Is this exclusion of NHCEs for ADP test legitimate? Options

[410b]

My company uses a 3rd party 401k tester.

I was studying and trying to understand the NDT test section of the results. When I inquired about the second test group which appeared to be a small group of NHCEs separated from the main employee test data I had submitted, I was told the testing procedure:

Eliminates/disaggregates NHCEs with less than one year of service as of the last semi-annual entry date of the plan.
Eliminates/disaggregates NHCEs under age 21 as of the last semi-annual entry date of the plan.

and that the tester's reporting department was confident the test had been done correctly.

The plan allows entry at age 18 but has a service requirement of 12 months which must include 1,000 hours of service. Plan entry dates are monthly, enrollments are quarterly.
Hiring and attendance policies of the company are such that it is impossible for a new employee to meet the 12 months of employment condition without also meeting the 1,000 hours of service condition.

I have stated to my plan administrator that I don't believe the testing is being done properly.

Since the plan does not allow participation until an employee has met the 12 months of employment condition, it is my contention that there are no employees with less than 12 months of service and eliminating/disaggregating such employees is a specious procedure which could cause the plan to be disqualified. I believe the testing should be redone without eliminating any plan participants due to the "service exception" mentioned in the testers manual.

I have also suggested that eliminating employees under 21 at a mid-year date is an aggressive approach. A more conservative approach would be to eliminate only the employees under 21 as of the last monthly entry date of the plan year.

I like "conservative" in relation to eliminating appearance of conflicts of interest or manipulation of tests.

Are my comments correct or incorrect?

Thanks.

[Jim Chad]

IMNTBHO I agree with you that your approach would be the more conservative one. I also agree with them that their approach is reasonable. This is a judgment call.

As the customer, you certainly have the right to request that they do it your way. And you should also expect to pay more if you take them out of their normal system.....and it might be a lot more. It might be a lot more if they have to do your Plan manually because their software is set up the other way.

FWIW I think you may want to trust their judgment.

[PAL]

I would use the enrollment/entry dates available under the plan. In the case above, I would include anyone age 21 or older with a year of service as of the last enrollment date (10/15?). If the plan has a year of service requirement, I would agree that it is unlikely that there should be any carve-outs.

From the IRS website - see page 6, last sentence in first paragraph on right column:

http://www.irs.gov/pub/irs-pdf/p6393.pdf

PAL

[Tom Poje]

this is a gray area depending on exactly which IRS official you might be talking to.
certainly the regs provide that you can test 'otherwise excludables' separately.
now, exactly how that works and who that includes has been an open debate the last few years.

1. does the document have to specify the use of 'otherwise excludables'. there seems to be some leanings in recent years that despite the fact this is a testing 'assumption', it might be best if it is described in the document.
2. do plan entry dates make a difference?
some IRS officials voice an opinion yes, other say no. The code / regs simply say refer to max age/servcie. (age 21/1 yr of service) and then the code says one must enter the first day of plan year or 6 months after meeting the requirements.
this ends up with 3 schools of thought.
you indicated the plan had monthly entry dates.
so, for a calander year plan, employee A is age 23 and hired 2/3/06. works 1000 hrs so 1 yr ends on 2/3/07.
one school of thought says monthly entry dates apply, so if ee is still working on 3/1/07 he is in the big test.
another school of thought says 'use 1/1 and 7/1 entry dates as max exclusion'. thus is ee is still working on 7/1/07 he is includable in the big test.
the final school of thought says you could have written the document to say use the absolute max entry date possible. (e.g.1st day of plan year or 6 months after meeting age 21/1 yr of service) thus this ee would be in the big test if still working on 8/3/07.
(Personally, until IRS says one way or another my leanings are in the last camp since I could have written my document that way - thus the people who are in the plan would have been excluded if I had written the document that way - hence the term 'otherwise excludable'.

since there is no requirement to use the 'otherwise excludable' assumption, there is certainly nothing wrong with including everyone, even if it makes a plan fail testing.

[410b]

Jim Chad,

Thanks for your response.

My problem is that as the taxpayer, my company and its plan administrator are responsible to the IRS. If they accept the tester’s judgment without understanding it, and that judgment is wrong, then there are accountability issues.

How is it is reasonable to apply this “if condition”:

If the plan's eligibility provisions are more liberal than what is required by the statute, disaggregation may be a permissible option.

to a situation in which the plan’s service eligibility requirement is NOT more liberal than required by the statute?


