Tom Poje

Not sure if I agree with comments on 'knowing what an E-Bar is' For example, I had Physics in High School, and another course in college. I could 'do the work' or 'know how to do the stuff' such as calculating resistors in parallel or in series. But it wasn't until I had a course in digital electronics that I really understod what was going on. FJR: you can never pick and choose, you have to follow the terms of the document. A document could be written to provide the minimum gateway to NHCEs only. If you did that, and had HCEs in separate classes then conceivably you could provide a 5% to 3 of the HCEs. Blinky: will you at least post a few more times so you can have more posts than I do?

sorry PATA, the closest is going to be the coverage and nondiscrim answer book. doing my best to add more examples, I did add a few on age weighted calcs for the new supplement this year. maybe when they publish the book all over again I can add a chapter for 'dummies' or you can sleep through one of my talks at the Fall ASPPA conference. I get to repeat the one on Safe Harbor 401k plans and Cross Testing (Though just getting through the basics of safe harbor doesn't leave a lot of time to toss in some info on how they effect cross testing.) Looks like I may have to eliminate the 'pension song'. And if you sleep through the talk you'll miss a few 'surprise' slides that weren't in last year. Hmmmm. How is it that I somehow worked Star Wars into this thing... On the serious side of the talk, it was really nice to hear a comment from someone who has been doing cross testing... "Oh that is what an E-Bar is and how they are generated".

Before the question comes up, Safe harbors have a further restriction - no hardships. see Notice 98-52 IV.H

your gut feeling is correct. comp = 210,000 max deferral = 14,000 415 limit = 42,000 so w/o match, profit sharing = 28,000 28,000 / 210,000 = 13.33% so gateway = 4.44% (lesser of 1/3 or 5%) if ee received 8000 in match then profit sharing is only 20,000. 20,000 / 210,000 = 9.52% so gatway = 3.17% since plan will undoubtabely be top heavy heavy, you really can't do much better than that. however, another problem arises. the HCEs total contribution remains the same, 42,000. in other words, to achieve a smaller gateway, you have replaced 3.8% profit sharing with 3.8% match. unless you do the same with the NHCEs (and if the NHCEs are not deferring much and so get little or no match) the test falls apart at the avg ben % test. Of all the cross tested plans I have run, I have very rarely had one pass the ratio percentage test and therefore not need the avg ben test. In fact, I suspect what very few of those that exist would do better as an age weighted plan, and thus avoid the minimum gateway entirely.

this depends on what type of SIMPLE A SIMPLE IRA has no comp cap if there is a match. A SIMPLE 401k limits the comp considered. so both web sites may be correct

I'm not sure its a matter of looking a the regs, but rather at the document. And if the document is silent...hmmm. If you use the APR at current age this would result, for example, in 'discrimination' between 2 ees both making the same comp, but one at age 65 and one at age 70 and the one at age 65 would receive a larger contribution. to get around this, documents I have seen say to use the same APR for both ees, resulting in the same contribution for both ees. (Apparently yours does not say this, but worse says nothing one way or the other) if you were to test for nondiscrim in this scenario the older ee would end up with a larger E-Bar because you are suppose to use the APR at current age if you are past NRA. The 'out' so to speak, is in the regs... "This would seem to raise the possibility of failing nondiscrimination testing. However, a plan will not fail discrimination testing merely because allocations are made at the same rate for employees who are older than their testing age (determined without regard to the current age rule in paragraph (4) of the definition of testing age in Treasury Regulations Section 1.401(a)(4)-12) as they are made for employees who are at their testing age. [Treas Reg § 1.401(a)(4)-8(b)(3)(ii)]" [i had written an example of this very scenario with E-Bars and everything else, so I have my 'quote' handy]

1.401(m)-2(b)(4)(i) 10% excise tax on excess aggregate contribution it does not say 10% on 'match related to excess contribution' so no tax

this is one of those points that is possibly arguable. technically there is no de minimus amount. under the self correction program, there is a note that says if the total corrective distribution is less than $50 then you are not required to make it if the cost of processing is more than its worth. now, does the term 'corrective distribution' refer to failed tests? or does this only refer to someone you paid out and found they had additional money coming? in other words, a year after the fact I rerun my ADP test and discover I fail by $3. do I get to exercise this option? if yes, it seems wierd I would have to refund the $3 if I complete my test within a year but after a year I could rely on the self correction program. does exercising this option 'significantly favor' the HCEs? I wouldn't think $3 would. does your document allow you to use any definition of comp that satisfies 414(s)? In other words, if you use comp - deferral would that shift numbers enough to pass testing?

