409A Guidance/JCEB Conference - Surprises?

[TCWalker]

One that I liked was the comment predicting the end of tandem/pour-back arrangements between 401(k) and NQDC plans. "They're close to being dead if not dead", was one comment I think. (The scenario discussed was a NQDC that pours deferrals into a 401(k) after year-end to max. allowable limits, then a similar dovetail comment was made about arrangements that work in reverse - k to the max, then to NQ.) "Employers are likely to de-couple these arrangement in the future to comply with 409A" was another afterthought.

Seems to me if this is simply a matter of the timing rule, there are other fixes. I guess we'll see if this position is re-thought as guidance matures in '05.

[Kirk Maldonado]

I think that they are wrong about "wrap-around" or "cash back" plans. I believe that it should be capable of designing them so that they should be considered to comply with section 409A.

[Guest Harry O]

Kirk,

Can you elaborate?

[mbozek]

How about elective contributions which are subject to a risk of forfeiture when they are made but are converted to 401k sal reduction at yr end when the excess becomes a vested NQDC?

[GBurns]

For these you will also have to address the issue of what is a "substantial risk of forfeiture", which was not defined.

[mbozek]

See Q-10 for definition of subs. risk of forfeiture.

[GBurns]

I doubt that you will find that useful as a definition. How could you use that to quantify "substantial" which must be quantified to be used?

[E as in ERISA]

Who made the comment about them being dead? Treasury?

[mbozek]

GB: Where does 409A require quantification? SROF is not quantified under IRC 83 which has been in effect since 1969.

[Alf]

Who made the comment about them being dead? Treasury?

Yes, Sweetnam

[Kirk Maldonado]

The speaker stated that there are two problems with these "wrap-around plans:" the timing rules and the amount rules.

I think that they can both be overcome if the wrap-around plan is properly drafted.

My guess is that the person who stated that these plans are dead probably isn't intimately familiar with how they operate and/or doesn't realize that they could be designed to avoid those rules. My hope is that somebody will prepare a thoughtful analysis of this situation and get them to reverse that position.

[Guest Harry O]

Kirk -

I guess I was hoping you would elaborate on how you would overcome these rules . . .

[GBurns]

Kirk

Will you also elaborate on "amount rules"?

[Kirk Maldonado]

GBurns:

Since the speaker didn't elaborate as to what he meant, I don't think that I can definitively say what he had in mind.

My point is that I'm not aware of any insurmountable obstacles contained in the statutory language or the notice that would preclude such an arrangement. It might be that future guidance will clarify why the government doesn't think that they work, but I don't see any obvious reasons why you couldn't make one work.

If somebody can see an inherent and insurmountable conflict, please identify it or them.

[Guest DMK]

I think that one potential issue, at least with respect to 401(k) tandem/spillover plans, from the IRS point of view is that an employee may be able to effectively control/manipulate the amount that is deferred into his nonqualified plan during the year by making changes under the 401(k) plan deferral election. There may be ways around this, but I understand that this is one concern that has been expressed.

[E as in ERISA]

I think that under 409A the deferral election should be as specific as possible, not subject to any future changes on the part of the participant. Any variations should be based on things outside the control of the participant -- such as decisions by the employer about increases in compensation or bonus amounts -- or possibly legal limits.

I had heard that plans that meet the "mirror" requirements under which amounts are transferred from an NQ to a 401(k) after ADP testing were not considered abusive. If there is no further discretion on the part of the participant after the initial election was made, then any "amount" issues should be satisfied. I'm assuming that the primary issue is that the "transfer" from the NQ to the 401(k) is not currently on the statutory list of distributions. Maybe the IRS could add it under "acceleration" rules that it is allowed to write. The timing of the distribution

would have to be established in advance, subject to acceleration only when ADP testing provides an opportunity for transfer.

I think that the IRS' bigger concern would be plans where the NQ deferrals begin only after the 402(g) limit is reached in the 401(k) plan. The participant has control over whether the 402(g) limit is reached. So the NQ deferral could be subject to manipulation. The IRS might think that the answer is for the NQ deferral election to be worded as being a specific dollar amount or straight percentage of compensation minus the 402(g) limit. So they think a specific election works. That may be fine if you think of it on an annual basis. But its more complicated if you have to administer it on a payroll basis. And many plans are designed that way because the employer is going to match up to a specific percentage of total compensation in both the NQ and 401(k) (no 401(a)(17) compensation limit). And then there are catchup issues.

