Discrimination Testing

[ERISAnut]

I am sorry to see this topic to take on an air other than 'neutrality' (or at least it seems to me) from numerous parties involved.

That should not be the point of posting. but things happen like that in the heat of the argument. I admit to losing it this Sunday when an indiviudal knocked at the door and (in my opinion) tried to twist and turn things that were in the Bible and kept changing words that weren't there, would try and change things from what I was pointing out, etc - but that is another story.

anyway

my original comment regarding cross-testing was simply meant to imply if you got to the point of cross testing you don't have a choice of picking and choosing who gets - the minimum allocation gateway would go to all ees who received top heavy.

however as pointed out, there is more than one gateway (unfortunately the IRS used the term 'gateway' in more than one place) - broadly available is also a possibility. and broadly available does not require passing the avg ben % test. though now that you pass broadly available you go on to cross testing and fail avg ben % test and at that point you would have to give at least some of the NHCEs an increase. and that could only be done, if I understand things correctly, through a corrective amendment.

This is an interesting plan. Assume plan fails avg ben % test as indicated.

Plan could pass on allocation basis by bringing enough 3% to 8% to pass.

plan could pass on accrual basis by providing 5% to all top heavy (assuming that is sufficient to pass avg ben% as well, or of course possible the ratio % for each HCE.

wonder which would cost more, or which one the employer would want.

and I am sure there are other possibilities.

Tom Poje, there was a simple question posted 5 multiple choice answers. Just provide your answer to clear it up. I agree with what you are saying; but just admit that the answer is "d"; and not "a". Give us that much.

[Mike Preston]

Correction, I never applied to use of the gateway to the average benefits test directly. I simply applied to use of the gateway to cross-testing, regardless of whether the coverage ratio test or the average benefits test is being used. Again, part of being an expert is knowing what train of thought you are on; and not running to the regs to bring in literature where the facts do not line up.

It was you that brought in a reg where the IRS explained the gateway was exempted because the benefits were primarly DB in nature where the average benefits test was being used to test DB and DC together. You used that same reg as your argument that by the mere fact that the average benefits test was being used ( not because a DB plan was involved) negated a need for the gateway into cross-testing. You lack of understanding on such a simple topic in New Comparability has led to a 5 page discussion that will likely prevent those looking to learn from understanding this concept.

Look back at my original post. While I attempt to break the question down into components and address each component to show the thought processes that pension experts apply on a daily basis, you bring in your own set of regs that apply to situations that aren't being address; just be you are confused.

I admit that I don't understand most of what you have written, above. It is so unclear that understanding what you have written is impossible.

Nonetheless, we soldier on.

I don't know what you are saying above, but you have stated, over and over, that one is precluded from using crosstesting in the average benefits test if the gateway requirements aren't satisfied.

I have stated just the opposite. That is, when performing the average benefits test, you can use EBAR's (that means crosstesting). There are no qualifications on my statement. One can use EBAR's. There is no requirement to meet any gateway threshold.

I am going to provide you with one more citation that you probably won't bother to read. This is from the 2006 ERISA Outline, written by Sal Tripodi. Does he satisfy your definition of an expert? In any event, he writes:

"Gateway requirements under section 401(a)(4) not applicable here. Under Treas. Reg. Section 1.401(a)(4)-8(b)(1)(i), a defined contribution plan is not eligible to be cross-tested for IRS Section 401(a)(4) purposes unless it satisfies a "gateway test" or one of the regulatory exceptions to the gateway test. This condition must be satisfied in order to perform nondiscrimination testing on a "benefits" basis, which is known as cross-testing. This condition does not apply if the sole purpose for converting allocation rates into benefits is to run the benefit percentages for the ABR test. In other words, if the defined contribution plan is still being tested on the basis of contributions (i.e., looking at allocation rates) for Section 401(a)(4) purposes, but the allocations are expressed as benefits to run the ABR test to support the coverage test under IRC Section 410(b) for that plan (or for any other plan maintained by the employer), the gateway condition under Section 1.401(a)(4)-8(b)(b)(1) does not have to be satisfied. The support for this interpretation is that the Treas Reg. Section 1.410(b)-5(d)(5)(1), which explains the mannser in which benefit percentages are calculated for the ABR test, does not cross-reference the gateway requirement under Treas. Reg. Section 1.401(a)(4)-8(b)(1)."