you are correct, the regs are not explicit about whether to include excess contributions, but the general consensus is that you should. if they were recharacterized as cath up, then you definitely wouldn't.

you have to follow the terms of the document, and yes, the document can say match every pay period so no true up is needed. Is one method favorable over another? guess that depends. If an ee defers nothing the first 6 months and then defers 10% the last 6 months you would have a true up at the end of the year. on the other hand if you match per payroll only, then he suffers the consequences for not deferring all year. on the other, consider Mr. Mega Bucks he who socks away $14,000 in deferral the first month. he would not get a match the rest of the year. these people need to realize not to defer so much so fast if there is not a true up.

ah, the safe way to go. I think the problem the IRS would have with the document (certainly once the new regs go into effect) it sounds like the document has safe harbor language in it, but with a caveat that the notice will decide if the plan is really safe harbor, and the notice will be treayed as an amendment. I would read the new regs to say that the current plan has no safe harbor language in it, but rather is amended to include such language in the case of a contingent safe harbor. this can be done on a year by year basis. the preamble to the final regs says A PLAN that uses the safe harbor method MUST specify whether the safe harbor contribution will be the SHNEC or the SHMAC and is NOT permitted to provide that ADP testing will be used if the requirements for the safe harbor are not satisfied. (e.g. in your case, the notice being issued) The safe harbors are intended to provide ees with a minimum threshold in benefits in exchange for easier compliance for the plan sponsor. It would be inconsistent with this approach to providing benefits to allow an employer to deliver smaller benefits to NHCEs and revert to testing. basically the document cited would seem to fail this condition. And I am not picking on that document - I have seen a few others that have similar language. again, how the IRS will handle things prior to the effectove date of 1/1/2006, I am not sure, since things were unclear. after that date, well.... this is my understanding of how to do it. by Dec 1, 2005 issue notice, we might go safe harbor. (if you don't, probably too late to go safe harbor in 2006) by Dec 1, 2006 issue notice stating whether plan is safe harbor for 2006. if yes, plan must be amended. it can be amended for 2006 alone or for all future years. if amended for 2006 alone, also issue notice for 2007 saying plan might go safe harbor for 2007. I know, it sounds like the notice is the amendment, but I guess there is enough of a difference in that there is no language in the document. also make sure document says current year testing rather than prior year testing. that is a requirement if you are going to use contingent notices.

Andy way over rates me, I have to look everything up. unfortunately I am beginning to get a better idea of where to look things uo. he refers to 1.401(a)(4)-8(b) the excpetions to the minimum gateway being 1. broadly available bands (would seem to fail this since one ee only gets 3% and another the same age gets 3% plus the age weighted) 2. age based allocation rates (while the plan does have this, that is not the sole contribution -so the question is does that take it out of this option) I'd lean toward the conservative side and provide the gateway to the terminated ee. he certainly has to receive the 3% SHNEC - no option there. now, once he has received the SHNEC, he has received a nonelective in one way shape or form. and anyone who receives a nonelective has to receive the minimum gateway. This person is not a part of the group 'age based allocation rate'. for example lets suppose the plan was top heavy, immediate elig to defer, 1 year wait for age based allocation. Plan is not safe harbor. At that point would you provide the minimum gateway for him? I think you would have to. I can't believe the addition gateway to a terminee with less than 500 hours would cost that much, and quite possibly the ee in question is not fully vested either. remember, the minimum gateway is not fully vested. Now, if the document does not have gateway language in it, then I guess this would have to be done through a corrective amendment.

ugh, you have to follow the regs which says that attributable match MUST be forfeited.(It never matters if these amounts are vested or not) It would certainly be nice if the document said forfeitures could be used to pay plan expenses, but that doesn't sound like what you have. I don't have a copy of the Corbel checklist handy, so I don't know if there is an item buried somewhere that gives you an option like 'forfeitures resulting from attributable match due to excess contributions will be used to....' Ugh. it sounds like you may be stuck with small amounts in the profit sharing accounts, if your document says that is how you handle any forfeitures. Oddly enough, if you could figure out a way to make the ACP test fail you could distribute the money, since you are allowed to correct ACP failures before the ADP failures. for instance, assuming document allows it, for testing comp use full year comp rather than comp from date of entry.