For example, what if during 2005 someone age 60 with $250,000 of annual compensation wants to maximize 401(k) deferrals (including catchups) and to defer an additional 10% of pay into the NQ up to a maximum of about $30,000 (knowing that there are some variable in compensation)...and actually gets an increase in June to $300,000 and gets a bonus of $50,000 in November. And the employer is willing to match the first 6% of deferrals into either plan...but also has set an administrative limit of 5% for HCEs in the plan. There are 26 payrolls a year. My understanding is that one of the issues is how to set up this election on the payroll system (for a company with 50,000 employees and 500 NQ participants). It would be easier to administer if you could tell the payroll system to deposit the deferrals into the 401(k) until the 402(g) limit is reached and then start contributing to the NQ as soon as the limit is reached in the 401(k). But if the participant decides to zero 401(k) elections early in the year, nothing will ever go into the NQ. Then I think that the match and catchup calculations become more complex.... How do you word them to apply correctly on a payroll by payroll basis and make sure the top executives get the maximums?

[mbozek]

The IRS should clarify why a wrap around plan could not comply with the rules for 409A. It would also be possible to make contributions by using funds that are not subject to 409A restrictions.

[Kirk Maldonado]

DMK:

Could you explain how that problem you suggested could ever happen in a wrap-around plan? What you are suggesting is fundamentally inconsistent with the way every such plan I've seen was operated as well and the logical underpinnings of how they are designed.

Are you sure that you are even talking about the same type of arrangement? Is your arrangement consistent with those discussed in the PLRS issued on such plans?

Were you the person advising the IRS on this point? That would explain their position.

[Kirk Maldonado]

Katherine:

I have basically the same question to you that I for DWK. Your posting contemplates an arrangement fundamentally different than that described in the PLRs and every such plan I've ever seen.

I think that IRS problem may be because they have the same misapprehensions about how these plans as DWK and Katherine. I had assumed that everybody would do their homework to find out how these plans work before voicing their opinions in public, but I'm beginning to think that my assumption was wrong.

I would be interested to see if you two still believe that it would be impossible to draft such a plan in a way that complies with section 409A, if you assumed the struture involved in those PLRs.

[Guest LeeNunn]

I posed the question to Bill Sweetnam and Dan Hogans directly during the call last Wednesday. I confirmed that non-qualified payout elections that are tied to qualified plan elections are permitted until Dec 31, 2005.

Wrap plans that distribute money from the non-qualified plan into the 401(k) plan will not be permitted after 2005 because the distribution is not one of the six events. That's when Dan (or Bill, I can't remember) made the comment about "not dead yet" and made the analogy to Monty Python ("bring out your dead" from Holy Grail).

Finally, I asked about diverting deferrals into the non-qualified plan after the deferrals reach the 401(k) elective deferral limit. Dan said that they have some concerns about this, and advised us to wait for the next round of guidance.

[Kirk Maldonado]

I disagree that there is a "distribution." How can there be a distribution when the amounts are not taxable to the individual? That is taking a very aggressive interpretation of the word "distribution."

[mbozek]

Are we getting the realm of esoteric distinctions in determining a distribution under 409A? Does distribution occur when funds are moved from the control of a NQDC to a 401k trustee or does a distribution only occur if the NQDC funds are made available to the employee, and not to another person? This question reminds me of the former US President who said that his answer to a question in a deposition depended on "what the meaning of is is." Aggressive interpretations may be the norm under 409A in order to insure that the IRS will collect the $1B in taxes that Congress estimated would be paid from 2005-2014.

