How many ways do you do 410(b) coverage testing before deciding to cor

[John A]

If a plan fails the ratio percentage test, it may still pass using the Average Benefits Test. However, the plan document may specify that if the plan fails 410(B) coverage testing, then additional participants will share in contribution until 410(B) is passed.

Before bringing in additional participants, do you generally do the Average Benefits test on both a contributions and benefits basis? Do you do the Average Benefits test both with and without permitted disparity? Do you do all the methods, including the Accrued to Date method?

Do you use Excel to do Average Benefits testing? What software do you use?

Thanks for any input.

[Guest]

plan document says 'if you fail 410(B) then bring others...'

if plan fails ratio test but passes avg ben test then you haven't failed 410(B) so you never get to the step of bringing in others.

I don't believe it works quite the same as an ADP test where you have 'options' available, but are not required to use them. e.g. shifting unused deferrals.

Based on your description, I would assume the plan fails ratio %, but the ratio % is greater than the safe harbor %, so you don't have to worry about the classification test.

Therefore you only have to pass avg ben % test. I suppose if testing on an allocation basis it would be fairly easy (if not imputing disparity).

If testing on an accrual basis, then an excel sheet would get a little more hairy. ok, a lot more hairy.

We use the nondiscrim test that comes with the pension software (Qtech/Relius), though I still verify a few people using an excel spreadsheet - just to make sure. Actually I use the spreadsheet to calculate additional amounts to pass failed tests, etc, so its not a one shot deal spreadsheet.

you have to update it every year to take into consideration the taxable wage base in order to impute disparity.

fun stuff, John.

Not sure what software you are using, there is nothing on it to do this testing???

[John A]

Tom, thanks for the reply - it's helpful.

We use about 3 different software packages, but mainly DATAIR and Trustmark. If the plan has always been on DATAIR and we've done the administration for the life of the plan, then DATAIR handles everything well. I'm not so sure about Trustmark's 410(B) capabilities. And I'm not so sure about DATAIR when it's a takeover situation. Sometimes it would be a lot cheaper to just add a person, rather than do a complex ABT, especially if it has to be done every way possible. Does Qtech/Relius automatically do the ABT on both a benefits and contributions basis, with and without permitted disparity, Accrued to Date, etc. (I believe DATAIR does)?

[Guest]

no accrued to date, but you do have the options of accrual vs allocation, imputing disparity, grouping accrual rates (though weak in this area - if I use this option, I do it off the system) and statutory exclusions. nothing for DB plans.

probably like Datair, it has its positives and negatives.

(If I was already running 2 or 3 softwares, I would be honest enough to say there is no way I would take on the expense and time to learn another system!)

[MWeddell]

A pension plan must have definitely determinable benefits and a profit-sharing plan must have a definite predetermined formula for allocating contributions. Hence, if you've got plan document language that adds more participants to prevent a 410(B) failure, it SHOULD be drafted in a manner that eliminates any discretion.

[John A]

MWeddell, I'm not sure I completely agree, and maybe I'm understanding my own question a little better. It appears to me that a plan sponsor has a choice of testing on a contributions or benefits basis. If the sponsor chooses to test on a contributions basis, the formula is then definitely determinable, and it appears to me that this choice can be made without ever checking on whether the plan would have passed on a benefits basis. Part of my question was whether or not the plan sponsor could make this choice and then do the test based only on this choice, or whether the sponsor was obligated to test on a both a contributions and benefits basis to establish a 410(B) failure. Are you saying that the employer does not have a choice, but is obligated to try all available testing methods?

[Guest]

personally I think the sponsor has to try all means to pass 410(B). In fact, I always figured that is what the client pays me to do, to figure out how to get the plan to pass, etc. well, of course, besides fudging the data! we have to keep our ethics. (I lost a client once because of that!)

I think where the question comes in is if plan fails, who do you bring in. most documents I have seen say start with the ee who terminated last, or had the most hours or whatever. that is definitely determinable.

It is also my understand that the IRS takes a dim view of document language that has the failsafe language to start bringing people in. I believe (of course I could be wrong) much of that language has been eliminated in newer documents.

[Dawn Hafner]

You also need to be careful here depending on how your document was drafted.

Many prototype documents provide that if you apply the "fail safe language" for applying accrual requirements (1000 hours, EOY requirement) that if the plan fails the PERCENTAGE test that you start pulling ineligible participants in based on the order listed in the document.

In other words, if you selected the fail safe coverage language the plan may not permit you to use the ABP test or the raio test, but you are stuck with the percentage test.

[MWeddell]

I agree that if the plan passes on a contribution basis (so that the fail-safe language is not triggered) then there's no need to also test on a benefits basis. I didn't mean to imply otherwise.