Are we learning, yet?

[Mike Preston]

It was you that brought in a reg where the IRS explained the gateway was exempted because the benefits were primarly DB in nature where the average benefits test was being used to test DB and DC together. You used that same reg as your argument that by the mere fact that the average benefits test was being used ( not because a DB plan was involved) negated a need for the gateway into cross-testing. You lack of understanding on such a simple topic in New Comparability has led to a 5 page discussion that will likely prevent those looking to learn from understanding this concept.

It is becoming a bit clearer now. You are absolutely correct in your assumptions and incorrect on your conclusion.

You believe that the citation in question, merely because it uses an example of a DB and a DC plan being tested together, does not apply to the case at hand; that is, a defined contribution plan that stands on its own.

You are absolutely correct that I believe the opposite.

So your protests that I've brought in an irrelevent regulation that does not apply to the case at hand is countered simply by saying that you are incorrect in your understanding of the applicability of the regulation citation in question. It most certainly does apply.

[ERISAnut]

The SOLE PURPOSE in this instance would be where this type of maneuver is requirement in order to peform the average benefits test. It is impossible to perform an average benefits test where the employer sponsors a DB and a DC plan without crosstest. Hence, the sole purpose would be in order for the average benefits test to be possible.

I know see your confusion. Glad I could help.

[austin3515]

Preston wins. Case closed. Can't argue with Sal. Could he have made it any more clear?

[ERISAnut]

Let me restate that is another manner.

If is possible to perform the average benefits test without cross-testing, the the use of cross-testing in the average benefits test would not be for the sole purpose of performing that test.

This is why I don't run to the regs to answer questions. They are written by attorneys (in many instances) and often require pension experts (such as myself) to break them down into layman's terms. Hence, bringing the regs into play often confuses the less experience who often struggle to understand simple topics.

So my suggestion is tone it down with the regs and let's have a real discussion about how to operate a qualified plan in order to maintain its qualified status.

[ERISAnut]

Preston wins. Case closed. Can't argue with Sal. Could he have made it any more clear?

Glad to see we have another confused one. Let me give you an illustration.

An employer sponsors a DC plan and a DB plan. The DC passes 401(a)(4) as a safe harbor plan based on the fact that everyone received the same allocation and 70% of the NHCE's benefited under the DC. The DB plan fails the coverage ratio test; and now the Average Benefits Test will be used.

Since the averge benefits test requires all plans of the employer to be tested, this will require the DB and DC plan to be included in the test. It is impossible to test these two plans under the average benefits test without crosstesting. (This does not mean that it is never possible to test under average benefits without cross-testing) Again, it is IMPOSSIBLE to test this DB and DC plan under the average benefits test without crosstesting (either convert the DB to an allocation or convert the DC to SLA at NRA). The sole purpose for converting DC into a straight life annuity would be so the plan can perform the average benefits test.

This is where experience makes the difference. BOY!!!

[Mike Preston]

You lack of understanding on such a simple topic in New Comparability has led to a 5 page discussion that will likely prevent those looking to learn from understanding this concept.

I think it is precisely the opposite. I pointed this issue out very early in the thread. If you would have understood it at that time, this thread would have been much different (and shorter!), and, most likely, more useful for the original poster.

But I am not going to sit back and watch you give BAD/POOR advice to the readers of BenefitsLink. It has been around a long time and you are a johny-come-lately. It has been around so long because, quite simply, many refuse to allow bad advice to stand unchallenged.

[Guest mjb]

I have no / expertise /opinion/ comments regarding the above discussion but this non quantative person has one question regarding the use of x testing and EBARS. I have seen illustrations of new comp plans that use a formula which provides a 5% minimum gateway to all the NHCEs and then a 21% contribution rate for HCES in their mid 40s and early 50s who have the max comp limit of 220k (44k contribution) as passing Xtesting using EBARS. No other details are provided in the illustration. To my non quant brain this formula violates the basic rule of not discriminating in favor of HCEs.