[Guest Mark Wincek]

Lee, thank you for your post. The response that you received was more positive than the feedback we received regarding NQDC plans that spill over into 401(k) plans during the JCEB teleconference last Monday. We did not have time during that discussion to drill down on spillover designs, but Bill made a general negative comment about them, without noting any relief under Q&A-23. On its face, Q&A-23 permits tying the payout under a NQDC plan to "a payment election made by a participant under a qualified plan." Because the amount that spills over from the NQDC plan into the 401(k) plan is based on the ADP results of the 401(k) plan (not a participant payment election), it would seem Treasury could take the position that the spillover is not covered by Q&A-23. On the other hand, the goal of a transition to the new rules that is not unduly abrupt and disruptive would be served by allowing some relief for spillover plans for 2005. Based on the exchange you participated in this past Wednesday, do you understand Treasury to be reading Q&A-23 broadly in the interests of transition relief for spillover plans, or is there another explanation for their suggestion that spillover plans are not dead yet? Thanks again.

[E as in ERISA]

Katherine:

I have basically the same question to you that I for DWK.  Your posting contemplates an arrangement fundamentally different than that described in the PLRs and every such plan I've ever seen.

I think that IRS problem may be because they have the same misapprehensions about how these plans as DWK and Katherine.  I had assumed that everybody would do their homework to find out how these plans work before voicing their opinions in public, but I'm beginning to think that my assumption was wrong.

I would be interested to see if you two still believe that it would be impossible to draft such a plan in a way that complies with section 409A, if you assumed the struture involved in those PLRs.

Kirk:

Why would you assume that the Treasury/IRS is talking only about the types of mirror plans described in PLRs when they say that the coordinated 401(k)/NQ plans are dead in the water. I certainly don't. The Treasury heard about a variety of tandem, mirror, spillover, wrap and other types of coordinated plans not covered by the PLRs when they were drafting the regulations. I assume that their comments are directed to many of those other plans.

You're saying that Treasury personnel must not be intimately familiar with how tandem plans operate or otherwise they wouldn't be saying that they are "dead in the water." And you're admonishing some of us for not doing our research before we speak in public. Yet you are saying that you've never seen the plan I describe...you don't seem to be aware of what is actually in operation out there. (Maybe you've only seen the plans you've designed yourself -- which are within the confines of the law -- and aren't aware of what everyone else is doing....?)

I am by no means saying that all of these designs are legal under either prior or current law. I was shocked when I first became aware of the extent to which they were being used -- with the assumption that they are legal! But I am saying that there are lots of plans out there with this design -- and they were discussed with Treasury when regs were being drafted -- and Treasury's comments are probably covering more than just the prescribed PLR designs.

The first post recognized that there are at least two designs: "The scenario discussed was a NQDC that pours deferrals into a 401(k) after year-end to max. allowable limits, then a similar dovetail comment was made about arrangements that work in reverse - k to the max, then to NQ."

I think that the first type is generally designed consistent with the PLRs that allow a transfer to be made from the NQ to the 401(k) after ADP testing is done, etc. I had heard that they were not considered abusive and they might be allowed to continue. I'm assuming -- like you -- that there will be some way to continue those plans...possibly after some clarification in regulations and rulings.

But there are a VARIETY of plan designs out there when you talk about the second optin -- the "reverse" -- where deferrals go into the 401(k) first and then the NQ. Specifically there are a number of plans where all deferrals during the year go into the 401(k) until the 402(g) limit is hit, then deferrals go into the NQ (the scenario I described). Deferrals into the 401(k) can be changed throughout the year or be ceased at any point, and its possible that no deferrals would go into the NQ because the 402(g) limit wouldn't be hit. I'm not saying that these are legal or have been okay'd by the IRS. There are certainly many that are very questionable under prior law based on constructive receipt rules. But make no mistake -- those designs are very definitely in use. I think that the Treasury is now very aware of their operations -- I know that they heard questions and comments about them while they were drafting regulations.

So I would not assume that they are only talking about the plans that are prescribed in PLRs when they make comments about tandem, mirror, spillover, wrap plans, etc. I think that everyone needs to be aware of these alternate plan designs when they discuss the comments that Treasury is making -- or you may be missing the boat on which design that they are talking about at any particular moment. The words used to describe them -- tandem, mirror, spillover, wrap -- are not "terms of art" that apply to only one specific design. The same words are used by different persons to describe both legal and illegal designs. I think that it is primarily the illegal plans that they are talking about when they are saying that they are "dead in the water." It was my understanding that they have always been very concerned with any plan designs that left any control in the hands of the participant (i.e., where the "amount" could be varied during the year).

