DB Offset Plans and the IRS

[Blinky the 3-eyed Fish]

After speaking with an IRS auditor, it appears they have on their radar to look at specific DB offset plans where HCE's have a higher benefit formula and NHCE's a lower benefit formula. It seems as if more than 60% of the nonexcludables have less than the 0.5% benefit AFTER the offset, then they are ruling that it doesn't meet the requirements of 401(a)(26). I am being somewhat general here because I got my first question from the IRS on this today and haven't gleaned some of the specifics.

Has anyone had a similar experience or heard anything more specific from the IRS?

[Guest dsyrett]

This sounds like an extension of the IRS's general position that a >= 0.5% accrual is required to be counted under 401a26.

[Blinky the 3-eyed Fish]

If so, it's a wide and misguided extension. 1.401(a)(26)-5(a)(2) clearly deems those to accrue a meaningful benefit if they would have accrued a meaningful benefit if the offset were disregarded.

[AndyH]

Blinky, have you learned anything further about this?

It would imperil many plan designs.

[SoCalActuary]

My understanding of the safe harbor floor offset is that it is used when you have uniform plan benefits. Your db plan does not. Therefore, the IRS is not permitting the safe harbor floor offset exemption.

[AndyH]

The requirement is that "The employees who benefit under the formula being tested also benefit under the other plan on a reasonable and unform basis".

So the uniformity requirement is on the DC plan, not the DB plan. There is a long discussion of this on this board somewhere.

[SoCalActuary]

Thanks for the clarification. Since this is a design I don't use, I will do more research.

Perhaps the past threads will give Blinky a good argument to use with the IRS agent. Maybe the agent is looking for uniform DB provisions here.

As a personal aside (not on point), my past problem with offset plans has been the funding volatility when the DC plan performance dramatically changes the DB offset. This is especially a problem when the DC plan has participant direction. It encourages smart employees to take the maximum gamble in their DC investment allocation, since they can't lose.

[Blinky the 3-eyed Fish]

Andy, I will update you if I learn more. We just submitted our response to the IRS agent.

SoCal, it is not a safe harbor floor-offset design. It is aggregated with the DC plan and general tested.

[AndyH]

Blinky, what are your thoughts on the interpretation of the "reasonable and uniform" rule.

Example, what if Class A received a 9% DB accrual and everybody else got .5%, both offset by the DC plan.

Does the DC contribution have to be the same percent?

Does the offset have to work the same, i.e. can you have Class A not be offset, i.e. they get both, but Class B gets offset?

This of course all presupposes the general test passes.

I wonder why the DC contribution might need to be "uniform" if instead you could simply alter the DB formula and achieve the same result.

I'm not sure if I'm being clear but I'm wondering if the offset plan needs to provide the same percent of pay to everyone. That makes no sense to me but I suppose you could argue that the "uniformity" rule requires that.

Comments?

[Blinky the 3-eyed Fish]

Regarding that rule, I have had to analyze it when the DC plan has rate groups and the owners get a higher percentage in it. So that means for the DB plan, the owners have a LARGER offset. I don't believe this to violate that rule as I feel one can certainly argue that the DC benefits are reasonable and uniform on a benefits basis and, again, the offset is greater. I have determination letters on plans with this scenario, but the (a)(26) inspection is questionable. It's under the umbrella of the coverage ruling and they don't seem to ask the right questions to make a proper ruling.

I am going to submit a plan where the owner's benefit is not offset at all, while the staff is. I have no plans like this, so we will see. In light of the IRS' quest, I doubt it will get through, but the net effect is minimal if it doesn't.

[Guest penman]

Blinky, have you received any more information on the ".5% accrual AFTER offset" issue you refered to in the initial post of this thread? Thanks.

[GBurns]

While it is easy to determine what is "uniform" it is always open to interpretation and disagreement as to what "reasonable" "meaningful" and "significant" mean. As a result we will have sporadic attacks under code sections containing those words.

So while it might be considered by some to be wide and misguided, "1.401(a)(26)-5(a)(2) clearly deems" that there is room for disagreement with interpretations. Let's see what the specifics will turn out to be of this and other cases, how wide a net is being cast and how aggressive they will be.

I personally expect a major thrust in this and more areas. I do not think that are many units within the IRS that will not aggressively try to find some way to jump on the "illegal and abusive tax shelter" bandwagon. Remember, it is not our interpretations and definitions that matter. Nor is it our chance for a photo op etc.

[Blinky the 3-eyed Fish]

George, I hope you aren't classfifying these arrangements as "illegal and abusive" tax shelters. The net effect of giving 0.5% in the DB or not is minimal as you can reduce the DC benefit by a nearly equivalent amount. The only result of mandating that a net 0.5% be given is the adminstrative problems with more people being in the DB (and eventually being paid out of the DB) with small accrued benefits.

