Reduction in Accrued Benefit

[jpod]

Fairly plain vanilla DB plan; uses high 3-year average compensation out of last 10 years preceding termination (not last 10 "years of service," but last 10 years). Employee works 20 years and leaves. Comes back five years later and completes 3 or four additional years of service and terminates again (younger than NRA). Upon rehire he took a lesser position paying less than he made when he left the first time. His high 3-year average drops significantly, so much so that even with the additional years of service his accrued benefit is less than what it was before he left the first time! Is there some overriding rule in ERISA that says this is not allowed to happen? Assuming there isn't, do DB plans ever have plan language to prevent this from happening?

[Andy the Actuary]

The accrued benefit cannot decrease by increase of age or service -- 411(b)(1)(G). So, assuming he was not paid a lump sum, wouldn't his accrued benefit be the greater of (a) his accrued benefit after the first 20 years and (b) accrued benefit at most recent date of termination using the average compensation the Plan provides at this time and all of his service? I.e., it would not decrease from what he had previously earned (provided he was not distributed a lump sum).

It might be helpful if you would state the "average compensation provision" actual wording.

[jpod]

1. There have been no distributions, lump sum or otherwise.

2. "Final Average Compensation" means Compensation averaged over any 3 consecutive years within the last 10 immediately preceding date of termination during which Compensation is highest.

3. I've heard before that some believe that 411(b)(1)(G) might apply in this situation, but I don't see how that's right. The benefit is reduced because of the combination of a long period of separation and a reduction in compensation when he came back.

[Andy the Actuary]

What is the basis (code, regulation, ERISA Outline Book, etc.) for your conclusion in (3) that "you don't see how that is right"??? That is, where do you read that it is okay to decrease the accrued benefit?

Another way to look at this is that you have two participants -- (a) is a terminated vested with an accrued benefit and (b) is the participant terminating now who will get the difference between what the formula produces and (a), or $0.

[jpod]

The statute and the reg say that the AB cannot be reduced "on account of" any increase in age or years of service. It is not being reduced on account of age, or on account of years of service. It is being reduced because his comp. went down. If there was published IRS authority for what you seem to be saying, the employer would be happy with that result result and could simply apply that rule administratively without having to amend the plan. Unfortunately there is no such authority, or at least I could not find it.

If the drafters of ERISA intended that an accrued benefit could never be reduced as a result of the operation of a perfectly valid formula, I think they would have drafted the statute to say so. They would have said something like "your AB can only stay the same or go up but can never go down." However, what they said is that it can't be reduced "on account of" any increase in age or years of service, which has to mean something much narrower than "it can never go down."

[david rigby]

The IRS let its position be known clearly in the 2008 Gray Book:

Gray Book 2008-42

Other DB Plan Issues: Declining Accrued Benefits

May an accrued benefit decrease during continued employment due to any of the following:

a) Increases in a Social Security offset?

b) Increases in covered compensation?

c) Reductions in average compensation?

d) Reductions in the maximum benefit limitations under 415 (other than legislation or changes in response to a variable index)?

e) Investment performance underlying a variable annuity?

RESPONSE

a) Yes, but only to the extent that the offset meets the restrictions specified in Rev. Rule 84-45 and is in keeping with the qualification rule stated in IRC §401(a)(15).

b) and c) No. In this situation, the reduction would be on account of increasing service since the reduction would not occur if the participant terminated employment. A reduction in benefits due to increases in age or service would violate IRC §411(b)(1)(G). This was the rationale behind the answer to Question 33 from the 2003 Gray Book which dealt with situation c) above.

d) Where a post age 65 actuarial increase would be limited by the compensation limit as capped by IRC §401(a)(17), the benefits must be started, or “suspended”, to avoid an impermissible forfeiture. Benefits accrued prior to the IRC §415 regulatory effective date would be insulated from having to make this change and could continue to provide actuarial increases in spite of the 401(a)(17) limit. In addition, note that for benefits accrued after the regulatory effective date and prior to adoption of plan amendments, regulation §1.411(d)-3 would limit the ability to add a suspension of benefits approach. Moreover, for participants who have attained age 70-1/2, suspension of benefits would generally not be permitted.

e) Yes. In this case the reduction is not on account of age, service or plan amendment.