Q is this type of contribution formula ever permissble under Ave benefit test, EBARs, or some other rule under 401a4? I hope there is a simple answer.

[ERISAnut]

Here we are again. I'm mister bad character and you are savior of the board. I have just given you some valuable insight that you may choose to consider in the near future.

When the regs are written, they are written to technical precision. Amatures such as yourself should not attempt to run there and quote them without being capable of determining the true meaning of what is being said. And then to attack my character for showing you how to articulate otherwise complex matters in layman's terms is weak.

Hope you have your napkin ready.

[Mike Preston]

Preston wins. Case closed. Can't argue with Sal. Could he have made it any more clear?

Glad to see we have another confused one. Let me give you an illustration.

An employer sponsors a DC plan and a DB plan. The DC passes 401(a)(4) as a safe harbor plan based on the fact that everyone received the same allocation and 70% of the NHCE's benefited under the DC. The DB plan fails the coverage ratio test; and now the Average Benefits Test will be used.

Since the averge benefits test requires all plans of the employer to be tested, this will require the DB and DC plan to be included in the test. It is impossible to test these two plans under the average benefits test without crosstesting. (This does not mean that it is never possible to test under average benefits without cross-testing) Again, it is IMPOSSIBLE to test this DB and DC plan under the average benefits test without crosstesting (either convert the DB to an allocation or convert the DC to SLA at NRA). The sole purpose for converting DC into a straight life annuity would be so the plan can perform the average benefits test.

This is where experience makes the difference. BOY!!!

Even when faced with a direct citation, you persist in your ridiculous interpretations. The relevant sentence you are attempting to mangle is:

This condition does not apply if the sole purpose for converting allocation rates into benefits is to run the benefit percentages for the ABR test.

What this means is that if the conversion of allocation rates into benefits is limited to testing under Section 410(b) [for the average benefits test], then the gateway doesn't apply. That is, if you convert for a reason other than 410(b) (which of course means 401(a)(4)), the gateway rules apply.

[Mike Preston]

I have no / expertise /opinion/ comments regarding the above discussion but this non quantative person has one question regarding the use of x testing and EBARS. I have seen illustrations of new comp plans that use a formula which provides a 5% minimum gateway to all the NHCEs and then a 21% contribution rate for HCES in their mid 40s and early 50s who have the max comp limit of 220k (44k contribution) as passing Xtesting using EBARS. No other details are provided in the illustration. To my non quant brain this formula violates the basic rule of not discriminating in favor of HCEs.

Q is this type of contribution formula ever permissble under Ave benefit test, EBARs, or some other rule under 401a4? I hope there is a simple answer.

The answer is YES. The magic number between 5% and 20% is 17 years. That is, if there are "enough" NHCE's at least 17 years younger than each HCE you posit, all tests will be satisfied.

"Enough" is a function of a number of factors, such as:

1) The plan sponsor's Concentration percentage (ratio of NHCE's to HCE's)

2) Whether the average benefits test is satisfied

3) Rounding

Assume 1 is 90% and 2 is YES. And let's ignore 3 for now. Then for each HCE you have identified you would need .2375 * 9 = 2.1375 NHCE's (that means 3 is your rough equivalent) who are at least 17 years younger than the HCE in question.

Therefore, if the plan in question had a 45 year old HCE getting 20% of pay and at least three 28 year old NHCE's getting 5% of pay, it would work.

The above is actually a simplified calculation as I ignored permitted disparity. In fact, the number is smaller than the 17 years I mentioned (think 16 instead).

[austin3515]

ERISA Nut. You're not very nice. I also think you think a little too highly of yourself. I'm not stupid. Anyone who says Sal Tripodi is wrong, and that one shouldn't look to the regulaitons for answers is indeed a NUT.

I think you should read the EOB paragraph that Preston attached and specifically indicate where in that concise clear as day paragraph Sal is mistaken.

[ERISAnut]

This condition does not apply if the sole purpose for converting allocation rates into benefits is to run the benefit percentages for the ABR test.

What this means is that if the conversion of allocation rates into benefits is limited to testing under Section 410(b) [for the average benefits test], then the gateway doesn't apply. That is, if you convert for a reason other than 410(b) (which of course means 401(a)(4)), the gateway rules apply.

