Jump to content

CPA says Gateway is only required contrib in a cross tested PSP


Recommended Posts

This is a takeover New-Comp Plan. We looked back at the 2012 and 2013 years and while the client made a gateway contribution, in each year, the rate-group test prepared by Relius showed a required contribution of around 9% of nHCE pay in each year. The CPA says that is OK because the pass the 410(b) Ratio Percentage test.My response was that the Plan is cross tested,hence it must pass BOTH the rate group tests and the average benefits test...he adamantly states "No way". I'd like a second opinion.

Here's some background and a few questions.

The Plan defines 2 Classifications [Rate groups] –Principals and all others. The Plan Document states the allocation method for non-elective contributions uses the “Participant Group Allocation Method”.

There are 2 principals and 6 nHCEs, one of whom was hired 5/1/2013. The Plan permits deferrals for employees with 6 months of service[entry date is monthly]. A Safe Harbor NE contribution is required for all employees meeting that same requirement. Class based Non elective contributions are also allocated to the same employees.

Each year the maximum 415 contribution is calculated for the Principal with the lowest compensation. That percentage is then applied to both Principals’ compensation. If a Principal with higher compensation receives a prospective allocation greater than the 415 limit, such Principal’s allocated contribution is capped at the 415 limit.

A gateway contribution is then calculated for the nHCEs equal to the lesser of 5% of compensation or 1/3 the highest Principal’s allocation percentage.

The Plan is a Non-elective Safe Harbor Plan so the 5% gateway is satisfied through the combination of 3% Safe Harbor + a 2% Non-elective contribution.

The Plan is then subjected to Rate-group testing under 401(a)(4).

Question 1

Must this Plan be tested under IRC 401(a)(4), The General Nondiscrimination Test?

If so, must the General Nondiscrimination Test include “Rate Group testing”?

Question 2

Are there any options available to pass general non-discrimination not using rate group testing?

Question 3

What does it mean for an otherwise excludable employee to be “separately tested”? If the Plan required only 6 months of svc for a Safe Harbor Non-elective, would that same Otherwise Excludable Employee be entitled to a gateway contribution?

Question 4

Is a Gateway contribution required for employees that have been employed less than one year/age 21 when the Plan requires a Safe Harbor Contribution for such employee? May those less than one year/age 21 employees be excluded from the 401(a)(4) General Test?

Question 5

The General test shows that approximately 9% of payroll must be contributed for nHCEs in order to pass the rate group test. The employer , in 2013 and 2012 contributed less than the required 9%. The 2013 tax return has been filed. Can one make self corrective contributions for employees in 2014 to correct that error or must some other correction procedure be used.

Link to comment
Share on other sites

Question 1 Must this Plan be tested under IRC 401(a)(4), The General Nondiscrimination Test? If so, must the General Nondiscrimination Test include “Rate Group testing”?

The plan has allocations for HCEs and NHCEs, so the allocations have to satisfy 401(a)(4) somehow. There are many option in 401(a)(4) to prove a plan passes. Your situation seems to describe a typical plan that tests the contributions on a benefits basis (cross-testing), thus the gateway requirement. Keep in mind, the gateway only allows the plan to test the contributions on a benefits basis, it does not mean the plan will pass testing on a benefits basis. Sometimes additional profit sharing above the gateway is needed to actually pass the test. Many folks, consultants and advisors call the entire amount gateway even though, in your example, the 5% is really the gateway and the rest is profit sharing.

Question 2 Are there any options available to pass general non-discrimination not using rate group testing?

Look at 1.401(a)(4)-2(b). you'll find the testing is bypassed for plans that allocate pro-rata as a percent of pay, plans that allocate a uniform dollar amount per person (regardless of pay but not over 100% of pay), plans that use integration - an allocation formula based on permitted disparity, and plans that provide a uniform amount per unit of service in the plan year. Then look at 1.401(a)(4)-8 (cross-testing). In section (b)(1)(i)(B)(1) and (b)(1)(i)(B)(2), you'll see the broadly available allocation rates exception and the gradual age/service schedule exceptions.

Question 3 What does it mean for an otherwise excludable employee to be “separately tested”? If the Plan required only 6 months of svc for a Safe Harbor Non-elective, would that same Otherwise Excludable Employee be entitled to a gateway contribution?