Penman, the only update I can give is that it does appear to be a wide net the IRS is casting in looking at these plans. We have stopped designing them for the time-being until further information is learned. I can tell you that ASPPA is aware of the IRS' actions and they are interested in cases. I specifically emailed the facts of my case to Brian Graff after I went to a talk of his and after he asked if anyone had such cases. It was his opinion that the IRS cannot abruptly change policies without issuing guidance. That is certainly the case here as we and many others have multiple determination letters approving such designs.

[Guest penman]

Thanks Blinky.

[mbozek]

Abusive tax shelters have to be desiganated as listed transactions by the IRS in accordance with the applicable regulations, e.g, certain 419A programs or acceleration of deductions to a qualfied plan. The problem with using terms such as significant, reasonable and meaningful in irs regs is that they cannot be uniformly enforced and are applied in an arbitrary and capricious manner. There is no way for a plan sponsor to know how to comply with regs that use such terms in operating the plan. The IRS will have to quantify its application of such terms in order to enfore the regs.

[GBurns]

Illegal or abusive tax shelters or transactions do not have to be listed. They do not even have to fall under the 'catch all category" used in the relevant Treas Regs etc of "Other Reportable Transactions". See the Regs and Circular 230 (current and Proposed).

[Guest Harry O]

mbozek -

That is the LAST thing the IRS should do. The tax is replete with references to "significant," "material," "reasonable" and "deminimis" which are obviously terms that are susceptible to interpretation. That doesn't make them unenforceable but it does leave the IRS open to second guessing by a court. And there is nothing wrong with that.

IMHO, much of the IRS's recent problems stem from doing just as you suggested -- attempting to write detailed, quantitative rules. This usually results in the IRS being hoisted on their own petard as smart tax lawyers twist the rules to support an unintended result. The 401(a)(4) regs are a perfect example of this shortsightedness. These regs craft elaborately detailed mechanical tests in support of a one line statutory sentence that says qualified plans can't discriminate. The result has been an explosion of plan designs that no rational tax lawyer would have blessed before the 401(a)(4) regs were issued. But now you can get these plans blessed by the government itself because they somehow manage to tick-tack-toe their way through the mechanical minefield of the 401(a)(4) regulations.

The last thing we need is more quantitative regulations.

[mbozek]

I would think the last thing that employers want is is a gotcha regulatory system that operates like the Red Queen's system of Justice ("First the sentence, then the verdict"). From a legal standpoint, the use of subjective terms is unenforceable if it results in arbitrary and capricious application of the law. Very few employers will go to court to fight over these terms because it isnt worth the cost and the court decisions have no consistency. The reason the IRS issued detailed 401(a)(4) regs in 1994 was because of intense criticism that the prior system of reviewing discrimination on a subjective basis (I know it when I see it) created immense discrepancies in plan benefit provisions approved by the IRS and encouraged forum shopping by counsel to find reviewers who would approve plan benefit formulas tilted in favor of HCEs. Prior to the regs, there were instances where different reviewers would take divergent views on the same provision in different plans of the same employer because of their views of vague language in the regs. Employers prefer a regulatory system where they know in advance what is permissible to a system where there is no certainty. The other reason in favor of objective standards is that they can be reviewed easily for abuse and changed in a objective way for all taxpayers, e.g., changes in the cross testing regs.

[Guest Harry O]

I dunno, I guess I thought that was what lawyers do - provide opinions rather than follow a tree diagram in a regulation to a pre-determined conclusion. Objective regulations seem to usually lead to abuse as they get twisted for unintended results. The IRS can't anticipate everything and then are left chasing the bus after the abusive practices come to light. Patching up a leaky reg is not the way to go, especially when the fix usually can't be imposed retroactively since taxpayers can argue they justifiably relied on the old objective regulation. Thus the aggressive taxpayers win again . . .

[Dougsbpc]

I agree with Mbozek.

There must be objective standards applied across the board. This explosion of questionable plan designs was, for the most part, neutralized by the minimum gateway requirements for non-safe harbor plans. Today, there is a very small percentage of all qualified plan participants who are affected by truly abusive designs thought up by smart tax lawyers. This in exchange for, well, an explosion of qualified plans adopted by small employers. This is a good thing. I remember those years prior to 1994 when very few small employers were interested in qualified plans.

Detailed, quantitative rules such as 401(a)(4) including the minimum gateway requirements result in more employers adopting plans and more employees having meaningful retirement savings.

[Blinky the 3-eyed Fish]

It escapes me how something less quantitative leads to clearer results. By definition exactly the opposite is true. Isn’t it just that the promulgations are not quantitative enough? If every last detail was spelled out, then it would be crystal clear exactly what is permissible. The fact that is not the case only means more detail is needed, not less.

