Safe Harbor Nonelective Plus Additional Nonelective

[austin3515]

3% SHNEC to all people eligible for 401(k) plus a 2% regular profit sharing contribution, integrated with SS but at 70% of the TWB. The owner and sole HCE is the only employee over the TWB. He earns $115K, with the TWB at 87K (i.e., integration level is ~61K. For the latter formula, you must have worked 1,000 hours during the Plan year.

According to the ERISA Outline Book, if there are different allocation conditions for two independent design based safe harbor formulas, the design based safe harbor is not valid and rate group testing must be performed.

The ERISA Outline Book goes on to say that if the additional PS contribution passes coverage on its own, then the rate group test will be passed (i.e., there is just one rate group).

When I run the rate group test it fails. The reason being when permitted disparity is imputed to his allocation rate, it assumes the integration level is $87,000. Because the integration level in my example is at 70% it's as though the plan is "super integrated." This results in the owner having a higher allocation rate than everyone else. I already verified that if the integration level was 100% then then his allocation is the same as eveyone elses and the rate group does pass. Am I missing something? OR are the ERISA Outline Book, and Relius both wrong????

What am I missing??? Is there an adjustment to imputing disparity when the integration level is a fraction of the TWB?

[Blinky the 3-eyed Fish]

There is no special adjustment for the general testing when the formula's integration is less than the TWB. That is why it is possible for a safe harbor integrated formula with the integration level below 100% of the TWB to actually fail the general testing. Of course they are safe harbors, so they never need to go there.

Now in your case, the situation you have is one that has been addressed on these boards before. I know I chimed in if that helps to narrow your search.

But basically to summarize past discussions, this type of situation with a SHNEC is very similar to the TH allocation situation. However with TH you are specifically allowed to treat people who only get the TH as not benefiting for coverage in order to maintain the safe harbor formula. With SHNEC that rule is not available, but I truly believe that the law has not properly caught up with the SH provisions. It is logical to be able to apply the same principles to SHNEC as to TH, although technically, there is nothing in the law that allows for this.

[austin3515]

So basically, under this scenario my plan is not in compliance with 401(a)(4)?

[Blinky the 3-eyed Fish]

First, I didn't read your first post carefully before I answered. My answer related to a situation where someone did not receive the PS contribution, but did receive a SHNEC.

But now that I am with you, you are correct that your formula does not have a safe harbor design and must pass general testing. If you fail, then consider an -11(g) amendment to correct the failure and perhaps a formula change to put the integration level at 100% of the twb.

[austin3515]

For not reading carefully enough you are forgiven...

For giving the answer I didn't want to hear, now that is another story altogether...

[jquazza]

I don't understand why you say you would pass at 100% of TWB integration level but not at 70%?

When you do your rate group testing, you can use permitted disparity, whether your integration level is 100% or 20%. The only thing that changes is the maximum disparity % (5.7%, 5.4% or 4.3%) In your case, it looks like it wouldn't matter anyway since you can only apply the permitted disparity on the additional 2% (remember, you're not allowed to apply permitted disparity to the SH contribution) and therefore the max disparity % should be 2% so no problem there. If you calculated your allocation with a 4.3% disparity, that's why you're failing, otherwise, if all the NHCEs make less than 61k+ and at least 70% of NHCEs benefiting at eh 5% rate, you should pass. If no, (document permitting) then can you pass with the ABT? If not, you can try cross testing.

[Blinky the 3-eyed Fish]

I don't understand why you say you would pass at 100% of TWB integration level but not at 70%?

In your second paragraph you acknowledge that you can fail at 70% of the twb integration, yes? 100% will guarantee that you pass on a contributions basis by imputing permitted disparity if you work through the math to prove that fact.

[austin3515]

Here's my math (my example is a 6/30/03 PY):

Owners Wages: 115,000

Integration Level: 87,000*70% = 60,900

SHNEC Contribution for owner: 115,000*.03=3,450

PS Compensation for Owner (w/Integration) = 115,000 + (115,000 - 60,900) = 169,100

PS Contribution for Owner = 169,100*2%=3,382

Impute Disparity on Owners PS:

1) 3,382/(115,000 – 50%*87,000) = 4.73%

2) (3,382 + 5.7%*87,000)/115,000=7.253

Lesser = 4.73

Plus 3% SHNEC, total allocation rate = 7.73% for the owner.