The above Response is a summary, prepared by representatives of the Program Committee, of the oral responses to the question posed to certain staff members of the Treasury and IRS, which represent only personal views of the individuals who provided them. Accordingly, the Response does not necessarily represent the positions of the Treasury or the IRS and cannot be relied upon by any taxpayer for any purpose.

Copyright © 2008, Enrolled Actuaries Meeting

All rights reserved by Enrolled Actuaries Meeting. Permission is granted to print or otherwise reproduce a limited number of copies of the material on the diskette for personal, internal, classroom, or other instructional use, on the condition that the foregoing copyright notice is used so as to give reasonable notice of the copyright of the Enrolled Actuaries Meeting. This consent for free limited copying without prior consent of the Enrolled Actuaries Meeting does not extend to making copies for general distribution, for advertising or promotional purposes, for inclusion in new collective works, or for sale or resale.

[Andy the Actuary]

The statute and the reg say that the AB cannot be reduced "on account of" any increase in age or years of service. It is not being reduced on account of age, or on account of years of service. It is being reduced because his comp. went down. If there was published IRS authority for what you seem to be saying, the employer would be happy with that result result and could simply apply that rule administratively without having to amend the plan. Unfortunately there is no such authority, or at least I could not find it.

If the drafters of ERISA intended that an accrued benefit could never be reduced as a result of the operation of a perfectly valid formula, I think they would have drafted the statute to say so. They would have said something like "your AB can only stay the same or go up but can never go down." However, what they said is that it can't be reduced "on account of" any increase in age or years of service, which has to mean something much narrower than "it can never go down."

Believe it is being reduced because of service. Had he come back to work but during his out period the plan had been amended to preclude participation or reparticipation, his accrued benefit would have remained at the frozen state.

Question: Following your comment: Suppose he had started his pension when he first left and then came back to work. Would you then reduce his pension?

[jpod]

This is all very interesting, and quite a conundrum. The separation from service complicates the thinking process, and I agree it would be even more complicated if the pension had commenced at the time of the first separation. I don't see any stock language in the DB LRMs that articulate rules comparable to the answers to (b) and © in the Gray Book questions; I find that odd.

I am not an actuary and I don't think I can get my hands on the Gray Book. Does the 2008 Gray Book identify the Treasury/IRS people who would have answered those questions? I suspect one was Holland, but he's gone.

[david rigby]

Does the 2008 Gray Book identify the Treasury/IRS people who would have answered those questions? I suspect one was Holland, but he's gone.

I don't think the identity is relevant, but this provides an opportunity for clarification. Please read again the paragraph above that begins, "The above Response..." The language of the Gray Book is NOT authored by the IRS, but is a paraphrase of opinions expressed by IRS and Treasury representatives. The Gray Book carries NO legal weight, but is intended to reflect IRS/Treasury opinions, especially on interpretations that may not have been anticipated or included in formal regulatory guidance. Sometimes, as the law changes, opinions in the Gray Book will change.

[jpod]

D. Rigby, I understand your caveat, but then why are you citing it like it's the gospel (you said "the IRS let its position be known . . . .")? Also, identity is very relevant. I would like to contact a knowledgeable person at IRS (hopefully one of the same people who offered the gray book answer) and inquire as to whether, if the plan follows the gray book answer, the plan could expect that if this issue is picked up in an audit the auditor would follow the unofficial answer given in the gray book, that's all.

[My 2 cents]

Consider what would happen if the person continued in service (i.e., no separation), with enough hours to earn full credit, but over an extended period of time pay is at a lower level (say more than 10 years)? For example, suppose that they were paid in part through commissions and increased competition or other factors have prevented the person from reaping the same level of commissions for many years, or suppose that the person is partially disabled and now employed full time but in a less well-paying position. When the higher, earlier compensation amounts begin to drop out of the 10 year period, is it necessary to calculate a grandfathered accrued benefit at the end of each plan year (as one is supposed to be doing for plans integrated at covered compensation)?

[jpod]

2 cents: Exactly! Your hypo illustrates how ridiculous the IRS position is, at least as to your facts.