Employees who are not age 21 and have not completed a year of service, or have done that but id not meet an entry date yet, are considered as "otherwise excludable employees" (the OEE group). Meaning, the statute allows those people to be excluded from the plan, but the plan chooses to let them in before 21 and 1 for some or perhaps for all purposes. These employees, if eligible for the plan, have the same rights as the other plan participants, but certain items do not have to be applied to them. For example, although the top heavy minimum does apply to them, the minimum gateway does not apply to them. So, if they are not eligible for profit sharing, then they will not get the gateway (does the plan require age 21 and 1 for PS purposes?). If the plan document you described gets them in after 6 months for PS as well, then since all of these NHCEs are in one group for allocation purposes, then they would have to get the same as all the NHCEs.

They also could be describing a component testing option (restructuring) when they say "separately tested". That's in 1.401(a)(4)-9.

Question 4 Is a Gateway contribution required for employees that have been employed less than one year/age 21 when the Plan requires a Safe Harbor Contribution for such employee? May those less than one year/age 21 employees be excluded from the 401(a)(4) General Test?

No, the gateway is not required for the OEE group. If the plan has all NHCE in one rate group and it requires 21/1 to get into the PS portion of that rate group, then those only getting SH and or TH are not entitled to get the PS under the terms of the plan as you describe.

Question 5 The General test shows that approximately 9% of payroll must be contributed for nHCEs in order to pass the rate group test. The employer , in 2013 and 2012 contributed less than the required 9%. The 2013 tax return has been filed. Can one make self corrective contributions for employees in 2014 to correct that error or must some other correction procedure be used.

There are a lot of ways to pass nondiscrimination testing. Just because a report says 9% is needed, I'm fairly certain that someone could easily come up with a report that shows they pass at a lower rate, maybe even at just the 5% gateway minimum if enough historical data can be provided to run multiple testing options. Unless, of course, the report stating 9% was done by Tom Poje, Tom Finnegan, Larry Deutsch, Kevin Donovan etc. etc.

Link to comment
Share on other sites

ok, I see while typing this John has responded. perhaps this will simply add to the comments

the gateway rules are found in 1.401(a)(4)-8
which is, quite conveniently entitled Cross testing
the opening sentence
This section provides rules for testing defined benefit plans on the basis of equivalent employer provided contributions and defined contributions plans on the basis of
equivalent employer-provided benefits...

so, lets suppose you have a plan in which everyone is in their own group.
for a given year everyone receives 5%.
you don't have to test on an accrual basis, you could test on an allocation basis (which would then NOT be cross testing, and so no gateway.

one of the gateways is 'broadly available' rates, which basically says if each allocation group passes you have satisfied the gateway. so you could have a group of NHCEs and HCEs at 9% and another at 3%.
personally I think their is a catch, missed by most (including many documents!)
the LRMs (Language Required Modification (or whatever that stands for)

had the following:
LRM #94
(Note to reviewer: There are other gateways that may be used in order for a defined contribution plan to cross-test using equivalent benefits under 1.401(a)(4)-8(b). The plan may provide for a different gateway other than the minimum allocation gateway (for instance, the broadly available allocation rate

requirement of Regulations section 1.401(a)(4)-8(b)(1)(iii) or the gradual age or service based allocation rate requirement of section 1.401(a)(4)-8(b)(1)(iv)); however, sample language for other gateways is not provided herein. If a sponsor wishes to use other gateways, it is important to ensure that the benefits provided under the plan remain definitely determinable. In order for plan benefits to remain definitely determinable, the plan document should specify which gateway is used. The plan document could allow adopting employers to elect between different gateways, but in order to provide definitely determinable benefits it is not sufficient for the plan document merely to specify that one of the gateway requirements will be satisfied.)
(Note to reviewer: No section 401(a)(4) failsafe language is allowed. The plan must pass nondiscrimination testing based on Income Tax Regulations sections 1.401(a)(4)-1 through 1.401(a)(4)-13.)

maybe they removed this requirement, or maybe it will show up in the next round of restatements, or whatever, but apparently documents 'should' specify what gateway will be used

Link to comment
Share on other sites

John, take the case where the wife has pay of $18,000 and contributes $17,500 to a non-elective safeharbor plan....I have always felt that the Plan had to pass the Plan Level Average Benefits Test [including all ee+er contributions]...before it got to rate group testing....so in the above example under most circumstances, it would fail that test.