I am saying that the term "sole purpose" has meaning. Otherwise, the author would've ommitted it from the statement.

Your are saying that "This condition does not apply if the sole purpose for converting allocation rates into benefits is to run the benefit percentages for the ABR test" has the same meaning as "This condition does not apply if converting allocation rates into benefits in order to run the benefit percentages for the ABR test". I am saying that the term "sole purpose" has meaning. So who among us is providing the weird interpretations?

[Guest mjb]

I am beginning to understand.

Q1 If there are 2 hces of different ages then would you need to have 3 NHCEs more than 17 years younger than each HCE e.g., 42 yo HCE requires 3 ee 25 or less and 50 yo HCE requires 3 more nhces age 33 or less or can you double count some or all nhces for both HCEs?

Q 2 if new comp plan is only plan, does the 5%/21% formula pass ave. bene test?

Q3 if the total of all ee is 10 or less and there are 2 HCEs would this formula ever pass xtesting?

[ERISAnut]

ERISA Nut. You're not very nice. I also think you think a little too highly of yourself. I'm not stupid. Anyone who says Sal Tripodi is wrong, and that one shouldn't look to the regulaitons for answers is indeed a NUT.

I think you should read the EOB paragraph that Preston attached and specifically indicate where in that concise clear as day paragraph Sal is mistaken.

Austin3515, I am not saying that you are stupid. You simply remind me of myself when I was inexperienced and stupid. There is nothing in what I wroted that came close to implying Sal Tripodi is wrong. We are merely discussing the meaning of a sentence. Let's leave the rhetoric aside from now a discuss the real issue.

Some of us understand when the gateway rules apply and some of us don't. The fact this I understand the rules and you don't does not mean that I am thinking highly of myself; but instead means that I do not think highly of you. I am trying to teach you something that you have no interest in learning. All you want to do is attack with your weak interpretations of someone else's literature. You should grow beyond that.

[Bird]

But I am not going to sit back and watch you give BAD/POOR advice to the readers of BenefitsLink. It has been around a long time and you are a johny-come-lately. It has been around so long because, quite simply, many refuse to allow bad advice to stand unchallenged.

Major props to MP for persisting on this. It's one thing to throw out opinions - which may differ - but to insist that regs don't mean what they say (or is the nut saying that it's cheating to cite regs?) is seriously misguided. Preston wins this hands down on reasoning, grammar, spelling, and politeness.

[ERISAnut]

But I am not going to sit back and watch you give BAD/POOR advice to the readers of BenefitsLink. It has been around a long time and you are a johny-come-lately. It has been around so long because, quite simply, many refuse to allow bad advice to stand unchallenged.

Major props to MP for persisting on this. It's one thing to throw out opinions - which may differ - but to insist that regs don't mean what they say (or is the nut saying that it's cheating to cite regs?) is seriously misguided. Preston wins this hands down on reasoning, grammar, spelling, and politeness.

You know, as I watched this board over time, I am often wondered where the experts are. Now I see. The experts would not bother to waist their time with a bunch of flunkies who have no real world experience with retirement plans. So before we go any further to insult each other, why don't you ask an expert that you may know personally to explain to you why I am right on this issue.

[austin3515]

ERISA Nut. You're not very nice. I also think you think a little too highly of yourself. I'm not stupid. Anyone who says Sal Tripodi is wrong, and that one shouldn't look to the regulaitons for answers is indeed a NUT.

I think you should read the EOB paragraph that Preston attached and specifically indicate where in that concise clear as day paragraph Sal is mistaken.

Austin3515, I am not saying that you are stupid. You simply remind me of myself when I was inexperienced and stupid. There is nothing in what I wroted that came close to implying Sal Tripodi is wrong. We are merely discussing the meaning of a sentence. Let's leave the rhetoric aside from now a discuss the real issue.

Some of us understand when the gateway rules apply and some of us don't. The fact this I understand the rules and you don't does not mean that I am thinking highly of myself; but instead means that I do not think highly of you. I am trying to teach you something that you have no interest in learning. All you want to do is attack with your weak interpretations of someone else's literature. You should grow beyond that.