one of the reason of the new regs was to make it clear that a safe harbor plan is driven by the document, not by the notice. In other words, the IRS has come back and said 'no, you really are not suppose to run things this way.' I realize that documents do exist that are worded this way, and how do you handle things in those scenarios, especially if you have a determination letter? well, I'm not a document expert or attorney or whatever. personally I think you are stuck. the IRS has stated a number of times that even if no notice was provided that does not free one from the safe harbor contribution obligation. in other words, a plan is either safe harbor or it is not. the only exception is the contingent notice, but even then one is required to issue the 30 day notice notice before plan year begins and then again plan year ends. and if one or both was not done, you have an operational failure.

At the fall 2004 ASPPA conference the IRS answer to the question was "If the safe harbor notice provide was the 'fixed 3% contribution', then no amendment to eliminate the 3% contribution is POSSIBLE (barring plan termination). This was Question 17. (Unlike a money purchase where you could reduce or eliminate the contribution during the year) now, 2 things must be remembered - comments by IRS agents at conference do not necessarily represent the actual Treasury position. Second, these comments were made before the issuance of the new regs. (I'd also add if the plan was terminated and then a successor plan was started I have a strong suspicion that the IRS would not take kindly to it) However, usually comments by IRS agents are fairly reliable. In this particular case, I think the 'out' is in the preamble (rather than the regs) which says A safe harbor plan could also have a short plan year in the year the plan terminates (w/o regard to the reason for termination or finacial condition of the employer) if the employer makes the safe harbor contributions for the short year, ees notified of change (I assume that means 30 days notice) and plan passes the ADP test. In either case, employer must make safe harbor contributions through the date of plan termination. in other words, the preamble seems to 'interpret' 1.401(k)-3(e)(4)(i) by replacing the words 'safe harbor matching contributions' with 'safe harbor contributions'

generally a safe harbor plan must be 12 months long. the new regs include situations in which the plan might be shorter than 12 months: Plan termination The plan follows the rules for a plan that was reducing or eliminating a safe harbor match, other than the requirement that participants have a reasonable opportunity to modify their deferral elections or The termination is a result of disposition or acquisition or employer incurs a business hardship as described in section 412(d) ( 1.401(k)-3(e)(4)) A safe harbor 401(k) plan may be amended during the plan year to reduce or eliminate the safe harbor matching contributions provided all the following conditions are met: 1. notice of suspension is provided to all participants 2. reduction or suspension of the matching contributions is effective no earlier than 30 days are eligible participants are provided the suspension notice 3. eligible participants are provided an opportunity to modify their deferral elections (and after tax contribution elections if applicable) 4. the plan is amended to provide that the ADP /ACP test will be satisfied for the entire year and the current year testing method will be used 5. the safe harbor matching contributions have been provided through the effective date of the amendment. 1.401(k)-3(g) and 1.401(m)-3(h) so, if you don't meet the conditions of 412(d) are you allowed to follow the guidelines for a safe harbor match (even though the plan has a SHNEC instead?) I'd tend to say no since you could have the SHNEC be a 'maybe'. If it is a case of hardship, then I would follow the guidelines as described for the match as close as possible.

if the HCE gets a bonus but also makes over the comp limit (200,000 as indexed) then you could still fail the compensation test as the exclusion of bonuses has no effect on their comp for allocation purposes. If the HCEs are the only people getting bonuses then I guess you would pass as well.

1. since the regs say the new rules apply to plan years beginning 1/1/06 and since the distribution applies to a plan year that began before 1/1/2006 I would assume the gap income rule does not apply merely because the distribution is taking place after 1/1/2006. 2. I would think your rule on gap income would be 'what does the document say' for example, a typical corbel document in regards to testing comp merely says something like 'any definition that satisfies 414(s)' that would give you plently of wiggle room. I have not seen the latest language they will have for gap income. perhaps they will have similar language 'any of the methods that satisfies...' or maybe it will list them 'gap income is satisfied by following one of these methods:" ........ oooops. by the way, if you wanted to you could follow the new regs right now, as long as you follow all of them. why you would want to do that is beyond me, but...

