that says you have to perform one test before the other the consequences are a different matter. its possible to self-correct an ADP failure a failure of coverage is more serious. you would have to go through VCP to fix if its late. and then, you would be stuck using the same method you might have used for nondiscrim testing - or if you switch, then you need to rerun the nondiscrim. I would say you don't really run one before the other, but rather in tandem, your goal should be to put the client in the best position. ......... As for not including those not eligible in the ACP test, consider the fact that it is possible to end up with a smaller different body count in the ACP test than in the ADP test. also, if you have a controlled group, you would not include all bodies in the ACP test unless you were permissively aggregating the plans.

and note the difference if you received 'more than' not 'if you received $x or more'

I think I stand (or sit) corrected. looking up the actually wording in the regs, I see that under 1.401(a)(4)-11(g)(6)(vii) it says for 401(k) plans QNECs under a 'corrective amendment' can only be given to NHCEs who were not eligible for a given plan year. If you are also failing a(4) coverage then you could provide an additional QNEC to the rank and file and the side effect would be passing the avg ben % test, which in turn would help pass the 401(k) coverage remember, the safe harbor nonelective is still a nonelective, so is treated as such for coverage. it doesn't help the 401k portion of the coverage if it only goes to NHCEs. however, if HCES didn't receive then it would make it easier to pass the avg ben % test.

if I understand your question (which is sometimes a big IF in my case) the regs are clear you can't aggregate a safe harbor with a non safe harbor for ADP testing. since you have to treat plans the same for coverage as you do for discrim testing that would seem to eliminate aggregating them for coverage. you indicated a ratio % of around 50%. you did not indicate what the avg ben % was. if greater than 70% then you still might pass coverage if ratio % > safe harbor%. depending on how bad (or good) the avg ben % test is, you might be able to increase the avg ben % test not by bringing more people in, but by providing QNECs under the corrective amendment process.

That can be found in 1.401(k)-5 Special rules for mergers, acquisitions and similar events. This is what it says: "Reserved" In other words, they haven't told us how to do it because it appears they haven't decided how to handle cases involving mergers of any 401k plans. that being said, since a safe harbor plan must be 12 months long, I would say you most likely can't merge the two animals, but there is no real guidance. one thing is clear, you definitely can't aggregate for testing.

the restructuring rules are found in 1.401(a)(4)-9© in regards to this question 1.401(a)(4)-9©(3)(ii) last sentence "...See 1.401(k)-1(b)(4)(iv)(B) and 1.401(m)-1(b)(4)(iv)(B) for rules regarding the inapplicability of restructuring to section 401(k) plans and section 401(m) plans." it should be very easy using a black magic marker to cross that requirement out, though perhaps white out might be a better choice. if you have enough 'otherwise excludables' you could disaggregate for testing, but that is about as close as you can come to breaking the plan into groups. you have that many people excluded from deferring?