Tom Poje

ok, this was a very round about way to get a report attachment. for whatever reason, I can't attach files, maybe it is internal security. anyway, the attached Relius report is an ADP correction report. It is a modified version of their report. This one will show the results of the ADP test under the 'old rules' - hence the proof that the plan passes if the refund was made. The report then shows the regular stuff under the new rules, indicating the amount refunded under old rules = amount under new rules. Also added a column indicating the deferrals after refund to show that the leveling method was used. Unfortunately, at least at 9.x relius you either have to replace the actual relius report to make this work or you could run the Relius report, and then open this report in Crystal and run it. that will also work. Now if I can figure out how to move this sheet to the Relius board. success, I have powers I never knew about!

first, lets get at least the terminology correct. the plan is being tested for nondiscrimination. (and lets preface all this by saying the same basic rules that apply to coverage apply to nondiscrim testing - the nondiscrimination test is slightly modified) for coverage you are required to disaggregate the 401(k), 401(m) and nonelective. keep that in mind - those are the 'bodies' you look at for your testing group. In particular, rate group testing is being performed. you said you failed, if that was true, then there is little you can do but make corrections somehow. I suspect what you mean to say is that the rate group test failed the ratio percentage test and therefore you are going to try to pass nondiscrimination using the average benefits test. there are 2 parts 1. nondiscrimination classification test - this is still a rate group test, only easier to pass since you only need to pass the 'midpoint' again, this group is based on the coverage tested group and is performed separately (the famous 'required disaggregation') for each group (e.g. 401(k), 401(m) and nonelective) thus, terminees with less than 500 hours could be excluded (remember this an option and not required, it may be helpful to include if there are HCEs that fall into this group.) so you are correct, if you didn't include them for coverage, you wouldn't include them here either. 2. for the average benefits percentage there is no disaggregation, you must include everybody who benefited. since the terminees could have deferred, you must include them as they did benefit. so, to answer your question, as to whether to include the terminees < 500 hours, the answer is both yes and no. it depeneds on what part of the test you are talking about.

I suppose the old exams are good practice for getting a feel of the type of questions that are asked. I had planned on taking the c-3 back in 1999, but the work load just never allowed me to really get the time. I did finally take C-3 in 2003 and get the CPC, but it was the ASPPA cram session that got me by thing. I think just sitting in for 4 sessions of intense review (eyes glazing over, brain begging for mercy) convinced me that I really had a good grasp of the material - that did more for me that looking at the old tests. And being able to take the test only a few weeks after the cram session really did the trick.

the SHNEC has to be 'at least' 3%, so one could provide a greater amount in a given year. however, assuming the notice specified the amount as 3%, I see no way to simply increase the amount to 4% after the fact. you would be in violation of some rule or something. I have never heard a notice specify that the SHNEC 'will be at least 3%', my gut feeling is the safe harbor should be definitely determinable in order to get the free ride on ADP testing, but I have nothing in which to back up that statement.

in other words you can't quit age 50 and then start collecting at age 55 and avoid the penalty. and in other words, the concept works much the same way as catch up eligible employees. as long as you turn the magic age during the calendar year, you are ok.

the only exception I can think of is if a an ee has excess deferrals (e.g. above the 402(g) limit, those amounts are treated as catch up before the ADP test is performed. It doesn't sound like that is what took place in your plan, but one never knows.

the only rule you have to worry about is that anyone with 3 years of service must be given the opportunity to remain under the old vesting schedule. many moons ago it probably made a bigger difference (e.g. when you had a lot more vesting schedules, ah the gory years of 'rule of 45' or whatever that was called)

my question would be along similar lines. I 'inherit' a plan and have no clue if terminees were reported (as they should have been). a participant gets paid out. To be on the safe side are you still allowed to report them as a 'D' even though they may have never been reported previously?

penman: for testing purposes ,you have to use a definition that satisfies 414(s) now, there are a number of definitions that satisfy this requirement. my reading of the documents is that they are more flexible than they used to be in other words,for example a few years ago the document said 'use total comp' - not much leeway. now it says any definition that satisfies 414s if I recall, the concept of 'consistency' from year to year applied to the determination of who was HCE - which would be in regards to something like rounding up or rounding down if you were determining who was in the top paid group.

actually you must follow the terms of the document. I believe the new Corbel documents all say "any definition that satisfies 414(s)" I think in previous years you had to be specific "comp from date of participation" or "total comp" I don't know what other documents definitions are.

someone in the office here printed the IAD report to a file, then opened it in word. when printing, set printer preferences to 'paper saver' and this printed 2 people per page. of course, that might not be an option with your printer. well, that is half as many pages. plus you could modify the IAD report to exclude ineligibles.

my understanding is that you can always use comp from date of participation for testing purposes (unless you have a poor document that says to use total comp or something silly like that) it does not matter if ee received top heavy based on full year. ee has met the gateway as far as I can tell since he has received a minimum based on comp from date of participation.

since plan has a last day rule, technically no one has accrued anything, and so the general rule of thumb has been you can amend the plan before the end of the year. of course there is always the possibility the plan also says no hours if one dies, disabled, normal retirement' and if someone indeed has hit one of those condition you would be out of luck.

in addition, watch top-heavy. if you say 'all ees get' then all means all. if you say 'all non keys get top heavy' then you can accomplish what you want. assuming of course it is a typical cross tested plan - very few are not top heavy