I don't think that it should be assumed that all of them are abusive. Many of them are used by large employers who want their management to get matching contributions on all deferrals (401(k) and NQ) but they want to minimize manual intervention in order to do that -- they want the payroll system to handle it. There are ways for the participant to define in advance what "amount" they want to defer -- in terms of a specific percentage or maximum dollar amount. But there are so many complexities. With changes in pay, variable compensation such as bonuses, comp limits, 402(g) limits, administrative limits on contributions based on prior year ADP, catch-up rules, etc. it is difficult to get the payroll and benefitss systems to do a true up during the year. It's much easier to do this by first putting all contributions into the 401(k) up to 402(g) limits and then fill the NQ after that.

Can you prescribe a design that will work under 409A. Or are they dead in the water, as Treasury indicates?

[E as in ERISA]

I think that it is kind of like assuming that the Treasury and IRs are only talking about the "legal form of haircuts" when they discuss how 409A applies to them. : )

[Kirk Maldonado]

I'm surprised that anybody is saying that you can have the amounts in excess of what is permitted under 401(k) go to a nonqualified deferred compensation plan. That alternative came up about the time of the first PLR on a wrap-around plan, and Treasury representatives made public comments that were picked up in the benefits press that those arrangements don't work. (I agree with them on that point.)

I agree with your comment in that I was myopic in assuming that when the speakers were referring to cash-back plans that their remarks included those of the type sanctioned by the PLRs. It could very well be (and I certainly hope) that they were focusing on some of the other variants.

All of the wrap-around plans that I've drafted were designed along the lines of the PLRs.

I hope that Treasury gets a lot of comments about why it should be possible to design wrap-around plans in a way that complies with Section 409A.

[mbozek]

I dont know how a 401k plan can be drafted to provide that excess 401k contributions will be transferred to a NQDC when the 402g regs explicitly state that the excess is included in the participant's income for the year contributed. I question whether the IRS ever issued a determination letter where this feature was disclosed given the disqualification of the plan if the the excess is excluded from income. The only way this could work would be if the excess 401k contributions were returned to the participant before 12/31 and an equivalent amount contributed to a NQDC from the ee's last pay for the year pursuant to a prior salary reduction agreement.

I also dont understand why the treasury reps would need to note that this type of arrangement is not allowed under 409A when it was never allowed under previous law.

[E as in ERISA]

My understanding was that under the old rules the way to coordinate between a 401(k) and NQ was to use a design like in PLR 9530038 or 199924067 (deferrals deposited in the NQ during year and moved to 401(k) after year end).

But there seems to have been a lot of loosening of plan design over the past several years. Some don't appear to believe that the NQ election needs to be established before the beginning of the year.

They don't transfer from the 401(k) to the NQ. They just put the first dollars into the 401(k). Then when the payroll system detects that the 402(g) is reached, it starts putting all future deferrals into the NQ. But if someone reduces, eliminates or increases their 401(k) elections during the year, then obviously the amount going into the NQ will be similarly reduced, eliminated or increased. So there is potentially a lot of flexibility during the year in regard to the amount of the NQ deferrals.... I'm not saying that was legal under old rules....just that is what has been happening....(I'm also not saying that it was done for abusive reasons...An NQ election could be definitively stated to be an amount or percentage of comp offset by whatever the 402(g) limit was for the year -- regardless of whether the 402(g) limit went into the (k) plan. But there became a need to have a way to coordinate between the NQ and (k)....because employers only want to match a certain percentage of pay....but matching contributions are made on a payroll basis...partly due to administrative limits for HCEs under prior year ADP testing..and catchup rules that essentially require you to define matching on a payroll basis if you don't want to match them. Its difficult to try to true up for hundreds of employees... I think that some changes were just made at the administrative level without the lawyers being aware of it....)

[Kirk Maldonado]

I had never seen the plan design that you mentioned, although I've seen several wrap-around plans with other design defects.

I'm not aware of any legal precedent that has occurred since the PLRs that would approve such an arrangement (or anything that predates those PLRs). If anybody who implemented a plan without either (a) conferring with legal counsel or (b) adequately warning the client that this is an aggressive approach, they have should be concerned.

[Guest LeeNunn]

Mark, I really enjoyed your ABA teleconference with Bill and Dan.