[david rigby]

I never cite the Gray Book as gospel. No one is required to use it as guidance. However, often it provides value to actuaries, plan sponsors, attorneys, primarily to understand the reasoning behind some of the IRS positions. Sometimes it provides answers to a particular set of circumstances overlooked in prior regulatory guidance; even then it does not carry the force of law, but might be considered significant.

The introduction to the 2008 Gray Book lists the IRS/Treasury representatives as:

The following representatives of the Treasury Department and the Internal Revenue Service attended the meeting:

Internal Revenue Service:

Tax Exempt and Government Entities Division:

- James E. Holland, Jr., Assistant Director, Employee Plans Rulings & Agreements

- Martin L. Pippins, Manager, Employee Plans Technical Guidance and Quality Assurance

Linda Marshall, Senior Counsel, Chief Counsel, Qualified Plans Branch 1

U. S. Department of Treasury: Harlan M. Weller, Government Actuary

[FAPInJax]

Consider what would happen if the person continued in service (i.e., no separation), with enough hours to earn full credit, but over an extended period of time pay is at a lower level (say more than 10 years)? For example, suppose that they were paid in part through commissions and increased competition or other factors have prevented the person from reaping the same level of commissions for many years, or suppose that the person is partially disabled and now employed full time but in a less well-paying position. When the higher, earlier compensation amounts begin to drop out of the 10 year period, is it necessary to calculate a grandfathered accrued benefit at the end of each plan year (as one is supposed to be doing for plans integrated at covered compensation)?

You are correct. A grandfathered benefit would be necessary.

[Blinky the 3-eyed Fish]

If left to my own devices, I would take jpod's position based on my interpretation of the 411 language. However, I have read this Gray Book answer before and would definitely be cautious if going against it. Also, I have seen plan document language that does not allow accrued benefits to decrease, so be sure to read the doc (obvious, I know).

[ak2ary]

While those of you who say that the Gray Book cannot cited as authority are correct, it does represent the IRS and Treasury's best thinking on an issue and, in essence, represents their formal position. These questions are prepared and submitted to the IRS and Treasury well in advance. They have the opportunity to not answer individual questions if they do not want to. They have vetted their answers at higher levels inside their agencies. So while you cannot cite them as authority, you ignore these answers at your peril.

This particular answer, I believe is different from the IRS position from the early 1980's when I first learned the 411 rules. But I know I was surprised by this same answer in the 2003 Gray book

[ERISAAPPLE]

Does anyone know if it is still the position of the IRS that a benefit in a final average pay plan cannot be reduced as a result of a reduction in compensation?

[My 2 cents]

I may be wrong in this, but so long as the highest possible early retirement benefit (i.e., the largest amount that would have been payable as an immediate benefit had the participant separated from service and claimed an early retirement benefit, reduced to the extent called for by the plan) is treated as a minimum, I think the IRS has backed off their earlier position under which the deferred accrued benefit payable at NRA is protected from the adverse impact of compensation reductions.

See question 34 in the 2015 Gray Book, in which the practitioner summary of the comments received from the IRS indicates specifically that this is a reversal of what was said in question 33 of the 2003 Gray Book and question 42 of the 2008 Gray Book.  This arose from the implications of the IRS's final regulations on hybrid plans.  But, of course, while the IRS was participating in the lost, lamented Gray Book process, they explicitly indicated that nobody could rely on what the Gray Book said.

[AndyH]

Nice to see a Blinky Post (although I disagree with his initial comment - which didn't happen often)

I bet this subject has come up on the ASPPA listserv if someone wants to research it.

[ERISAAPPLE]

How do we buy the 2015 Gray Book?  I did not see it for sale on the Conference of Consulting Actuaries website.  I know we can't rely on it, but I would like to buy one.  

[My 2 cents]

There appears to be a reference to the 2015 Super Gray Book ($250) on the Conference of Consulting Actuaries website under archives.  Perhaps it can be purchased (as a CD-ROM) that way.  It would (if available) encompass all of the Gray Books from 1990 through 2015.  The IRS declined to participate in that process beginning in 2016.

[ERISAAPPLE]

I saw the reference, but I did not see a way to buy it.  Maybe I will call them.  I would like to get one before they are no longer available.  I know I can't rely on them, but I can't rely on any other secondary source either.