Revoking the Irrevocable

[Guest erisafried]

Quick question for those of you who design and maintain "sports car" DB plans (you know--the ones doctors, dentists, and lawyers like):

Broadly, this issue arises under Reg. Section 1.401(k)-1(a)(3)(iv) [(v) in the proposed 401(k) regs] which allows for a one-time irrevocable deferral (or contribution, where DB plans are involved) election for folks who are newly-hired or newly-eligible for a plan. If you touch all the bases, this election is not regarded as a CODA.

For, say, a brand new DB plan that a partnership is installing (which will be cross-tested with an existing DC plan of some sort), the practice seems to be to allow the partners to make a one-time election to participate in the DB plan. If you elect to participate, you take home less money and the amount you elect is transferred to the DB plan instead.

Leaving aside the issue of whether the one-time election option is even available in this scenario as a result of participation in an existing plan (a strict reading of the regs and related guidance suggests that it may not be), suppose this partnership subsequently fell under the spell of a consultant who promised even greater riches through more agressive cross-testing, etc. In order to be able to take advantage of said changes, the participants would obviously need to be able to revise their "irrevocable" election. Can they do it?

Despite what seems to be a pretty clear prohibition on revokable irrevocable elections, are employers actually allowing them anyway and just playing audit roulette? Has anyone gotten the IRS to approve changes to elections when the terms of the plan at issue change dramatically?

I suspect that this sort of thing is, in fact, going on, and I am trying to figure out what rationale is used to support it. Any thoughts?

[Blinky the 3-eyed Fish]

Are you worried that the partners need this irrevocable election in order to receive monies in the DC and/or DB plans or otherwise it will be deemed a CODA? If so, don't be. As long as benefit decisions are made on the employer level (yes, even though the partners may be controlling their personal plan savings behind the scenes), then this will not currently be raised as an issue by the IRS.

[Guest erisafried]

That's certainly part of the concern.

So, just to see if I've got this straight...

DB plan (cash balance to be specific) provides for tiered contributions by partners. When you sign up, you do so with the understanding that (1) you're in for as long as you're a partner; and (2) the plan sponsor will automatically deduct the amount specified in the plan from your earned income based on your age and length of service.

Plan sponsor (doing the bidding of the partners) decides to substantially increase the tiered contributions and, because some partners are worried about cash flow issues, wants to give all partners a second bite at the apple (i.e., to either stay in the plan or opt out).

If the plan sponsor was to just "unilaterally" ("Pay no attention to that man behind the curtain!") increase the contributions but the participant group stays the same, I wouldn't get nervous. What gives me pause is the fact that the plan sponsor wants to give partners an opportunity to essentially make a new election--which looks more like a CODA type of deal versus something "imposed" by the plan sponsor (i.e., not an election).

Because of the sensitivities here, no way, no how is this going to be implemented without a prior IRS ruling. I am trying to figure out whether I am going to look like a dope by going in to seek one if the regs, etc. are squarely against me. Or, as you suggest, maybe it's standard practice to just gloss over this point (Mr. Consultant certainly did) because the IRS is not really hunting this particular critter at the moment.

[Blinky the 3-eyed Fish]

Let's forget about the DB plan for a moment, because the same concept applies to a cross-tested plan. There are many designs where individual partners are in their own rate groups, in fact there are many designs where each individual is in his own rate group. Each year for the partners their contribution is generally treated as a directed expense, thereby reducing their earned income. Who is really controlling how much is contributed for the partner? Of course, it's the partner. In other words, they are deferring income rather than taking immediate cash, a CODA. Does the IRS know this is occurring? Of course. However, to play the little game it is important that any changes to the document or decisions on what amounts go to each allocation group be made on an employer level. That is what keeps it from going over the final hurdle of being a CODA.

A DB plan in this situation is the similar. To fluctuate the profit sharing contributions for a partner no amendment is needed to the plan, so it's a bit cleaner to do this year to year. A DB plan would require an amendment for each partners' level of benefits to change things. Do multiple amendments for multiple partners and you have yourself a nice paper trail of changes. It just looks worse, even though it's really the same idea as the PS plan. That paper trail may pique the interest of an IRS reviewer or it may not.

So that brings the DB plan into a gray area. How many changes are allowed? Will the IRS care if every partners' benefit is changed every year? Will they care if it happens every few years? I don't know the answer.

[Guest erisafried]

[still trying to get my arms around this...]

Would you allow a partner who previously elected not to participate in the plan to opt into the plan after the increased contributions are available (or vice-versa)? If so, I suppose the argument you'd be making implicitly (and maybe this is the point you're making--it's Friday afternoon, and this is making my head hurt) is that all partners are "participants" but some don't have any accruals and some do (based on the schedule appended to the plan). Following the amendment, there'd be no new opt-in or opt-out. Instead, the plan sponsor would essentially adjust the contributions for $0 participants (and others) to some higher level (at their behest). It'd be a non-election election.

The plan at issue has been reviewed pretty thoroughly by the IRS, and we got no heat for scheduling individual benefit accruals for partners. So, I feel like we're in reasonably good shape there. Unfortunately, the plan docs don't provide that all partners are participants--just the ones who opted in when the plan was adopted, as well as any new ones that came aboard since then. I could reform them now to make them consistent with the foregoing argument, but my guess is that would look kinda suspicious to Jim Holland and his actuarial stormtroopers ("You know that old plan document that excluded the non-electing partners? Well, we take that back. Those guys were ACTUALLY participants who didn't receive accruals. Yeah...that's the ticket.").