And I am sure mbozek will agree, who needs more lawyers to provide opinions?

[mbozek]

B: The Question is who will pay for more legal opinions, not whether more lawyers are needed.

H: Clients dont want to pay for legal opinions on how to maximize benefits for HCEs under the a4 requirements when they can provide discriminatory benefits under a SERP or stock options. They prefer to pay an investment advisor to increase portfolio gains on plan assets. But if you know of a client who wants an opinion on minimizing compliance under the a4 regs I will be glad to provide one.

[AndyH]

Very well-stated comment IMHO,

I guess I thought that was what lawyers do - provide opinions rather than follow a tree diagram in a regulation to a pre-determined conclusion.

[SoCalActuary]

mbozek - SERP & stock options work fine for some companies.

You are, of course, aware that many of our clients don't have these choices.

A large class of employers don't run their business for the eventual value of their stock, since they aren't even allowed enough retained earnings to build stock value.

For those employers, these issues should not be dismissed out of hand.

[Guest penman]

Regardless of what side of the fence we fall on, last fall at local ASPA meeting Paul Schultz said that the days of the of the bright line regs like 401(a)(4) are over. I am paraphrasing but I am not skewing the message. He said that going forward the IRS doesn't want to write regulations that give a roadmap of exactly how to get to the edge of the envelope. Instead, they want the practitioner to use more judgement on what they feel is right. I believe he said "what a good person would do".

[mbozek]

This is ironic since the large consulting companies lobbied the IRS and treasury to create the bright line tests so as to avoid complex, expensive disputes over the plan formulas under the revised non discrimination requirements of the 86 act. The argument was that plan sponsors should not have to continuiously defend the plan formula to different IRS agents who changed their intrepretations of the rules and should be able to rely on a favorable determination letter which would reduce the amount of resources spent by IRS agents on qualfied plans and promote efficient administration. The preamble to the final a4 regs states that the regs allowed the IRS to reduce the frequency of testing qual plans and descriptions of the quality of data that may be used to substantiate compliance with the a4 regs.

Dispite what the IRS says, complaince will not depend on what a good plan advisor would do in applying the rules but what a good IRS agent would do in applying vague and inarticulate regs that use terms such reasonable, significant, etc. If you read between the lines lines the IRS is saying that it does not have the resources to write detailed regs because of budget cuts. As an example, the recent proposed regs on 403(b) annuites require a written plan document to state all of the material terms of the plan but does not describe what constitutes material terms.

[WDIK]

It is very unlikely, in my opinion, that any amount of detailed regulation will be enough to deter those individuals that want to push beyond the "edge of the envelope." On the other hand, failing to give clear guidance in the hope of relying on standard of "what a good person would do" seems doomed to failure.

[GBurns]

I have a very cynical view. I have been watching the tax shelter related arena for quite a number of years, probably starting in 1989 after seeing the surge in activity caused by '86.

I watched how the IRS approached split dollar, reverse split dollar, springing cash value, 419 plan designs and many financial services industry concepts etc before BOSS etc came to the fore. This caused me to also be conversant with the preparer penalties etc and what is now called the tax shelter regulations.

It seems that someone decided that the IRS needs more power after realizing that many taxpayers would test the IRS in court and probably win many cases. After this became a reality the threat emerged that more would challenge the IRS and the IRS might not be able to handle the volume.

So IMHO the decison was made to "load the dice" for the future. How? No more "bright line" no more definitions, no more clarification. That way everything was subject to question and interpretation, that way many taxpayers might be willing to compromise rather than to head for court.

To further "load the dice" there would be fewer Treas Regs and fewer Rev Rulings. Notice the number of Notices that have been issued . In some cases a short non commital Rev Ruling is issued alongside a lengthy Notice. Why a Notice? A Notice has next to no standing.

IMHO the decision to use Notices was so as not to be held to anything hard and fast. It was necessary because the number of courts and cases that have been ignoring or setting aside Rev Rulings and GCMs etc has been increasing, and courts have been moving towards accepting the use of PLRs not as precedent but as explanations of the IRS position on an issue. The result is that the courts have been leaning towards preventing the IRS from espousing a position (PLR or otherwise) then striding to the podium to espouse a different position. Add to this the Supreme Court's decision in MEAD or the many cases such as Grande v Allison regarding Proposed Regulations and agency rulings.

In other words, much that the IRS previously relied on has either come into question or failed them. So they had need for a new approach using attacks based on different support rationale, namely vagueness and subjective opinion.

[could be me maybe not]

Comments such as those by Paul Schultz paraphrased by penman should be dismissed out of hand because the majority of errors or unethical digressions are likely to be done by those not in attendance at such sermons.

[Guest penman]

The point is the IRS's mindset.

[could be me maybe not]

And it is a good and important point. My comments were intended to opine on the absurdity of it.