However, before additional contributions are made to additional participants, there shouldn't be any employer discretion. Also, no one working for the employer should have discretion or else the 411(d)(6) regulations tend to call this employer discretion. Hence, if the plan document language simply says "if the plan does not satisfy Code section 410(B) ..." then it should be interpreted as requiring one to try any possible testing methodology for which the plan might demonstrate that it satisfies 410(B). If you don't want to be obligated to test every which way, then the plan document ought to specify in what manner one tries to satisfy 410(B) before making the fail-safe contributions.

[Dawn Hafner]

I agree MWeddell about testing under all methods to pass 410(B). I just think some administrators are not even aware of how limited their choices are when they select the fail safe language in the adoption agreement. For plans that will want to use ABP an addendum to the adoption agreement is often necessary.

[MWeddell]

Dawn, your point is also correct. I was replying to John A's most recent post and didn't mean to imply any disagreement with your post.

[John A]

MWeddell,

The consensus does seem to be that the plan sponsor is obligated to try every possible method to meet 410(B) before using the fail-safe language unless the plan document language clearly indicates otherwise. Clearly, if any method is found in which the plan passes, no further testing is necessary. However, the consensus seems to be that if the plan fails on a contributions basis, the sponsor is obligated to go on to test on a benefits basis (again, unless the plan document clearly says to only test on a contributions basis - the sponsor cannot choose to test only on a contributions basis if that basis fails). The fail-safe language can only be used after every conceivable way of passing 410(B) has been tried.

[MWeddell]

I agree with your above post, John A. I think it's unanimous.

[John A]

This probably falls under the category of beating a dead horse, but what about the interest and mortality tables used to convert contributions to a benefits basis? Surely the plan sponsor can choose an interest rate and a mortality table (and whether or not to use salary history)to do the testing and then test on that basis, correct? The sponsor would certainly not be obligated to try several different combinations of interest rate and mortality table to do the testing (even though one combination might result in passing while another results in failing), would it?

[MWeddell]

John, you wrote "The consensus does seem to be that the plan sponsor is obligated to try every possible method to meet 410(B) before using the fail-safe language unless the plan document language clearly indicates otherwise." I think that includes examining various mortality and interest rate assumptions unless the plan document specifies what assumptions to use. I would think that the person doing the testing could use some judgment about whether there's any chance that a specific combination could improve the test and is not required to run every theoretically possible combination.

[John A]

I'm not sure this changes my mind about MWeddell being correct, but page 8.26 of the 1999-2000 ERISA Outline book states, "If the plan fails the ratio test, the employer has two choices - expand coverage to pass the ratio test or pass the average benefits test." This would seem to imply that the employer is not obligated to do the ABT and could choose to simply expand coverage if desired. However, I do not see any cite.

[Guest]

I would read that, try passing avg ben test, and if not then expand coverage.

see 1.410(B)-2(B)(1) A plan satisfies [coverage] if and only if it satisfies on of the test (B)(2) - (B)(7).

(B)(5) through (B)(7) are your exceptions found on question 3 of the schedule T, (B)(4) is for special ESOPs.

That leaves you either ratio or avg ben test.

If it passes one of them [even if you don't know it], I don't know how you can justify expanding the group covered.

From talking to others, I get the idea they think if you fail ratio %, you fail, end discusiion, and they don't realize the avg ben test.

I see this different, than say, the ADP test where you 'may' test separately otherwise excludables, or you ' may' shift unused deferrals.

[John A]

Tom,

I agree with you, but that is not the way the ERISA Outline Book says it. It says the employer has 2 "choices"; it does not say the employer must use 1 choice before getting to use the other. In this case, I believe the wording in the book is wrong, or at least misleading, since the employer does not get to choose (does not have "choices").

[Guest]

your point is understood, but by implication the following is also true:

when performing coverage I MUST always run the ratio percenatge test. I am not even allowed to consider another test unless I need to. then if I fail, I should proceed to other means.

But why this logic? If my plan passes Avg Ben Test, I wouldn't even need to run the ratio % test. (although since it shows up as part of the test, it only makes sense to perform the ratio percent test)

Or suppose my plan passes both Ratio and Avg ben % test. Which I would think applies to most plans. why is it mentally I feel I should check the box 'plan passes ratio percentage rather than avg ben test? both tests are valid.

but I can honestly tell you that even I mentally have this notion that says if I check one box I might as well send the form in on flourescent paper 'cuz the IRS is goona make me for prove the avg ben test.

(I would add this particular example is the same reason there are so many opionions on the Bible. Is it an either/or scenario or a both/and. Yes, you could read it one way, but it is just as easy to read it another way.)

[Archimage]

I am curious about putting another factor in this discussion. Let's say you finally pass coverage on a benefits basis. Would the gateway minimum apply in this situation where an employee was not eligible for the regular allocation due to the hours requirement but was eligible for the top heavy minimum?

[AndyH]

Archimage, the gateways do not apply for coverage testing.

[Archimage]

That was my thinking as well. Then again I initially didn't think you would have to give the gateway to participants that only received the top heavy or SHNEC. With all the fun gateway stuff I just wanted to be sure.