[AndyH]

If I may interject.

Erisafried, are you of the (false) impression that the arrangement you describe is somehow an accepted qualified plan design that might be approved by the IRS? You are describing a scheme. "Opting in" and "opting out" are not acceptable plan terms for a DB plan.

What you describe is intended to be shielded from IRS scrutiny, not submitted for approval. Your apparent skepticism is well founded. I know these arrangements are not uncommon to some law firms, but these terms that you describe are not exactly in the SPD.

Setting up one of these things is a business decision that is not without risk.

[Belgarath]

Since I take a rather strict constructionist view of such questions, I find all this somewhat mystifying.

We don't do "designer" DB plans, and the documents that I see are generally prototypes, but the language is unambiguous. You can IRREVOCABLY waive participation. If that's what the document says, then you can't revoke it just because you think you might get away with it. Or at least you shouldn't...

If the document says that it is REVOCABLE, then you can apply for a determination letter to confirm if this aspect is acceptable. Or if there's already a letter allowing such a scheme, then you should be ok.

It seems pretty black and white to me, which makes me assume I'm missing the point.

[Guest erisafried]

You're not missing the point at all. Now, the consultant and the plan sponsor...different story.

I mainly wanted to try to get some sense about how flexible people were in dealing with this issue. I, too, had the sense that these sorts of practices go on out in the field but didn't want to be in a position of advising my client to follow the rest of the lemmings over the cliff.

I think the regulations are pretty clear about how this is supposed to work (or not work), but I was also anticipating objections from the consultant (don't be a stick in the mud--EVERYBODY does it like this, etc.). The plan sponsor here wants to believe the consultant and is probably going to be looking for reasons to discount the conservative (and correct, I think) position that I will be taking. Not that I have it in for consultants or anything...just don't like getting knee-capped in client meetings by agressive ones.

[Blinky the 3-eyed Fish]

I have to admit I am confused as to the exact intent of all of this. It is extreme partner flexibility in the DB plan? When would the decisions be made to "opt in" or "opt out"? I sounds like it's a design where benefit accruals are frozen each year and then increased to the extent each partner wants them to be before the 3/15 after the year-end. Then the benefits are re-frozen and this process continues year to year. If so or if it's something along those lines, then that is something I would not do, especially without the IRS's blessing in full.

[AndyH]

Chicken.

[AndyH]

ERISAfried, how about sharing the background and qualifications of the "consultant"?

[GBurns]

erisafried,

I think that you are approaching this from the wrong direction.

You are trying to prove your side rather than forcing the scheme "promoter" to prove that their design can fly. Allowing them to say that this is what others are doing is not enough.

They should be able to show conclusively where this design is clearly allowed. They should by now have at least a few clients whose legal advisor/tax consultant approved such a design in writing. These legal advisors/tax consultants should be willing to discuss what they looked at and verify that you are looking at the exact same design.

If no such "evidence" can be produced, then logically that is not what others are doing and others have not seen it as okay and the design might not be compliant.

It is a sales trick to shift the burden of proof to someone else who does not have the details nor has the intricacies of the design. You can never be right and so the sales rep wins by default.

Do not continue to be "suckered" let him prove his position.

[Guest erisafried]

Blinky: It's not really "extreme" flexibility that they seem to be going for. Rather, they mainly want to significantly increase the accruals (which is effectively increasing the contributions made by individual participating partners) AND use that increase as a basis for allowing both current participants as well as partners who previously elected not to participant to change their minds. The theory is that the significant change to plan terms justifies giving everyone a second bite at the apple. Although there could be further changes to the plan in the future, there is no intention to continuously vary the accruals from year to year.

As it goes, my impression is that plan is currently sort of a model-T design (i.e., servicable but not sexy). IRS agent Torquemada gave the thing a pretty thorough going-over shortly after it was adopted and concluded that the design was permissible. Mr. Consultant now wants to come in and strap on a supercharger.

Andy: Let's just say that the consultants come to us from one of the "Final Four" and from another firm that you've probably heard of. The whole deal has the aroma of a "tax product" which always makes me a bit nervous.

GBurns: I hear you, brother. I haven't seen the back-up for the proposal as yet, so I will reserve judgment for the time being. I have found, however, that there is generally an inverse relationship between the strength of a tax position and length of the research memorandum necessary to support it--we'll see if this one is true to form.

As a practical matter, I think I will win this one ultimately, but the fact that the client already thinks it can do what the consultants suggest means that I need to be well armed before going into battle.

[Blinky the 3-eyed Fish]

Well then what's the problem? It would have been idiotic for the first partners that elected to not have accruals in the plan to have signed an irrevocable election, so I assume the document excluded them from the plan or included them but with no accruals. A DB plan design is not completely inflexible so that once a benefit decision is made it cannot be changed ever. Now that you stated more detail I am unsure of your concerns.

[Guest erisafried]

I think I see what you mean. So far as I know, no one signed any irrevocable elections, so really what we are talking about is a plan sponsor decision to prospectively modify the eligibility and accrual provisions of the plan rather than participant "elections" (even though this view is moderately fictional given the fact that the participants--or some of them, anyway--control the plan sponsor).

I was focused on the way the consultants presented the issue and that prompted my concern initially. The issue was presented with more of a partner/participant focus than a plan sponsor focus and that got me thinking about the one-time irrevocable election regulations.

I still think that the "fictional" element involved in the change makes it prudent to seek IRS guidance before moving forward. At bottom, I guess my question was really whether people felt much concern about about the fictional aspects in practice or whether it was just something that tends to get finessed.