Everything I read says it must pass BOTH average benefits and rate group.

Any thoughts??

Link to comment
Share on other sites

For passing nondiscrimination regarding the nonelective allocations in the plan, if each rate group has a 70% or greater result, then the average benefits test is not needed for nondiscrimination testing regarding the nonelectives. This is a fairly common plan design: a low-paid spouse who works for the owner defers the max but gets a small (if any) profit sharing.

If any rate group is under 70%, then to pass 410(b), the average benefits percentage test must pass and 1.401(a)(4)-2(c )(3) basically allows the nondiscriminatory classification test portion of 410(b) to be the midpoint between the safe and non-safe harbor percentages.

Look at 1.401(a)(4)-2(c )(1): "The employer contributions allocated . . . are nondiscriminatory in amount . . . if each rate group under the plan satisfies section 410(b)."

Is this not included in what you've read?

Link to comment
Share on other sites

Thanks John, but I can tell you that its very rare to find a rate group that passes at 70%...most pass at or above the midpoint because that's the fundamental purpose of the classification....are you saying all your rate groups pass at 70% or above???

Link to comment
Share on other sites

Example:

2 HCEs: Owner at max pay, spouse at $25,000.

2 NHCEs: one young, one old.

Each participant is in their own rate group (grouping by any other means limits your testing flexibility)

Component plan #1: Owner and young NHCE: passes on a benefits basis: 100% of the NHCEs in that component plan are in the HCE benefit rate group.

Component plan #2: Spouse of owner and older NHCE: passes on a contributions basis: 100% of the NHCEs in that component plan are in the HCE allocation rate group.

Each rate group is over 70%. No average benefits percentage test needed.

Gateway required for the whole plan however because cross-testing in one of the components.

Link to comment
Share on other sites

Thanks John, but I can tell you that its very rare to find a rate group that passes at 70%...most pass at or above the midpoint because that's the fundamental purpose of the classification....are you saying all your rate groups pass at 70% or above???

For what it is worth I have general tests that pass the rate group test with all the ratios over 70% all the time.

It is just a matter of how aggressive your plan wants to be.

I had an ESOP once that allocated the non-Elective ER based on this formula:

25% of the contribution was allocated YOS for vesting (for elig employees)/YOS for vesting for everyone in the numarator

75% of the contribution was simple comp/comp formula

Every year it pass the rate group test with all the groups being over 70%. What it did was reward long term employees but the rate group test was fairly easy to run.

Link to comment
Share on other sites

Yes, many times the demographics prove that a much less expensive option is to pick the one or two specific NHCEs and give them a little more PS if that gets all your groups over 70%.

Sometimes it's as easy as saying: the profit sharing allocation this year is 5% of pay, but not less than $500. Of course the $500 minimum only affects a few of the lowest paid NHCEs, bumping their allocation rates into one or more of the HCE allocation groups.

Link to comment
Share on other sites

OK, I did a little research and ran your idea by some other actuaries...here's one that reflects what they all said...

"What you are talking about now, Charles, is referred to as "restructuring".

When you restructure into separate testing groups, each testing group istreated as if it were the only plan. And, yes, each testing group can use adifferent method to pass 401(a)(4).

A warning, though. Each restructured plan must independently pass 410(b).

Most of the time we pick and choose who is in each plan for testingpurposes. Unless, magically, the folks we pick and choose constitute areasonable business grouping you will be precluded from using the ABPT.Hence, when testing whether the plan being tested passes 410(b) you almostalways use the 70% threshold (even if the plan satisfies the ABPT with roomto spare). "

Here's what Jeff Wadle says [Jeff's a serios regulations guru]...

"I do not think there is any way you can argue for using different methods for determining the accrual rate for different rate groups – the consistency rule should stop that.

