Cash Balance Plans - "Normal Retirement Age"

[Guest TooMuchFreeTime]

I have a client with a Cash Balance hybrid plan with an unusual definition of Normal Retirement Age; The earlier of age 65 and five years of service.

We filed for a determination letter on termination and this is the sticking point with the Service. They've cited Laurent v. PricewaterhouseCoopers in saying the definition violates ERISA, we've cited Fry v. Exelon, the 1.401(A)(4)-12 regs and 411(a)(8) saying that it doesn't.

There's no question that going forward, the new 1.401(a)-1(b)(1)(i) regs would prohibit the use of this definition. However, the plan was frozen years ago, and future accruals are not at issue.

I've talked to others in the field who were surprised we've been experiencing this much pushback and thought that the Service had settled on a position that these Normal Retirement Age definitions were permissible, but have been unable to identify any specific instances.

Has anybody else dealt with this issue? Has seen a favorable resolution?

[SoCalActuary]

Welcome to the new kinder IRS....just kidding.

The "earlier of" issue is worth fighting from the IRS position, because it grants authority for inservice distributions at NRA.

What do you lose if you give that position up?

[Guest TooMuchFreeTime]

What do you lose if you give that position up?

The issue is that this is an old, old plan. The sponsor froze it, then tried to terminate back in 1999 prior to the det letter filing moratorium on hybrid plans. Since that time, it's gone through two or three actuaries, and the only benefit information we have is a single benefit amount. If the sponsor was forced to go back and perform the whipsaw calc for all the participants, I don't know if that's even possible.

[AndyH]

I don't pretend to know the answer, but I'd like to understand the question better.

Why would you think that such a NRD would comply with anything? (No, I have not read the cases).

[ak2ary]

I have attended and participated in sessions where this topic was discussed with IRS speakers a few years ago who said they didnt like it but didnt see a way around it which is why the phased retirement regs outlawed it. I guess now that they have a case to latch onto..

The reason you would think that it is ok is because the law (prior to the Phased retirement regs) put no restrictions whatsoever on Normal Retirement Age. In fact existing guidance recognized that plans were allowed to have unreasonable Normal Retirement Ages, hence we had a rev rul that said it was not necessarily reasonable to assume for valuation purposes that someone would retire at NRA.

This is primarily a large plan cash balance issue, they used it to avoid whipsaw and for more leverage in their 401(a)(4) teststing...small plans didnt pick it up until much later (when they saw how well it worked)

Under the law NRA was simply the age of full accrual, the age that triggered full vesting, the need for actuarial increases/suspension notices etc...

In fact in the mid 70's, IRS withdrew existing NRA guidance (very similar to the phased retirement regs) because, I understand, they believed they had no authority to go beyond ERISA. When this happened, I was busy riding my Big Wheel so my memory is a little cloudy.

[AndyH]

Thanks for the info.

[Guest TooMuchFreeTime]

Why would you think that such a NRD would comply with anything?

When you say it like that, you make it sound dirty.

The answer generally, as noted by ak2ary, above, is that it's right because nothing says it's wrong.

However, there's more to it than that. When compared against definitions that are widely accepted (and expressly permitted under the Code), this definition uses the exact same metrics (age and service). The only difference is that it uses the EARLIER of the two events rather than the LATER. The effect is that in most cases, normal retirement age will occur much sooner. However, as noted by the regulations and notices referenced by ak2ary, prior to the new regs, there was no such thing as "too soon." So, how can you attack a normal retirement age for being too soon if there's no such thing as too soon?

You need to be very, very creative.

In the Laurent case, Judge Michael Mukasey (that name should ring a bell) did just that. He re-interpreted the old '78 notice (which he also mis-cited) to require that a Normal Retirement Age definition must be written with reference to an actual age, and only an age. Thus, any definition that incorporated a non-age-based metric (like, say, Service) is invalid. Not only is this NOT what the Notice says, but it's in direct contradiction with numerous provisions of the Code and Treasury Regulations. If this were the case, the vast majority of defined benefits would be similarly disqualified. (As this case was decided less than a week before his retirement from the bench, I wonder if he maybe checked out a little early).

The IRS has gotten in on the act, as well. While in the past they've taken the stance of "there's nothing prohibiting it, but we really don't like it," more recently they've stretched to find something to point to besides the Laurent case. Most recently, they've argued that they "expect" such a plan to violate the accrual rules of 411. Not only does a cursory reading of the 411 regs make it clear that they're talking about something entirely different, but you'll be hard pressed to come up with an interpretation that captures these plans, but doesn't also include all of the plans using a "later of" normal retirement definition.

All of this in light of an Internal Revenue Code that explicitly permits Normal Retirement Age definitions using Service as a component, using an "earlier of" approach, and even those that are measured by service EXCLUSIVELY with no reference to age whatsoever.

Going forward, it's clear these provisions (because of the possibility of very early retirement dates) would run afoul of the new regulations. However, retrospectively, all the Service has to go on is a bit of bile in the back of their throat. That's not supposed to be enough, but unless you want to subject your client to the time and cost of litigation (where, you hope, somebody might actually read the law before trying to apply it), you'll have to live with the Service's own review where all they need to do to support any position is say "this is our position."

[Mike Preston]

TooMuch....bravo. Not that I necessarily agree with the use of earlier of NRD's, since to me they do seem sort of.....dirty, but I absolutely agree with you that bile shouldn't be enough to enforce a position that is essentially based on "we don't like it". That kind of thinking is used over and over by those in the government who are basically saying: it is too costly for you to fight me, so I'll make up my own rules. If I had to write a paragraph, this week, for every time that has happened in the past, just to me and my clients, I wouldn't have time to sleep or eat or....well, you get the drift.

I remember very clearly that there were a number of reviewers at the IRS who just didn't like the fact that 401(k) deferrals are included in running the average benefits test. So, when they saw a test that would have a different result if 401(k) deferrals were excluded, they made a point of noting that the regulations didn't allow 401(k) deferrals to be included in running the average benefits test. It got to the point where one of our national organizations had to get a "sit down" with the powers that be in order to have them send the message, from the top down, that following the actual wording of the regulations was kind of important.

In any event, have you seen the article written by Ira Cohen extolling the virtues of the "earlier of" NRDs? I know he was at PWC at the time (that is, he had already left the service). I think it was published in an accounting journal. I'm sure SOMEBODY has a copy of it, somewhere, although I just tried to locate mine and it isn't jumping out at me.

That article may have some citations indicating why he felt that not only was it allowable, but that it was in the best interests of the public to promote it. Of course, it was written quite a few years ago, so its usefulness might be limited.

[david rigby]

Well stated. Essentially, this points out the ERISA and IRC were written long before anyone thought of hybrid plans, and the statutes simply do not recognize that they are neither DB nor DC as originally contemplated. If the IRS wants correct statutes and regulations, they should argue in favor of Congressional action that creates a new paradigm to cover hybrid plans, instead of forcing as is.

[Guest TooMuchFreeTime]

In any event, have you seen the article written by Ira Cohen extolling the virtues of the "earlier of" NRDs? I know he was at PWC at the time (that is, he had already left the service). I think it was published in an accounting journal. I'm sure SOMEBODY has a copy of it, somewhere, although I just tried to locate mine and it isn't jumping out at me.

That article may have some citations indicating why he felt that not only was it allowable, but that it was in the best interests of the public to promote it. Of course, it was written quite a few years ago, so its usefulness might be limited.

I was able to locate a copy of the letter as part of the amended complaint in the Laurent case. For anybody's who's interested, it can be found HERE.

If anybody has any additional thougths, I'm still open to ideas/suggestions/heckling.