Backloading and 3% Accrual Rule

[Andy the Actuary]

An employer wants to consider a benefit formula that is 1 1/2 % of average compensation per year of service, maximum 50%. Further, he wants employees who have completed 15 years of service at age 65 to get a 50% normal retirement benefit. This appears to satisfy the 3% accrual rule of 411(b)(1)(A).

However, this means an employee hired at age 50 has accrued a 21% benefit at age 64 and then his benefit increases to 50% at 65. This smells worse than last weeks fish.

Assume there are no HCEs.

Can anyone explain why this dog don't hunt?

[Mike Preston]

Because it doesn't satisfy any of the three accrual rules? Why do you say that the first paragraph describes a benefit that satisfies the 3% rule?

[david rigby]

I agree with Mike. Perhaps there are some facts not yet in evidence?

Of course, you can wrap the formula in the fractional rule.

[Andy the Actuary]

Because it doesn't satisfy any of the three accrual rules? Why do you say that the first paragraph describes a benefit that satisfies the 3% rule?

Please someone help with this. The Normal Retirement Benefit for someone entering the Plan at the earliest age is 50%. 3% of 50% is 1 1/2 %. The benefit formula caps after 33 1/3 years. 411(b)(1)(A) states (Notice 2008-7):

Section 411(b)(1)(A) provides that a defined benefit plan satisfies the requirements of the 3% method if, under the plan, the accrued benefit payable upon the participant’s separation from service is not less than (A) 3% of the normal retirement benefit to which the participant would be entitled if the participant commenced participation at the earliest possible entry age under the plan and served continuously until the earlier of age 65 and the normal retirement age under the plan, multiplied by (B) the number of years (not in excess of 33 1/3 years) of his or her participation in the plan.

Please explain how the example misapplies this rule?

[Blinky the 3-eyed Fish]

I have to say I don't see the problem here. The max benefit is 50% of average comp. 3% of 50 is 1.5% so a participant can never have less than 3% of the max benefit times years of participation.

[Andy the Actuary]

Thank you Mr. 3-Eye. Would appreciate if any of the naysayers would be kind enough to explain their positions.

Perhaps, I'm sniffing the wrong fish trail. Could it be possibly that the age 65 benefit is age rather than service-related and that's a no-no?

The ERISA position was incorporated to prevent egregious backloading, which is what the suggested formula can produce.

[Mike Preston]

Why am I still concerned about the fish? Let's presume that the basic formula is 50% accrued over 33 and 1/3 years, which is the 1.5% multiplier. There is another formula/level is there not? The one that provides the uptick to 50% at 65/15? The more I write about this the more I think that it does, in fact, satisfy the 3% rule. But I agree with your initial assessment.

[Andy the Actuary]

Mike, thank you. Virtually everyone has the same impression. Phew. Yet, all agree the numbers appear to work.

There is some possibility that no one has ever thought to do this (I certainly hadn't considered it). And, no examples appear that would suggest this is doable or not doable. My gut is if you asked the usual authority, the answer would be "no." It doesn't need to go that far as this would frustrate the initial purpose of ERISA which was to nuke arrangements where employees were fired at age 64 to deny pension benefits. I simply hate to advise "you can't do it" when I'm unable to articulate why! But, I will and since this does not involve my domain -- N_x and D_x -- I would urge the client to seek a legal opinion, which they would have to do anyway because I have not seen and I doubt that there is a prototype plan that provides for such scheme.

[ak2ary]

I agree that this meets the 3% rule. So all of your problems are solved (as soon as you figure out how to slip the 29% annual accrual through the general test).

[Andy the Actuary]

I agree that this meets the 3% rule. So all of your problems are solved (as soon as you figure out how to slip the 29% annual accrual through the general test).

Yup, this is anything but a design-based safe-harbor and should have absolutely no problem passing nondiscrimination testing so long as you don't have any HCEs with fewer than 33 1/3 years at retirement!

[Mike Preston]

I agree that this meets the 3% rule. So all of your problems are solved (as soon as you figure out how to slip the 29% annual accrual through the general test).

Yup, this is anything but a design-based safe-harbor and should have absolutely no problem passing nondiscrimination testing so long as you don't have any HCEs with fewer than 33 1/3 years at retirement!

Or you have a PS plan that you aggregate with this plan, test on the basis of accrued-to-date and restructure until the darn thing works. Piece-a-cake.

[ak2ary]

Sure but if you were aggregating with a dc you wouldnt need this silly design for the db to begin with

[Mike Preston]

Sure but if you were aggregating with a dc you wouldnt need this silly design for the db to begin with

Possibly. It is, quite possibly, a design that is desired by the client for, get this, employee benefit reasons. Tell us, Andy!

[Andy the Actuary]

This was very interesting and while Mike is beining tongue-in-cheek in his request, it is worth sharing how a situation can evolve.

A 300 participant Client has a 50% flat benefit reduced for fewer than 15 YoS. The accrued benefit is project/prorate. (This is a non-safe-harbor safe-harbor with minimal general testing). The client wanted to reduce benefits for future hires and we had discussed keeing the 50% for the current population and reducing 50% to 35% for future hires and felt they should pass general test for a long while if not indefiitingly. Though I brought this up a couple times since our intial discussions, we hadn't discussed this for 6 months.

Then, out of the blue, I get an email that "they" had caucused and come up with 2% per year of service, maximum 25 YoS. When I demonstrated that this was a richer formula than what they had for the current population, they came back with the formula in question.

Their motivation was clearly to provide the same retirement benefits for the career employee but greatly reduce benefits for employees who terminated employment prior to age 65. This would have been a good pre-ERISA formula and they would have been fair in administering it (i.e., not fired people at age 64).

[mwyatt]

Will be interesting presenting the funding results at age 64, given the new PPA world of unit credit funding. Might want to have your client aware of the modest ballooning in TNC when that kicks in, to say the least...

[Andy the Actuary]

Will be interesting presenting the funding results at age 64, given the new PPA world of unit credit funding. Might want to have your client aware of the modest ballooning in TNC when that kicks in, to say the least...

Let assume (likely incorrectly^*) the client would adopt the formula and that miraculously, it passes the general test. The particular client has used aggregate funding using realistic, long-term assumptions, I would continue to urge they continue funding on this basis so long as greater than the PPA minimum.

^*Apart from all else, this formula flies poorly from a p.r. perspective.

The point you raise extrapolates to small stable populations where the ucnc escallates (like term insurance) with passage of time. PPA appears to accomplish what ERISA was trying to prevent: Pushing too much funding into the future.

[mwyatt]

Darn tutin' it does. Amuses (or basically amazes) me that the new 430 funding method was basically declared an illegal funding method for salary based plans pre PPA.

[Mike Preston]

Catch as catch can.

[SoCalActuary]

Darn tutin' it does. Amuses (or basically amazes) me that the new 430 funding method was basically declared an illegal funding method for salary based plans pre PPA.

Sounds like the financial economists were disrespectful of the IRS.

I recall how Wick & Holland reacted in January 2006 after reading the Senate bill. This was clearly different than their knowledge base.