However, you could use the restructuring rule to divide the plan into different separately-tested component plans. And test one component plan using EBARs and the other just using allocation rates. Maybe this is what was really being suggested? Note that this is somewhat different than using different methods for different rate groups. If there was an NHCE with both the highest EBAR and the highest allocation rate, they would have to be a zero in one of the component plans.

Restructuring usually works when you have a few younger HCEs like children of the owner who would have EBARs way too high to fit in a rate group. But have allocation rates below a number of NHCEs.

A separate question is whether the rates used for the ABPT can be different than those used for rate group testing. My opinion is no, they cannot be, although others disagree. So, if I did restructuring into two component plans – one using benefits basis and EBARs and one using DC basis and allocation rates, I would calculate two different ABPTs for use with those component plans (if the rate groups did not pass at 70%). When I have restructured this way, often the EBAR component plan ABOPT works fine – but not the allocation rate component. So would then just make sure rate groups in allocation rate component pass at 70%. "

So I assume you are "restructuring"????

Link to comment
Share on other sites

not necessarily. you (and apparently whomever you discussed this) are forgetting (or not realizing an important piece)

the avg ben % test includes all contributions (deferrals + match + profit sharing)

whereas your rate group test only includes profit sharing.

or for a real simple example, imagine a young owner child deferring 10,000. it will blow the avg ben pct test out of the water.

but if that child receives little or no profit sharing, the rate group for the child would be above 70%

I would add a lot of people forget about this possibility happening..

Link to comment
Share on other sites

Tom is right, of course. The example I provided was both restructuring and has over 70% in each rate group. Let's do an example without the restructuring:

2 owner HCEs, age 55 and age 45

4 NHCEs

Testing the entire plan on a benefits basis (the ebar for the younger owner is higher than the ebar for the older owner)

Rate Group 1 (Age 55 HCE):

3 out of 4 NHCEs are in the rate group of the age 55 HCE, and both HCEs are in that rate group, so that group is: ( 3 / 4 ) / ( 2 / 2) = 75%

Rate Group 2 (Age 45 HCE):

If only 1 out of 4 NHCEs are rate group of the age 45 HCE, you would get ( 1 / 4 ) / ( 1 / 2) = 50%

That is not 70%, so you can either: run the average benefits percent test, or allocate more $ to just one of the other three NHCEs.

In this example, the plan only needed to give $200 to one of the other 3 NHCES to get them in the age 45 HCE rate group. In contrast, it would cost the employer an extra $3,000 in employee benefit costs to get the average benefits test to pass.

So with an extra $200 allocation, we now get 2 out of 4 NHCEs in the age 45 HCE rate group, so that group is: ( 2 / 4 ) / ( 1 / 2) = 100%

Both rate groups are over 70% so the average benefits percentage test is not required. Of course the employer could still say, no let's allocate $3,000 extra instead of just $200 to one person.

Sometimes the best option is to combine restructuring with cross-testing. Also look at the actual dates of birth: testing using age last or age nearest can make a noticeable difference - and be sure to impute disparity if the HCE comp exceeds the integration level.

Remember to try testing using average compensation if the owner compensation drops and to look at the possibility of testing using compensation from date of entry if that will help you (or apply both of those assumptions).

Or you can test using compensation with or without including deferrals in the compensation definition.

Will the top paid group election affect the results - may want to look at that sometimes, if it's not past the end of the plan year.

Even more fun is to test using the accrued-to-date method (using the annual method is an alternative).

If a plan needs the average benefits test to pass, you also have the option of averaging the ebar for any individual in the average benefits percentage test (rare).

And don't forget to check out the rate group banding option (rare, but more useful in larger plans). There are more testing options to name here, but who's really still reading this at this point anyway? Oh, and if your document is written without flexibility for testing, then forget most of that above - you'll have to follow the document and miss all the fun (sad face).

When you add a DB plan (traditional or cash balance) into the mix, you also may need to try varying assumptions are differing crediting rates, A.E., mortality - sometimes 8.5% GAM71 gets you to a lower NHCE requirement, other times it's IAM83 Female with a post retirement rate of 7.5% (especially if using the DB benefits to satisfy a portion of the gateway). And watch out for the combined plan deduction limits if the DB plan is not subject to PBGC.