All other NHCE’s (who for simplicity all earn the same amount).

Total Wages: 30,000

Total SHNEC: 900

Total PS: 600

Imputed Disparity:

Allocation rate is 5.7% or less, so 2% x 2 = 4%

Plus 3% SHNEC = 7% allocation rate for NHCE's

Therefore, the Owners rate group will fail because no NHCE’s benefit. I think we all agree that the whole point of imputing disparity to a formula integrated at 100% is to give everyone the same allocation rate, so I won't go through the math on that piece of it.

In this example, it just so happens that the gateway requirements seem to be satisfied so I could use cross testing, but let’s assume that I CANNOT.

I would like to think that the imputing disparity rules are adjusted if the integration levels are less than 100%, but if they are, neither me or Relius seems to be aware???

[Blinky the 3-eyed Fish]

I would like to think that the imputing disparity rules are adjusted if the integration levels are less than 100%, but if they are, neither me or Relius seems to be aware???

Asked and answered. The imputing disparity rules are not based on the plan's formula so why do you think the integration level of the formula should have any impact on the calculation? Or are you calling for a regulatory change?

[austin3515]

Blinky (or should I say Jack McCoy? Don't you have to object first?), I agree with you, but it seems that there should be a special exception carved for this situation--i.e., it's a safe harbor formula to integrate at less than 100% of the TWB, so there should be a way to get imputing disparity to equalize everyone's rates if the integration level is less than 100%. But the point is moot because there is no such exception that I am aware of...

[Blinky the 3-eyed Fish]

The solution is for you to run for Congress, get on a committee and pursuade your fellow legislators to enact changes to this perceived injustice. But you better get cracking or your crusade will be delayed for 2 years. Your slogan:

Pensions for All! Yeah Baby!

[austin3515]

Smashing...

[austin3515]

Just put a call into Corbel for support. I'll let ya know what they say.

[Tom Poje]

why not just read 1.401(a)(4)-7 which describes in detail how to impute disparity.

in particular paragraph (b)...

if the plan took into account full disparity and used the TWB as the integration level.....this is used to determine if plan...satisfies the general test....and average benefits percentage test...

there are no exceptions.

at the same time, why not argue that you should be able to use 6.2% rather than 5.7% since that would reflect the actual Social Security percentage that is taken out from a paycheck.

[austin3515]

http://www.ssa.gov/pubs/10035.html

The link above indicates that 85% of the payroll taxes are allocated to OA insurance and 15% to Disability insurance (i.e, 6.2 x .85 = 5.27% = OA insurance). 401(l)(2)(A)(ii)(II) clearly states that it should be the rate attributable old age insurance, or 5.7% whichever is greater. How can we use the 6.2%?

[Tom Poje]

you can't. the original regulation would have increased the % as soc sec went up. in fact, in 1988 it went from 5.7 to 6.05, but in 1989 it was set back (somewhat permanently) at 5.7.

no, I don't recall running a plan at the 6.05%, I am looking at a table showing the %s.

[austin3515]

To clarify then, this plan design does not work and will never work?

Isn't it a little scary that a nonstandardized prototype could so easily violate the design based safe harbor requirements? There must be a ton of plans out there with the same problem, no?

Thanks Tom, as always!

[Blinky the 3-eyed Fish]

To clarify then, this plan design does not work and will never work?

It's not a safe harbor, but that doesn't mean it doesn't "work".

[austin3515]

I thought we "proved" that it will never work? Whoever has the highest comp will be in a rate group all on their own, that rate group will have the highest allocation rate, and no NHCE's will be in there, because none of them earn more than the TWB (okay maybe a few do). So therefore, the rate group will never pass (absent cross-testing, assuming gateway is satisfied).