But, IF, IF, IF (let me emphasize again, IF) I mistakenly assumed you called me stupid in a prior post, surely you don't expect to me believe your last post doesn't?

No response needed, nor desired...

[wsp]

Wow....This has gotten way out of hand. Sorry about that. And sorry about delays in my response as I've been out of office with daycare issues. Family comes first!

Someone about 6 pages back (I think it was Mike) asked about numbers as they doubted the plan truly failed. I'm hoping I'm wrong and it actually passes but here goes.

36 total employees

12 excluded min age & service

2 excluded <500 hours

22 non excludable

7 non excludable are HCE

16 benefit

7 of those are HCE

Ratio Percentage 60

Going to Average Benefits I get 100% HCE Average of 6.73 and NHCE Average of 4.12.

2 of 7 HCE's receive comp at 190k and happen to be about 40 years old. Most of staff is about same age

Top Heavy goes to all non-key participants who are employed at end of year even if not getting year of service. participation for deferral was 3 months. So, you do have participants receiving 3% at decent level of compensation. Further, three members of TH group were born in late 70's I have 7 people in TH group and 10 in the Not Eligible group (1 due to december hire and 9 due to term; six of which termed >500 hours)

So, I think the plan still fails after 3% TH is given. Plan document says that TH contribution has to be "at least 3%" hence the original question. Why not increase that group up to plan's passing the ABPT test.

I realize this has evolved into whether gateway must be given and for my specific case it might be moot as I have to give 5% to pass anyways (I had thought it was 4 but it's not).

Now, the argument against increasing TH contribution is 1) goes against plan document which isn't true as plan document doesn't limit me. To me "at least" means I can go higher. And it's a top heavy minimum not a top heavy maximum (isn' t it???) 2) It would create a separate rate class. But isn't that what TH already does? And aren't I already including them once I go to the ABPT anyways?

If I can't do that, what's the criteria that I use to bring people back. Change the entrance requirement? How do you make a change like that applicable to only 1 year? Seems to me that making such a change only brings that TH group into the mix anyways so what's the difference in what I'm saying and you are saying?

Sorry for such base questions...10 years of my experience was with medium sized plans that the extent of the discrimination testing was adp/acp testing and the occasional refund calculation. Coverage and other testing was never an issue and plans never went top heavy.

[Mike Preston]

This condition does not apply if the sole purpose for converting allocation rates into benefits is to run the benefit percentages for the ABR test.

What this means is that if the conversion of allocation rates into benefits is limited to testing under Section 410(b) [for the average benefits test], then the gateway doesn't apply. That is, if you convert for a reason other than 410(b) (which of course means 401(a)(4)), the gateway rules apply.

I am saying that the term "sole purpose" has meaning. Otherwise, the author would've ommitted it from the statement.

Your are saying that "This condition does not apply if the sole purpose for converting allocation rates into benefits is to run the benefit percentages for the ABR test" has the same meaning as "This condition does not apply if converting allocation rates into benefits in order to run the benefit percentages for the ABR test". I am saying that the term "sole purpose" has meaning. So who among us is providing the weird interpretations?

You, in my opinion.

If we accept your opinion that "sole purpose" has the, uh, purpose of rendering the average benefits test subject to the gateway, we find that the following sentence:

"In other words, if the defined contribution plan is still being tested on the basis of contributions (i.e., looking at allocation rates) for Section 401(a)(4) purposes, but the allocations are expressed as benefits to run the ABR test to support the coverage test under IRC Section 410(b) for that plan (or for any other plan maintained by the employer), the gateway condition under Section 1.401(a)(4)-8(b)(b)(1) does not have to be satisfied."

has no meaning.

I'm not willing to believe that the entire sentence I repeated directly above has no meaning.

I'm beginning to see a pattern. I believe that while you were in your government position, it was made clear to you that you should ignore the literal wording of the Code/Regulations and merely carry out department policy. You mistakenly believe that others should be similarly instructed and that your years of experience provide you with a looking glass through which you instinctively know what the regulators meant to say.

Sorry, but that just doesn't cut it.

