Does this offset plan appear to pass 411?

[Gary]

A plan has a formula that is 1.5% of avg pay for each year of service, offset by 66 2/3% of Social Security Benefit ("SSB").

This formula produces an accrued benefit of 0 (zero) for many years, since the offset is so high.

It would appear to violate 411 in design.

However, the Plan provides that the SSB be based on pay while in the Plan only, with no projection to age 65 or prior to date of hire.

Does anyone know, if this feature then makes the Plan pass 411 in design?

Of course, since the SSB has relatively high minimums, the accrued benefit is still less than 0, for many short service employees.

Any comments would be appreciated.

[Alonzo]

This plan probably fails because of the Social Security bendpoints. The fractional rule is the only method that would give you a chance of passage, and you will find that at least some participants are getting less than the fractional rule benefit. The plan undoubtedly flunks the 133% rule.

[Guest]

Alonzo,

Please help educate the social security-challenged!

What exactly do you mean by social security "bendpoints" and how does this impact the backloading analysis?

[Gary]

The Plan sponsor states that the Plan meets the 4/3 rule by design. Based on the gross portion of the accrued benefit, which is simply 1.5% per year, I agree it would pass. However, the SS offset causes benefits to be zero for short service employees. The gross portion of the accrual is a unit credit accrual, and the offset is 66 2/3 % of the SS benefit, with no unit credit accrual or service prorate. So it would seem to violate backloading issues and fail 401(l) as well. The SS benefit is supposed to be based on pay with the employer only, and one might say that this sufficiently applies the accrued SS benefit and does not have to be prorated. However, we still have zero accruals for several years for new employees and the result still causes backloading.

The question is, how does it specifically violate 411 (not just by saying a backloading violation), and fail for eg. the 4/3 rule? And what is a suggested or possible remedy to this formula?

[Alonzo]

This plan does not pass by design. My guess is your client, or the persons that advised your client are relying on Reg. 1.411(B)-1(B)(2)(D), because of its title and its reference to "social security benefits". But you need to read that reg very carefully, because it is the source of your problem.

If you read the regulation absolutely literally, you assume that the participant's social security benefit will never change and the participant will continue to receive compensation at his current rate. Assuming the participant has a typical compensation pattern (prior year compensation lower than current year), you will have 4/3 test failures in years where a participant gets a substantial raise.

I believe the consensus on interpreting this reg is to assume "social secutity benefits" means to assume the social security formula remains the same As you know, social security is essentially a career average plan that provides a fairly high portion of compensation to the lowly compensated, and a lower portion of compensation to those who make close to the wage base. (The place where the formula changes is the "bend point".)

To test to see whether a participant will meet the 4/3 rule, you assume that the participant has earned the compensation he's earned in the past, and will earn his current year compensation in the future, and use that to determine both the gross benefit and the offset. Because of the bend points, an employee with a typical accrual pattern will have jumps in his accrual rate.

One note. the 4/3 rule says that all projected accruals in the future will nt exceed the current accrual rate. A gradual acceleration is not acceptable.

Note. I learned this rule by some very harsh experience. A major client was sued over a similar fact patern. They're still in litigaton.

[david rigby]

"...the 4/3 rule says that all projected accruals in the future will not exceed the current accrual rate."

Almost. This rule is found in IRC 411(B)(1(B) and is referred to as the "133-1/3 Percent Rule". It states that the accrual rate for any later year is not more than 133-1/3% of the accrual rate of any prior year.

[Alonzo]

Excuse the glitch. I meant any projected future year can't be more than 133-1/3% than the year being tested.

[Gary]

Alonzo and Pax. Thank you for your excellent responses. I will read the reg you cited very closely. I agree with your methodology of frozen comp and SS benefit. However, my concern was that you compute the benefits for each year whne doing the 3% rule and the fractional rule, but the 133 1/3% was just design based, thus if the formula was 1.5% on the gross end and the SS benefit complied with 71-446, then the rate was 1.5% and maybe passed the 133 1/3% rule. Of course, when testing the benefit accrual year by year, it obviously fails, since the accd ben. is zero for several years.

So, now my question is: Where in the regs of the 133 1/3% rule, does it say that you have to test the actual benefit each year, as opposed to just observing the rate based on the plan formula?

In that litigation case you were referring, who discovered this error and filed suit? That is, was it an individual or a lawyer, etc.?

[Alonzo]

The case was brought by a fairly famous plaintiffs attorney in the benefits area. My understanding is that the guy had an ad in the AARP's magazine that advised participants who thought they were getting a raw deal with their pension should contact him. The participant who came in just missed out on some early retirement subsidies, and wanted to sue on the basis of the way his service was credited. The attorney took a look at the formula, and felt he had a profitable class action on his hands, based on the accrual formula.

There's no Rev. Rul. 71-446 exception to the accrual rules. This is why most offsets you find are pro-rated.

[Gary]

Was the company from Chicago?

Back to our original point. So you stand by the position that you don't just go by the gross accrual rate, but you compute the actual pattern of accruals for a participant, based on the required assumptions of level pay, etc.

One point about 71-446. Section 11 (Early ret under offset plans) states that for a benefit payable no earlier than 65, the plan is properly integrated if (in one case) "The old age insurance benefit which the EE would be entitled to at age 65 is based upon the assumption that he will not receive after severance any income which would be treated as wages for purposes of the SS Act." This implies to me that if the SS ben is based only on comp while covered by the PLan, then no pro ration is necessary and it just has to be less than the max allowable offset (83 1/3%). SO that was where i thought that perhaps the SS ben was done legitimately and that then the Plan by design strictly based on the gross formula, passed 411.

However, all of this is moot, if you have to check the ACTUAl accrual pattern as I understand you believe and what I think as well, but for various reasons play devil's advocate.

[Alonzo]

The 411 accrual rules were part of ERISA, which was passed after 71-446 was issues. Unless I had some post-ERISA authority, I would not rely on the 1971 revenue ruling.

[Gary]

One final comment. So then if say the avg pay = 40,000, and the SS ben = 12,000 (at age 65) and the person has five years of service his accd ben for 133 1/3% accrual rule purposes is under our current plan design:

.015 *5 *40,000 - 2/3 * 12,000 = 3,000 - 8,000 = 0, thus still no accd ben

And for 411 testing his accd ben projected after say 15 years (but computed today) would be:

.015 * 40,000 (remains constant)*15 - 2/3* 12,000 (remains constant)=

9,000 - 8,000 = 1,000.

Clearly, this goes from a 0% accrual rate to an accrual rate greater than zero (whenever it actually occurs) and violates the 4/3 rule.

This demonstrates how it fails 411's 4/3 rule, given that the accrual pattern is to be carried out for testing as opposed to just basing the test on the plan formula.

Is it confirmed that this test must be done, by carrying out the accrual pattern?

[Alonzo]

That's my reading. But, since you have no idea of who I am or what I know, I'd suggest you look at the following:

Reg 1.411(B)-1(B)(2)(i)(B). (This is what the testing rule comes from. Actually, you should reread all of -1(B)(2)

Rev. Rul. 78-252. (Which illustrates the operation of the accrual rules with respect to an obviously flawed PIA offset formula.)

Good luck. This is a tough one to resolve.