[Blinky the 3-eyed Fish]

First, never say never. Second, the person with the highest comp could be an NHCE. Third, I didn't say absent cross-testing. Fourth, purple monkey dishwasher.

[austin3515]

But don't you think that this has to be a common oversight? Isn't it troublesome that a prototype can have such a treacherous 401(a)(4) pitfall?

[Blinky the 3-eyed Fish]

Yes, it keeps me up at night.

But seriously, I agree, but that is why it is important to know the rules, otherwise many aspects of plan administration can be treacherous, like the jagged cliffs of Acapulco.

[austin3515]

You have a way with words... Purple Monkey Dishwasher????

[Blinky the 3-eyed Fish]

There is some that will get that line, while others will not. The first to respond correctly to the meaning of the reference wins nothing.

[austin3515]

Somethign with the game "telephone," right? Is that how you get them to repeat whet they said???

This thread is falling apart...

[AndyH]

http://www.geocities.com/SoHo/Den/6460/1999mar31.htm

[austin3515]

Restructure the Plan's into two component plans:

Plan 1: Employees getting SHNEC only,

Plan 2: Employees getting SHNEC and integrated contribution.

Both are design based safe harbors.

Plan 1 and Plan 2 must pass coverage independently.

Another chapter is closed...

[Blinky the 3-eyed Fish]

I don't think your proposed solution resolves anything. You still have the only person above the TWB is an HCE. You ran through the math to show that the people below the TWB have a lower contribution percentage imputing permitted disparity. So that leaves your restructuring to mandate that the HCE is the only one in his restructured plan. That doesn't pass coverage though.

[Kevin C]

Interesting thread!!! Sorry, but I feel the need to stir the pot just a bit more.

1. With the 3% non-elective safe harbor contribution included in the general test, doesn't Notice 98-52 prohibit you from using imputed permited disparity in the general test?

2. Aren't standardized AND nonstandardized M&P plans required to use 401(a)(4) safe harbor allocation methods? The situation being discussed easily fits into the adoption agreements for the both the standardized and NS protoype documents we use.

When the logic starts telling us that we have to 401(a)(4) general test some plans using standardized prototypes, it seems a good time to step back and look at the big picture. Did the IRS make a huge mistake in how they integrated safe harbor 401(k) provisions into the M&P program? Or, did we miss something in getting here? I have a difficult time accepting the idea that we can have a standardized prototype plan that can possibly fail 401(a)(4).

I'm starting to wonder if the IRS considers the 3% non-elective safe harbor contribution as a part of the CODA. Consider the following:

401(k)(12)(A) IN GENERAL. --A cash or deferred arrangement shall be treated as meeting the requirements of paragraph (3)(A)(ii) if such arrangement --

401(k)(12)(A)(i) meets the contribution requirements of subparagraph (B) or ©, and

401(k)(12)(A)(ii) meets the notice requirements of subparagraph (D).

[Blinky the 3-eyed Fish]

Quick comments

1. You can't impute permitted disparity on the SHNEC, but it doesn't say you can't impute permitted disparity on other contributions just because you have SHNEC.

2. Nonstandardized prototypes don't have to be safe harbors. Now I don't work with standardized DC prototypes, but it sounds like there is something missing.

I'm not sure what you mean by the CODA comment. If you are saying SHNEC should be classified as a deferral and not a nonelective contribution, well then I disagree.

[Kevin C]

I'll try to be more specific.

From Notice 98-52:

B. Use of Safe Harbor Nonelective Contributions to Satisfy Other Nondiscrimination Tests

A safe harbor nonelective contribution used to satisfy the nonelective contribution requirement under section V.B.2 may also be taken into account for purposes of determining whether a plan satisfies §401(a)(4). Thus, these contributions are not subject to the limitations on qualified nonelective contributions under §1.401(k)-1(b)(5)(ii), but are subject to the rules generally applicable to nonelective employer contributions under §401(a)(4). See §1.401(a)(4)-1(b)(2)(ii). However, pursuant to §401(k)(12)(E)(ii), to the extent they are needed to satisfy the safe harbor contribution requirement of section V.B, safe harbor nonelective contributions may not be taken into account under any plan for purposes of §401(l) (including the imputation of permitted disparity under §1.401(a)(4)-7).