You go from one extreme to the other. While you don't want to parse the regulations, you now want to parse an expert's description of those regulations as if he were writing regulations. As I've already stated, and as you've already disagreed, the meaning of "sole purpose" in that description is to draw a distinction between applying EBAR's for the sake of 410(b) or for the sake of 410(b) and 401(a)(4). As an educator, Sal is about the best there is. But your inability to understand his description proves that even the best educators can't reach everybody.

I bid you peace.

But I will not allow you to post an unanswered message that implies the gateway is required when the 401(a)(4) test, if required, is performed on the basis of contributions. The sole reason that the gateway is required is if the 401(a)(4) test is done using EBAR's when a DC plan is involved, acknowledging that even in that circumstance the gateway is not required if one meets the exceptions. To repeat using other words, the gateway is never invoked predicated on the basis of the average benefits test using cross testing.

[ERISAnut]

You're benefiting 100% of the Nonexcludable HCE's. In order to pass the compensation ratio test you only need to benefit 70% of the nonexcludable NHCE's (11 in this case).

You are already benefiting 9 of the 11 needed to pass the compensation ratio test and not worry about crosstesting or average benefits testing. Likely those additional two employees would normally receive the 3% TH minimum; but this will not be a TH miniumum once you amend to bring them into the allocation formula for the plan. Another 5%, for a base allocation of 8% for the two brought into the plan would get you were you need to be. Others who are not brought into the plan will be left at the TH minimum of 3%. THIS IS IN RESPONSE TO WSP'S ADDITIONAL QUESTION.

[Mike Preston]

But, IF, IF, IF (let me emphasize again, IF) I mistakenly assumed you called me stupid in a prior post, surely you don't expect to me believe your last post doesn't?

Look at it this way. Who amongst us would want to be held in high esteem by the Nut? His ideas are truly bizarre. He believes that regulations and Code are indecipherable. Do you? Of course not. He believes that regulations and Code should be ignored in favor of some ethereal principles that only he knows. Do you? Of course not. He sees language that is crystal clear and interprets it in a bizarre manner. Do you? Of course not.

I have been called inexperienced and incapable by him.

Austin, welcome to the club. I am happy to line up in your camp.

[wsp]

You're benefiting 100% of the Nonexcludable HCE's. In order to pass the compensation ratio test you only need to benefit 70% of the nonexcludable NHCE's (11 in this case).

You are already benefiting 9 of the 11 needed to pass the compensation ratio test and not worry about crosstesting or average benefits testing. Likely those additional two employees would normally receive the 3% TH minimum; but this will not be a TH miniumum once you amend to bring them into the allocation formula for the plan. Another 5%, for a base allocation of 8% for the two brought into the plan would get you were you need to be. Others who are not brought into the plan will be left at the TH minimum of 3%.

Can I discriminate in that regard? Or do I use DOH and bring in earliest hires? And, if I amend to bring in someone based on hire...how does that not apply next year for people who are already employed? I guess..how do you word that to make it applicable to only this year? Without it being a loss of a benefit for others who are employed? Seems like it's changing the entry date to me...

[ERISAnut]

You're benefiting 100% of the Nonexcludable HCE's. In order to pass the compensation ratio test you only need to benefit 70% of the nonexcludable NHCE's (11 in this case).

You are already benefiting 9 of the 11 needed to pass the compensation ratio test and not worry about crosstesting or average benefits testing. Likely those additional two employees would normally receive the 3% TH minimum; but this will not be a TH miniumum once you amend to bring them into the allocation formula for the plan. Another 5%, for a base allocation of 8% for the two brought into the plan would get you were you need to be. Others who are not brought into the plan will be left at the TH minimum of 3%.

Can I discriminate in that regard? Or do I use DOH and bring in earliest hires? And, if I amend to bring in someone based on hire...how does that not apply next year for people who are already employed? I guess..how do you word that to make it applicable to only this year? Without it being a loss of a benefit for others who are employed? Seems like it's changing the entry date to me...

Even though they were employed on the last day, they each worked less than 1000 hours during the year. I doubt they would all have the same number of hours. So your amendment would simply bring in the employees starting with the one having the highest nubmer of hours during the year. Try that one for starters.