I think spillover plans will have some relief until Dec 31, 2005, not so much from Q-23, but from Q-21. This is consistent with Treasury's observation that employers aren't expected to have been clairvoyant. Q-23 focuses on payout elections and Q-21 focuses on deferral elections.

After 2005, spillovers from NQ to Q will be a problem because the spillover is not one of the six permitted distributions. Spillovers from Q to NQ depend on the design. Treasury worries about NQ deferral elections that depend on qualified deferral elections, so the answer seems to be hardwiring. A plan should be able to assume that 401(k) elective deferrals occur as originally elected. In other words, a change to a 401(k) election has no effect on the NQ deferral election.

This creates administrative headaches, but Treasury seems to have little sympathy for plan administration. It's just a question of cost to the employer.

Another challenge in tying non-qualified deferral elections to qualified deferral elections is the deferral of bonus compensation. The timing of the elections are different. I'm sure Treasury will give us plenty of guidance on this as we prepare for 2006.

[E as in ERISA]

They were not very clear, but my understanding of what they said today was that they were least concerned with their approved design and most concerned about other designs where changes in the 401(k) change the nonqualified amounts.

[Guest LeeNunn]

That's not the way I heard it. I think Treasury believes that it's obvious that transfers from NQ plans to qualified plans don't work after 2005. They're more concerned with NQ elections that are contingent on qualified elections because that's less clear. This is not the same as a distribution from the qualified plan to the NQ plan.

Suppose an executive elects to defer 10% of salary - first to the qualified plan and then to the non-qualified plan. Some plans, or some payroll systems, increase the NQ deferral if the employee reduces the qualified plan election. This is a problem. The solution is to essentially make the entire election irrevocable for deferrals in 2006 and beyond.

[Kirk Maldonado]

LeeNunn:

At the last teleconference they said that those arrangements were dead in the water. This teleconference they backed off an awful lot. I interpreted their remarks as meaning that some arrangements do work, but they've not worked out exactly what the acceptable parameters are.

[E as in ERISA]

I gleaned that they were only concerned about the "amount" issue today -- and its definitiveness. Accordingly, I interpreted their admittedly vague comments about what was okay to include the PLR approved method because I had previously heard that they don't consider that abusive. They indicated that they were primarily concerned about situations where the amount going into the NQ would be affected by changes in the 401(k) (and that is not what happens in the PLR situation).

[TCWalker]

I agree Kirk, I would interpret the attitude today as a softening on the "wrap" issue, I suspect in part do to a growing awareness of the prevelance of these arrangements with large employers. I think my notes left it as a matter of "consistency with legislative intent and construction of 409A" subject to future guidance.

[Guest Harry O]

Lee -

The plan you described does have a problem, except it is the qualified 401(k) plan that has the problem. That type of plan violates the contingent benefit rule under the 401(k) regulations as is not a qualified cash or deferred arrangement. That is nothing new with 409A, it is a qualified plan problem.

[mbozek]

At a recent seminar on 409A a speaker stated that the IRS had issued favorable determination letters for 401k plans that transferred excess contribuions to a NQDC plan. The employees elected to contribute the max ADP % to the 401k plan and the excess would be deposited in the NQDC. He said that under this election there would be no tax of the excess as a refund under the 401k plan regs because it was never an asset of the 401k plan. Don't know how the plan got around the requirement that 401k contributions cannot be conditioned on making contributions to another plan or program but I suspect the IRS never saw the election form.

This led me to think that the IRS wanted statutory standards for NQDC so that it could flush out all of the aggressive arrangements created by tax counsel which were below the IRS radar screen. Now the NQDC rules will be transparent and create a level playing field for all employers.

[E as in ERISA]

mbozek -- So had they only in effect authorized it from the 401(k) standpoint as opposed to the nonqualified standpoint? And did they discuss the timing of those distributions from 401(k) to nonqualified? Even if the qualified plan could get a determination letter, could there still be an issue with constructive receipt rules from the nonqualified standpoint. There had been some early 90s nonqualified PLRs that allowed the transfer from the 401(k) to the nonqualified and then I thought that they pulled them when they issued 9530038.

[mbozek]

I dont have any more info other than the NQDC contributions were made after the ADP limit was reached.