My understanding of the general test is that all employer contributions are tested together in total. Are you saying that you apply permitted disparity to one piece of the total and not the other in the same test?

From Rev Proc 2000-20, Section 3:

.09 Changes to General M&P Plan Requirements --This revenue procedure makes several changes and clarifications to the requirements that apply to all M&P plans. Significant among these are the following:

1 Rev. Proc. 89-9 and Rev. Proc. 89-13 prohibited the issuance of opinion and notification letters for plans that contain or may contain multi-tiered benefit structures. This prohibition has been reformulated as a general requirement that the allocation or benefit formula in a nonstandardized M&P plan must satisfy the following uniformity requirements of the regulations under §401(a)(4) pertaining to safe harbor plans. In the case of a nonstandardized defined contribution plan, the allocation formula must be a uniform allocation formula, within the meaning of §1.401(a)(4)-2(b)(2) of the regulations, or a uniform points allocation formula, within the meaning of §1.401(a)(4)-2(b)(3)(i)(A). In the case of a nonstandardized defined benefit plan, the benefit formula must satisfy each of the uniformity requirements of §1.401(a)(4)-3(b)(2). In addition, each nonstandardized plan must give the employer the option to select total compensation as the compensation to be used in determining allocations or benefits and each nonstandardized defined benefit plan must automatically or by option allow the adopting employer to satisfy one of the design-based safe harbors described in §1.401(a)(4)3(b)(3), (4), and (5). (Of course, standardized plans and nonstandardized safe harbor plans continue to be required to satisfy design-based safe harbors described in the regulations under §401(a)(4).) Thus, for example, an M&P plan, other than a uniform points defined contribution plan, may provide for disparity in the rates of employer contributions allocated to participants' accounts provided the plan satisfies §401(l) in form. Exceptions to the uniformity requirements are provided for Davis-Bacon plans, plans that would fail to satisfy the requirement only because of the plans' top-heavy provisions, and plans that have continued to apply certain limitations under the Code that were repealed by GUST.

I read that to mean that non-standardized prototypes are required to use 401(a)(4) safe harbor allocation methods. I also noticed that the prototypes we use do not have general test language in the base document. Most of our plans use prototype documents.

My comment about about the 3% SH contribution possibly being considered a part of the CODA was not intended to imply that they are deferrals or that they should be treated as such. With mandatory disaggregation of 1) CODA, 2) match and 3) everything else, I'm just wondering if the IRS thinking is that the 3% SH belongs in the CODA portion for testing, not in the portion for everything else. It may be a little far fetched, but it would eliminate the problem with some standardized prototypes having to be general tested. If the IRS didn't miss the boat on this one, they had to have a reason for thinking it isn't a problem.

[AndyH]

OK, my turn.

No, you don't throw a SHNEC in with another employer contribution for general testing. Your understanding is incorrect. You must break out the SHNEC for the permitted disparity calculation, as Blinky spouted.

I think you are correct about DC M&P plans, there technically are no "cross tested" or tiered prototypes. I am not sure about non-safe harbor DB prototypes.

I also have no idea what you are saying about a CODA.

[Blinky the 3-eyed Fish]

My mistake on the nonstandardized prototypes. We only deal with these on takeover situations and I realized that while cross-testing on some of the plans was set up using these documents, they were submitted for individual letters.

[Kevin C]

We deal mostly with prototype documents, both standardized and non-standardized.

The standardized prototype we use allows requiring 500 hours OR employed on the last day of the year to receive the PS contribution, even if a 3% SHNEC is used. So now we have an IRS approved standardized prototype that may not have a 401(a)(4) safe harbor allocation method. But, having a 401(a)(4) safe harbor allocation method is one of the requirements to be a standardized prototype. I may be overreacting, but I think this is a huge problem, unless the IRS has rationalized another way to look at this issue.