To clarify " Short PY will create a year of vesting service." most likely it will. vesting has to be measured in 12 month periods, so despite a short plan year, you would still look at a 12 month period to determine if an ee has 1000 hours. see (Labor Reg 2530-203.2©.

just what part of 'irrevocable' means 'unless you put in a 401(k) feature' or when the regs say it is irrevocabe 'on the employee's behalf to the plan and to any other plan for the employer (INCLUDING PLANS NOT YET ESTABLISHED) for the duration of the employee's employment..' 1.401(k)-1(a)(3)(iv) New regs is 1.401(k)-1(a)(3)(v) - worded slightly different but it says the same thing again, it doesn't say 'INCLUDING PLANS NOT YET ESTABLISHED unless they are a 401(k)' emphasis mine, but not meant to be interpreted as screaming or any such thing.

Blinky - I forget when I tripped across that piece of info - a number of years ago, and it sort of has stuck in the back of the thick skull. now that you have more posts than me I would like to say "you're a better man than I am..." but given your 'state of being' it wouldn't quite be true. hmmmm. I am fishing a quote that fits you.

as described, the following occurred: deferrals made in Dec 2003 due to error never deposited until 2005. the regs are clear on this one 1.401(k)-2(a)(5)(i) this refers you to 1.401(k)-2(a)(4)(i) which requires deferrals to be deposited no later than end of the 12 month period following the year to which the contribution relates. (wow, I have never seen this one come up before!) thus, you may not, I repeat NOT use these in the ADP test. instead they must satisfy the a(4) test as if they were the only nonelective contributions for the year. (I doubt you could cross test because that would require the gateway!) now, the cite I provided is from the new regs, but I know it exists in the old regs as well, I just am too lazy to thumb through the book to find it since I have this one handy.)

regs say "may be amend after the first day of the plan year and no later than 30 days before the last day of the plan year 1.401(k)-3(f)

many many moons ago, back when I was doing software support, I figured out a way to trick the old glorious Pentabs system into doing an ideal salary calc for an age weighted plan - using of all things, the target benefit module. It started from a service call in which the client asked if it was possible, and I said no, but give me a day and I'll figure a way around it. send me the plan to test. I'm sure I figured it out on a Friday night - no one else wanted to work that night, since I'm no party animal it didn't matter to me. I thought it was great, cuz those nights were slow for service calls, and it gave me a chance to figure out solutions to such ridiculous questions. well, I expected a simple 10 life 1 ideal salary calc - instead I got a 100 life with 20 'owners' to figure ideal salaries on. what a test! but the solution I came up with actually worked. well, I'm not in support anymore, I have a desk covered with work, and appreciate even more why software people get paid for the time they spend doing such things. I actually have an idea how it might be possible to do this in the spreadsheet, at least on a fairly simple level, but getting the time to even try it is another thing. Go and figure, my boss doesn't like me spending my time doing stuff for nothing! (I do take that as a compliment that you even mentioned my name in your post!)

agreed wit heverything you say, (or just simply 'greed' in the case of HCEs) you are correct if the HCE makes less than the TWB imputing disparity won't help. The example given was both ees above the wage base so I didn't consider that, but it is possible to have an owner (or worse in a poorly written group, the owners kids with less than the wage base)

working through the formula: C Rate = allocations / (plan year comp – ½ TWB) D Rate = [allocations + (5.7% * TWB)]/comp assuming a 15% allocation to all groups: all NHCE would be 15% + 5.7% = 20.7% 100,000 would be 20.01% (D rate) 200,000 would be 17.51% (D Rate) Assumed plan year was 2003 and TWB 87,900. But this was for example purposes only. I did not read your question properly, I would assume the document says allocate comp to comp. when that happens the HCEs will have smaller e-bars if tested on an allocation basis. the reason being is that for testing purposes, part of the contribution will be considered to be 'integrated' though it really isn't. or put another way, it would be possible to allocate the owners group more than 15%, but because comps are different you wouldn't equal the effects of having an integrated plan - each HCE would receive the same %. but they could receive a higher % than the NHCEs. and at that point you might even consider the dreaded component plan testing.

In addition, I would say there are no 'delinquet years' as there is no requirement to file a schedule P

to make things clear: in this case testing will be done on an allocation basis rather than an accrual basis. If you are using 100% of the TWB, and max disparity of 5.7% then all ees should end up with the same E-Bar. If two people have different comps, it doesn't matter, they receive the same % of pay. when you impute disparity they still will have the same % when finished. If one of the ees is above the TWB, then before imputing disparity one will have a greater %, but after imputing disparity they will have the same % about the only cases I can think of where someones ebar might be different is if you used calendar year comp for allocation purposes but comp from date of entry for testing purposes.

the regs only speak of 'adding a safe harbor feature to a pre existing profit sharing plan. 1.401(k)-3(e)(2) but I imagine the process of converting the money purchase to a profit sharing plan can all be done in one step. you can only defer on amounts from the start of the 401(k) feature - that has always been the case regardless of whether the plan is safe harbor or not. when you say use comp fro entire year, for what other purpose? 1.401(k)-3(a)(2) specifically says you can limit comp to period of participation, so I would take that to mean the reverse also holds, you could use full year comp - but at that point I am not sure why you would, when you could make a profit sharing contribution that is subject to vesting in addition to the safe harbor