410(b)(4)(B):

(B) Requirements may be met separately with respect to excluded group
If employees not meeting the minimum age or service requirements of
subsection (a)(1) (without regard to subparagraph (B) thereof) are
covered under a plan of the employer which meets the requirements of
paragraph (1) separately with respect to such employees, such employees
may be excluded from consideration in determining whether any plan of
the employer meets the requirements of paragraph (1).


410(a)(1)(A):

(1) Minimum age and
service conditions
(A) ) General rule
A trust shall not constitute a qualified trust under section 401 (a) if the
plan of which it is a part requires, as a condition of participation in the
plan, that an employee complete a period of service with the employer or
employers maintaining the plan extending beyond the later of the
following dates—
(i) the date on which the employee attains the age of 21; or
(ii) the date on which he completes 1 year of service.


The company’s 401(k) plan specifies age 18 and 1 year of service as participation requirements.

Employees of the company that do not meet the minimum service requirement of 410(a)(1)(A)(ii) are not covered under any plan offered by the company.

How is it then reasonable (or legal) to take a database of plan participants (all those who are eligible to defer regardless of whether they are or are not doing so) and disaggregate some who purportedly have less than a year of service when they had to have a year of service to be in the list in the first place?

[410b]

PAL,

Thanks for your response. I need to print and read that document.

The sentence you pointed me to leads to another question regarding dates.

It seems like I have learned something more about 401(k) testing each year I have pulled data for the testing company.

Fiscal year plan, FYE April. Employer does enrollment and change paperwork quarterly, first of month. May 1, etc.

Because of this, the data that I prepared for the testing company originally worked off those quarterly enrollment dates. As I began to understand more about what was going on, I spent a long time one year studying plan documents and discovered that our plan said that we have a monthly entry date, so I am now preparing the test database for the testing company that way.

I think you are suggesting combining those dates in a way I had not considered before.

My company’s plan, people qualifying in Feb. enter in Mar. Qualifiers in Mar. enter in Apr. Qualifiers in Apr. do not enter until May, and I have excluded them from the test database. February and March qualifiers are not allowed to complete enrollment paperwork and actually make deferrals until May 1.

I am interpreting your comments to mean that I do need to continue including the Feb. and Mar. qualifiers (Mar. and Apr. plan entry) in the testing database. However, for the ADP/ACP test, I think you are telling me that they can be excluded/disaggregated and put in the separate test group along with the under 21 employees since they are not allowed to actually make deferrals until May 1.

[pmacduff]

FWIW - I think that the "otherwise excludable" in this instance is the age 21, not the year of service. (Your plan eligibility is age 18, right?)

I agree with those who say this is a more aggressive approach, but not necessarily incorrect or not allowed.

my 2 cents

[BG5150]

We have never had any problems disaggregating participants who were under 21 or would have entered the plan the next plan year if the plan used semi-annual entry dates.

Would the test have passed if everyone was included? Some practitioners just automatically disaggregate the populations as a default. If the plan will pass with everyone in it, then all is good.

For the excludable people: is the census data correct for those people? Be sure they are really excludable.

Also, for "service less than a year": You could actually have people with a year of service who could be excludable. Say a person is hired in August 2006. One year anniversary is August 2007. With monthly entry dates, the person would come into the plan on Sep 1. With semi-annual entry dates, the person would come into the plan 1/1/08. So even though the person is eligible for the plan and has more than a year of service, he or she can be tested separately.

[410b]

Tom Poje,

Thanks for your response.

In regard to the otherwise excludable, I would be interested in your opinion on the "carving out issue" too. I tried to express my thought struggle on that more clearly in my response to Jim Chad.

Your comments are helpful.

I have been focused on the idea of “getting someone to see the test is wrong”, which I believe to be the case with the service part of the exclusion. However a plan amendment could address the date issue for the age part of things. I think I need to be mentioning that as part of the solution too.

Again, setting aside the question of whether the service exclusion was correct and considering only the “date to use” question, I was asked if what our testing company is doing was “industry standard” practice. I think your response tells me that instead of 1 “standard” there is a range of accepted practice and that each practice has some logic to support it that can be traced back to a code section.

Your post is very helpful to me in understanding those different date choices.

The test failure that you mention is one of the reasons for my concern over this issue.