Interesting thoughts. I would hold the answer from the Answer Book might need to be expanded on, but I am not sure if it is incorrect. If a plan fails the 414s comp test what is the correction? run your ADP test using a definition that satisfies 414s. but there is no test to perform in a safe harbor plan. In fact the regs (certainly the preamble to the new regs makes that clear - a safe harbor plan 'is not permitted to provide that the ADP testing will be used if the requirements for the safe harbor are not satisfied")Thus, at that point, are you allowed to increase your safe harbor contribution using a definition of comp that satisfies one of the safe harbor of 414(s)? I am not sure under what rule you could do that. That defeats the intention of safe harbor 401(k) plans. (I see the ERISA Outline Book simply says the plan will generally use a definition that satisfies safe harbor 414(s) otherwise it will HAVE to satisfy the 414s comp test (Emphasis mine) Thus, I suppose if you were 100% guaranteed to pass the 414(s) compensation test every time, you could get by - but can you guarantee that? In addition, if you were providing the safe harbor on a payroll by payroll basis (thus no true up) I guess that would require you to perform a 414s test every payroll.

gateway only comes into play if you want to cross test

if plan has a last day rule, then no one has accrued a benefit until the last day. this generally means you could amend the formula e.g. suppose you had one class = owners you discover they have hired the son. since young son is an owner your cross test will fail miserably unless you amend. if you only had 1000 hour stipulation, once you get passed mid year it would be too late to fix the current year.

I have no problem it being posted. the file size if 5000+ KB so it takes forever to upload. I suspect I can no longer post files because of all the security in the office. For an actual plan - use Relius for cross testing for studies - use the spreadsheet mainly cuz it is so easy to try different combinations. for checking data - I will do a hand calc from time to time by the way, I suspect the spreadsheet has not been updated for the new limits, comp, etc. I'm no excel expert but I did manage to figure out how to add the latest covered comp tables - (used if imputing disparity) my memories of the Floridata printer - having 2 computers share a printer, and you would look at the individual sitting across from you and 'guess' what report he was printing by the sound - you really could tell.

while I appreciate the comment, you shouldn't rely on my opinion - it is really the regs - see 1.401(a)(4)-3(d)(3)(ii) ... Accrual rates within a range may not be grouped under this paragraph if the accrual rateswithin the range genearlly are significantly higher than the accrual rates of the NHCEs in the range... There are no guidlines what would be considered significantly higher, but you are very observant Merlin - setting the HCE at the top bound would seem to fail the smell test of this reg. good job, 10 points. I know you can manually set the range(hint hint hint - I would do it that way), but I have only been able to do this for one HCE - I am not sure if the intention is to be able to set more than one range at a time, maybe I am doing something or wrong, or maybe there is glitch in the system

I suspect that report is my variation of Fred's. (especially since it says 'w report' cuz I wanted a print out of just contribution data. If you used Pentabs, did you use the Floridata Dot Matrix Printers. Man those things wre built to last. I bet those suckers would still run even today.

nrehme - if you send me your e-mail I can send you the file (click on my name and you can send an e-mail through the benefits link directly) as for how does one learn - that is real interesting. I worked for Corbel many moons ago, back in the glory years of DOS of Pentabs, and they came out with the nondiscrimination module. There were no instructions on where the numbers came from, and seeing how I figured I would take calls on the system I taught myself by simply running different contributions, and mathematically figured out where the numbers came from. trial and error. practice. sort of like learning math in grade school, repetition. There certainly wasn't a lot of material available at the time. I suppose it would have been great to have a lot of material available but if it had been I doubt I would have learned as much as I have (or at least as 'deep' as I have) without actually crunching the numbers by hand. And I would be the first to admit I might have made some errors along the way, or could have done things better - I think that is why people throw out cautions about simply plugging numbers and going with the results. Good luck! Books are a great resource, attending seminars helps, but I suspect most of those are aimed at people who have a grasp of the material to start with. At last year's fall ASPPA conference during the talk I gave I walked through the mathematics of simply depsoiting money in a bank account and watching that one deposit grow with interest. And then suddenly I ended up with an E-Bar - I suspect few saw it coming - very effective. It was great to hear someone tell me they had been doing cross testing for a few years but now the numbers made a lot more sense.