In a final average pay plan, is it customary to lock in any prior year

[Guest Linda Ursin]

In a final average pay plan, is it "customary" to lock in any prior year's accrued benefit and pay that benefit if it is larger than the benefit derived from the plan formula? This could happened where there is a significant decrease in comp. I can find no basis in 411 for doing this.

------------------

[david rigby]

No it is not customary. In fact, my understanding is that it is possible for an accrued benefit to go down if compensation goes down. Normally, the accrued benefit is defined in a formula which is based on service and final avg comp. The formula need not contain any reference to "last year's accrued benefit", etc.

However, the benefit at NRD cannot be less than the benefit payable at any ERD.

Also be careful with plan amendments.

[This message has been edited by pax (edited 12-01-1999).]

[Gary]

It would appear that the benefit could go down if pay decreases, but isn't avg pay defined over a period of 10 years? Where you could use earnings from several years past, thus maintaining the benefit based on the higher avg pay.

[david rigby]

Sure, average pay might be defined that way, but it might not. Average pay might not even be "final average", could be "career average". Etc, etc.

[Larry M]

To expand upon Pax's comments, the key to the question is found by looking at the plan document to determine the benefit.

It is rare the plan will use the term "final average pay". Rather, it will define the benefit in terms of a multiplier of "average compensation".

If "average compensation" is defined by using the participant's salary over his years of service or participation in the plan, then it is considered a "career average plan".

When we reduce the period over which the compensation is averaged by eliminating some of the early years, we have a "final average plan"

These final average plans may define average in any number of ways, the most common of which are

a. that of the last n years prior to termination of employment;

b. that of the highest n consecutive years in the last y; and

c. that of the highest n years any time

"n" is usually 3 to 5"

and "y" may be five to ten

If a is the definition, then we do not look back to determine whether the average was higher in earlier years.

If b or c are the definitions, then we look back to determine which years within the period created the highest average.

[again,as with the entire universe of qualified plans, there are exceptions to "we don't look back" comment...for grandfathered benefits or early retirement benefits]

[This message has been edited by Larry M (edited 12-24-1999).]

[Guest]

I agree that you really have to look at the plan language definition of what the final average pay is and then stick with that, even if the person may have had a lower salary in later years that would affect the formula.

[Guest Barnard Walsh]

ERISA does not allow the reduction of accrued benefits, (or does it?)

[david rigby]

Actually, the prohibition against a reduction in accrued benefits is that an *amendment* may not provide the reduction. If comp decreases, then the normal operation of the DB formula may result in a decrease in the accrued benefit.

See IRC 411(d)(6)

[Guest]

This issue has always intrigued me. I'll put on my pro-employee hat for a moment . . . Why doesn't a reduction in FAP from a prior year not violate the following IRC and ERISA requirements:

1. The normal retirement benefit must be the GREATER of the benefit at normal retirement age or the benefit available at any earlier retirement age. IRC 411(a)(9).

2. Benefits cannot decrease on account of increasing age or service under the plan. IRC 411(B)(1).

[Gary]

The issue of whether an accrued benefit can decrease due to a decrease in average compensation continues to be a perplexing one.

In question 9:25 of the 2005 Pension Answer Book, Stephen Krass addresses the issue of "can an accrued benefit decrease because of increasing age or service". Then goes on to give an example of a plan where the formula is 3% of AAC times years of service, where AAC is defined during the 3 consecutive plan years over the last ten that produces the highest average.

And then says "IRS representatives have opined that, if the definition of AAC causes a reduction of accrued benefit, the prohibition against reduction WILL be violated.

[david rigby]

They have opined that viewpoint "ad nauseum". But, have they ever shown where it is supported in the statutes?

[Gary]

With that said, then perhaps an administrator can take either position as long as they are consistent.

I have a small plan (one participant) that actually says that the accrued benefit should never decrease. Which to me is reasonable to say that the accrued benefit should be at least as much as the end of the prior year's accd ben.

However, it can produce funky results for eg.

A participant begins a business at age 45, starts a DB plan at age 55 and has NRA as age 62. And he provides for a pension of 100% of avg comp (3-year) pro-rata on service where service commences at age 45 for accrual purposes. He uses compensation beginning with the eff date of age 55 for benefit accrual purposes.

Say he earns $25,000 in the first year, then the AB would be 25,000 * (11/17) or 16,176 after one year (415 limit is 1/10 of 170,000).

In the 2nd year he has compensation of $0.

His AB would be 12/17 * 12,500 or 8,824, but since his prior year AB was 16,176 it would remain at that higher number.

A bit funky!