HCE % participation has been increasing. NHCE participation remains low. The ADP test failed 3 years in a row (appropriate corrective action was taken). The ADP test passes this year, only because of what I consider to be an incorrect procedure. A situation easily correctible now, much more embarrassing later, if it is wrong.

This post has been edited by 410b: Jul 2 2008, 12:30 PM

[410b]

pmacduff,

“… more aggressive approach, … not necessarily incorrect …”

Comment noted. Thanks.

BG5150

What are the service entry requirements for your plan, 1year or less than 1 year?

All employees would have failed ADP test for 4th year in a row.

Employee census data was created based on 18 and 1 year of service. Testing company exchanged emails with me on questions, fixed clerical errors, and then made exclusions from that database based on their criteria.

In regard to your last paragraph, from my post to Jim Chad above:
How is it is reasonable to apply this “if condition”:

If the plan's eligibility provisions are more liberal than what is required by the statute, disaggregation may be a permissible option.

to a situation in which the plan’s service eligibility requirement is NOT more liberal than required by the statute?

I cannot come up with a logical way to support the concept of reaching into a container and taking out something that was never put in in the first place.

This post has been edited by 410b: Jul 2 2008, 12:09 PM

[410b]

Tom Poje,

Reading again, seeing more that I missed.

Re "final thought" is that the date 6 months after (08/03/07) or the plan entry date after that 6 months period you mentioned (09/01/07).

[Mike Preston]

There is a confusion here. Maybe more than one. Maybe more than two.

1) The testing company has indicated that the actual entry dates are semi-annual (that would be 5/1 and 11/1).

Eliminates/disaggregates NHCEs with less than one year of service as of the last semi-annual entry date of the plan.
Eliminates/disaggregates NHCEs under age 21 as of the last semi-annual entry date of the plan.

But you have also indicated that "Plan entry dates are monthly."

Both can't be correct. Which is?

I'm going to assume that the entry dates for the plan are indeed monthly and that the testing company has used the semi-annual entry dates solely for purposes of determining whether a participant can be carved out.

As others have pointed out, you can not find agreement, even at the IRS, as to whether the superimposition of semi-annual entry dates solely for the purpose of determining whether an individual can be carved out is acceptable. Depending on what industry conference you might have attended in the past three years, you will find an IRS representative, possibly on tape, indicating one way and you will find another IRS representative, also possibly on tape, at another conference indicating the opposite. There are no citations I'm aware of which provide a more formal IRS view on the choice and the IRS view on both choices.

By the way, this issue has been around for years and years and the IRS has flipped and flopped on it for years and years. It is hard for me to imagine that the IRS, if push came to shove, would hold fast to either approach as being the only legitimate alternative. But as also pointed out, you, as the client, can ask that the test be done in the most conservative manner, if you wish, and the testing company should be able to do so, even if it is at additional cost.

2) Assuming that the testing company's description is solely for carve out purposes and that monthly entry dates are required by your plan's document, I also note that you have stated that the employer processes enrollment forms quarterly. If I understand you correctly, this is a qualification failure. That is, if an employee hired in February of 2007 would be eligible to enter the plan on 3/1/2008 and that employee is not given an opportunity to commence deferrals until 5/1/2008, then the plan has not operated in accordance with its terms. If this is a correct description of what is going on then the EPCRS (Employee Plans Compliance Resolution System) provides for a correction, which you should embark upon. This may involve corrections for many years and be quite expensive. If the correction is necessary and the plan sponsor finds the default correction prohibitively expensive, you might give consideration to applying to the IRS for an alternative correction under EPCRS. Competent legal counsel should be engaged if you find that correction is necessary, in either event.

3) You have indicated that there are certain participants who are being carved out, but you haven't given us an example birth, hire and work history for anyone carved out in this manner. You hit upon one category, that would be those hired in the last quarter of the year (2/1 through 4/30) who are not actually allowed to defer until the next year (5/1 following the 12 month anniversary of the 2/1 through 4/30 hire date) who definitely should be carved out if the actual entry dates are quarterly (but see the above item if the entry dates are actually monthly). If the entry dates are monthly, then those carved out would be those hired 4/1 through 4/30 of the year before. However, it is possible that those individuals aren't in your database at all.

Note that if the entry dates are indeed monthly and if the employer has been inappropriately excluding those from deferring once they have entered the plan until the next quarterly processing date, you will find that these individuals will be reflected in the final test exactly as if they were properly carved out. This is because the default correction under EPCRS is to have the employer make a corrective contribution to these individuals equal to the average of the NHCE deferral rate (with an adjustment for lost earnings). Hence, the actual NHCE average will be the same as if they were excluded.