Link to comment
Share on other sites

  • 4 weeks later...

you forgot that the Wolverines used to be dependable on fielding a good team every year.

my only point was that it is possible restructuring was not needed or even be referred to.

I have had co-workers who look at the test results and since the plan fails avg ben pct test that the plan fails testing. but the tests in question passed ratio % for each HCE.

Link to comment
Share on other sites

John, not to worry! :-)

Tom, it seems that they recruit folks who are fond of sucker punches these days. With Miller down for the year it looks like Michigan State is all that stands in their way within the conference.

As far as what I forgot, it is hard to argue that I forgot to take into consideration something that I specifically said is usually irrelevant.

Besides, you guys are discussing rate group testing and I was discussing the requirement that each component plan satisfy 410(b). As you know, a completely different kettle of fish.

I'll fade back into the background now. Cheers.

Link to comment
Share on other sites

Remember to try testing using average compensation if the owner compensation drops and to look at the possibility of testing using compensation from date of entry if that will help you (or apply both of those assumptions).

John, remember that the two can't be combined. If you are using testing compensation from date of entry you have to be using annual comp. If you are using average compensation then you can't use comp from date of entry.

Link to comment
Share on other sites

Remember to try testing using average compensation if the owner compensation drops and to look at the possibility of testing using compensation from date of entry if that will help you (or apply both of those assumptions).

John, remember that the two can't be combined. If you are using testing compensation from date of entry you have to be using annual comp. If you are using average compensation then you can't use comp from date of entry.

But you can exclude certain 12-month periods of compensation if, during such periods, the plan disregards the employee's compensation for determining benefits - see 1.401(a)(4)-3(e)(2)(ii)(A).

See also 1.401(a)(4)-3(e)(2)(ii)© for allowing certain months of compensation to be excluded as well, but I've rarely had to apply that one.

And, also, accrued-to-date testing must use average compensation. See 1.401(a)(4)-3(d)(1)(i) and 1.401(a)(4)-3(e)(2)(ii)(A), with exception for certain formulas.

Link to comment
Share on other sites

Remember to try testing using average compensation if the owner compensation drops and to look at the possibility of testing using compensation from date of entry if that will help you (or apply both of those assumptions).

John, remember that the two can't be combined. If you are using testing compensation from date of entry you have to be using annual comp. If you are using average compensation then you can't use comp from date of entry.

But you can exclude certain 12-month periods of compensation if, during such periods, the plan disregards the employee's compensation for determining benefits - see 1.401(a)(4)-3(e)(2)(ii)(A).

See also 1.401(a)(4)-3(e)(2)(ii)© for allowing certain months of compensation to be excluded as well, but I've rarely had to apply that one.

And, also, accrued-to-date testing must use average compensation. See 1.401(a)(4)-3(d)(1)(i) and 1.401(a)(4)-3(e)(2)(ii)(A), with exception for certain formulas.

Things can get very confusing in this area, real quick. I think we have reached that point in this discussion. The use of the word "But" seems to indicate you disagree with something. However, the first two things you mention are generally associated with drop out years and drop out months. I wasn't discussing either. Maybe I'm just not understanding what you are disagreeing with.

Link to comment
Share on other sites

It's just a reminder, not a disagreement. Many times participants enter on the first day of the plan year and the plan is written to ignore the service prior to that plan year of entry for benefit accrual purposes, in those cases the compensation prior to entry can be excluded (and sometimes that actually helps). The observation of this exclusion may lead some to think compensation prior to entry can be excluded, when really it's actually certain 12-month periods can be ignored for averaging compensation, which is perhaps why you were pointing it out. My post was only trying to lightly touch upon a small portion of the myriad of testing possibilities for Charles to think about.

It appeared the percenption was that the plan required 9% for all NHCEs. From my experience, most plans have each participant in their own allocation rate class. Thus, giving all NHCE the same percent of pay is an employer decision and is only needed when the plan has very few NHCEs and all are needed to be in a certain HCE's rate group. This particular case perhaps has very few NHCEs?

Also, one main question asked was about the average benefits test (for 401(a)(4) purposes). So we pointed out that test is not required when each HCE rate group meets the 70% threshold, and included a couple of ways to get a plan to that, one of which can include restructuring.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...