401(a)4 testing required?

[Guest RACHELP]

Situation is as follows:

401(k) Plan, top heavy - no match

6 HCE

8 NHCE 3 Terminated during plan year

Compensation only counted from date of participation (calendar year plan , 1/1 7/1 entry)

Plan fails ADP

QNEC given to all nhce (hces do NOT get qnec) - no conditions to receive allocation

3% min given to all participants except 3 nhce not there on last day of plan year

am I correct that I have automatic pass for 410(b) but must now test for nondiscrim because I have two benefit formulas with different allocation conditions?

this fails (a)4 b/c 3 terminated nches only getting tiny QNEC and because one HCE who got in middle of year is getting top heavy min on all comp but 414(s) comp is half year

what can I do here to make it work - bring in one term nhce - do I have to do this by amendment?

Help

[Tom Poje]

ugh, maybe.

you said plan year was calendar, so I assume you are talking 12/31/2004.

that means to fix with corrective amendment you have 9 1/2 months. too late.

(Though I suppose if you are talking about one of the hurricane effected areas you could argue for more time)

anyway:

Coverage: you indicated all NHCEs received so coverage passes at 100% (at least for nonelective portion)

assuming all could defer, so 401(k) portion passes at 100% as well.

401(m) - unknown based on data provided.

as for nondiscrim, your conclusion is basically correct, but lets review what has taken place.

QNEC given to all NHCEs, no HCEs. this is considered a safe harbor formula. It is ok to exclude HCEs.

nonelective - if this was also provided to all ees then it would also be considered safe harbor and under the multiple plan formula rule no further testing is needed. however, as their is a last day provision (e.g. it is a top heavy) then testing is needed.

now the regs say a plan must pass 401(a)(4) with QNECs

a plan must also pass 401(a)(4) without QNECs

1.401(k)-2(a)(6)(ii) of new regs 1.401(k)-1(b)(5) of old regs depending on what you have handy.

obviously testing with QNECs should pass. well maybe not obvious since QNEC is small. hard to say w/o knowing numbers.

now w/o QNECs only 5/8 NHCEs get or 62.5% so ratio % fails, but that would be enough to pass midpoint, so if you can pass avg ben % test, I think you are ok.

as for testing comp, some documents describe 414(s) comp as any definition that satisfies 414(s) can be used, in which case you could use full year comp.

(comp from date of entry being allocation comp not testing comp)

imputing disparity should help.

plan could be tested on accrued to date basis, which is a little more complicated, but uses not just current year but would include some of the prior year contributions as well.

You didn't indicate if any of the terminees had less than 500 hours. certainly when testing with QNEC you would have to include them in testing since they received a contribution.

I have never seen the issue addressed before, but I guess when testing w/o QNECs you would be able to exclude those ees since you can exclude terminees with less than 500 hours.

of course, I could be wrong about all the above, but that is my understanding of the rules.

[Guest mparker2028]

I follow along, but I must tell you this is for 2003 Plan Year

Aside from the various violations/failures this client has which I won't go into

the doc says must pass 410(b)1 and gives instructions for how to bring in new people.

does this language preclude using avg benefit test in (a)4 or am I mixing apples and oranges

all term participants worked >500 hours

do you know where in relius I could change testing comp (if my doc even allows to use any def of 414scomp)

there is no match for this plan - it's discretionary

I can't seem to pass any way I do it

since this is 03 year - have any suggestions (we are in the midst of considering EPCRS in any way shape or form)

thanks for your help

[Tom Poje]

in Relius you can override testing comp - without opening up a plan I believe it is a subscreen along with computed comp.

the language of bringing people in is for coverage and I don't believe that applies to nondiscrim testing. (You are correct, such language precludes the use of avg ben test, but I am pretty sure that applies to coverage only) and you are passing coverage since all NHCE received a QNEC.

you cant 'self correct' nondiscrim failure, so somehow have to figure out how to get the sucker to pass.

but if I understand things, you have 5 NHCEs at 3% 5 * 3 = 15

8 nhce total, so 15 / 8 = 1.875

HCEs are at 3%, so 70% * 3 = 2.1 so fail avg ben % test without the deferrals (which would have to be included. but since you fail ADP test, including deferrals makes things worse. but that is testing on allocation basis. so cross testing might work depending on ages.

on the other hand, lets suppose all 8 received top heavy in 2002.

then you would have

8* 3 + 5 * 3 = 39 for 2 years. 39 / 8 nhce/2 years = 2.4375

HCEs al got 3% so their avg is still 3%.

70% * 3 = 2.1 so 'passes' avg ben % test but you still have to add back deferrals (so that is a guestimate number anyway), but at least you are in the running.

its a bit more complicated than that, that is a crude illustration only

[Guest mparker2028]

The document is a Corbel Volume Submitter

doc does say that any definition of 414s can be used but must be applied to all participants

this hce with half year comp is the only participant who entered midyear

if i overide and use full year comp instead of half year for this hce for (a)4 testing, do i also have to use full year comp for adp test?

it doesnt affect adp results - he deferred nothing

thanks for all your help

[Tom Poje]

I don't think you need to change your ADP test.

The adp test is the nondiscrim test for deferrals and you use what options work.

for example, if you used otherwise excludable for the ADP test, there is no requirement that you do the same for the a(4) test.

you probably noticed there are two comp fields - one for rate group and one for avg ben % test, so it sounds like you need to override twice.

ah, good universal language description - at least that is how I would read it.

[Guest RACHELP]

How do you run the general test in Relius on employer contributions without QNECs? Do you do it by reversing the QNEC transaction? If I do that and run the general test on an accrual basis, I pass on accrual basis - I get past the gateway b/c 3 termed people are getting nothing by way of top heavy min and qnec has yet to be allocated?

This seems like cheating - is it?

Some day this job isn't worth it.

[Tom Poje]

I think you are correct, the only way you can run it on Relius would be to reverse the QNEC and simply not include it.

as to the other part of your question, I am really not sure.

if I understand the regs correctly, In order to cross test, one has to first pass through the gateway.

Once all ees who have received some type of nonelctive (including QNECs) have also received enough to satisfy the gateway, then I proceed to cross testing.

at that point I run with whatever numbers I have under the rules I have (testing with and without QNEC)

this results in an odd situation. an ee who receives a QNEC only gets bumped up to the gateway with an additional nonelctive.

now, when I perform my cross test without QNEC the ees contribution is less than the gateway. that seems strange, but I guess once you pass through the gateway you simply test following the rules.

I am not sure if you get to a point of running the test 'before the QNEC' and therefore dont have to worry about the gateway. I think the QNEC exists, buried deep in there, and simply isn't used.

but I have no clue. no wonder I never recomend giving a QNEC and also having to cross test.