So, can you confirm which is the correct definition of entry under the terms of the plan and whether or not the employer has inappropriately been precluding people from deferring between their actual plan entry date and the next quarterly processing date?

[410b]

Hi Mike,

Thanks for the response. I am going to answer you in two separate posts, because I am still not getting a good answer to my original basic question.

I am seeing the term "carve out" used in responses to my question, so I will use that here.

This statement showed up in one of my search hits this week:

"If the plan's eligibility provisions are more liberal than what is required by the statute, disaggregation may be a permissible option."

Our testing company's manual includes a statment along those lines. That same concept is referred to in some of the posts above, for example the last paragraph of BG5150's post.

It seems to me that that statement also implies a reverse.

If the plan's eligibility provisions are NOT more liberal than what is required by the statute, disaggregation IS NOT a permissible option.

My plan is NOT more liberal than the statute in regard to service requirements. It equals the plan. Therefore I don't think it is correct to do "carve out' based on a service requirement. I DO have to work on understanding the date issues which you and Tom Poje have raised. However, my first, basic, core question is not which date pattern is the proper one to use for a "carve out". It is whether or not "carve out" is correct at all for our plan. (Conceptually, it seems to me that the testing company is saying that they can take rocks out of a bucket I provide, even though I never put any rocks into it.) Because I can't follow the logic of the test "carve out" procedure, I have put myself in the position of saying to our Plan Administrator, our CPA, and our testing company that I believe it is incorrect to be doing a service based "carve out" at all and that it could cause our plan to be disqualified by allowing higher than proper HCE contributions. I am wanting to know if I have made a correct or incorrect analysis of that issue and what the support is for whichever answer is right.

If my analysis is correct, then I think we should be able to say to the testing company that they have not been performing code compliant tests on our account and that they should do proper tests at no additional charge to us. If I am wrong, I have a very red face and another annual learning experience about 401(k) testing. In either case my company has a better knowledge about the testing properly supporting HCE contribution limits.

[Blinky the 3-eye...]

My plan is NOT more liberal than the statute in regard to service requirements. It equals the plan. Therefore I don't think it is correct to do "carve out' based on a service requirement.

Ah, but your service eligibility IS more liberal because of the entry dates. You could have entry 6 months after satifying the 1 year of service requirement. You have monthly (or quarterly) entry dates. Does what everyone is saying make more sense now?

This post has been edited by Blinky the 3-eyed Fish: Jul 2 2008, 03:20 PM

[Kimberly S]

My plan is NOT more liberal than the statute in regard to service requirements. It equals the plan. Therefore I don't think it is correct to do "carve out' based on a service requirement. I DO have to work on understanding the date issues which you and Tom Poje have raised. However, my first, basic, core question is not which date pattern is the proper one to use for a "carve out". It is whether or not "carve out" is correct at all for our plan. (Conceptually, it seems to me that the testing company is saying that they can take rocks out of a bucket I provide, even though I never put any rocks into it.) Because I can't follow the logic of the test "carve out" procedure, I have put myself in the position of saying to our Plan Administrator, our CPA, and our testing company that I believe it is incorrect to be doing a service based "carve out" at all and that it could cause our plan to be disqualified by allowing higher than proper HCE contributions. I am wanting to know if I have made a correct or incorrect analysis of that issue and what the support is for whichever answer is right.

If my analysis is correct, then I think we should be able to say to the testing company that they have not been performing code compliant tests on our account and that they should do proper tests at no additional charge to us. If I am wrong, I have a very red face and another annual learning experience about 401(k) testing. In either case my company has a better knowledge about the testing properly supporting HCE contribution limits.

For the moment, ignore the entry date issue and focus on a more clearly defined piece of the puzzle. Your plan is more liberal than the statute because it allows people in at age 18, and the statute says you can make them wait until 21. Anyone in your plan who is 18, 19 or 20 can be excluded from the test because you didn't have to include them in the plan and therefore should not be penalized because you were generous and let them in.

Does that help clear up the concept of carving out certain people?

FWIW, testing by excluding "otherwise excludable" employees is an industry standard, albeit with some questions raised above about exactly how to define that group.

This post has been edited by Kim Sheek: Jul 2 2008, 04:48 PM