Floor Offsets versus DB+DC Combos

[AndyH]

Seeking opinions on non-safe harbor floor/offset plan designs.

Lately I have see many floor offset proposals and fairly new plans done by others. Some floor offset plans appear to comply with the Schultz memo ( .5% accrual before offset and uniform offset) and some clearly comply with nothing. I admit to a bias against non-safe harbor floor/offset plans on the grounds of concerns about the consistent interpretation of 401(a)(26) by the IRS, some ambiguity in the regulations, and the general "thin ice" that I perceive them to be bullt upon when there is a $0 net accrual.

In contrast, a DB/DC combo (cash balance or traditional) can do much the same with the complication that lots of people might have small benefits in the DB plan. But a CB accrual is less powerful for testing purposes.

Do others share my view or are others ok with full offsets and general testing of non-safe harbor floor/offset plans?

[SoCalActuary]

We do both. But we rarely use the floor-offset since the offset can fluctuate in an unpredictable way.

[tymesup]

If you allow participant direction in a floor offset, participants may decide to invest in high risk alternatives. Heads they win, tails the DB plan makes up the difference.

[AndyH]

Good points.

[Dougsbpc]

If the DB benefits to EE's are small to begin with, self-directed investments in the DC plan are not much of a problem. However, there can be a fluctuation problem for those with higher benefits. Before the plan is adopted recommend pooled investments for the DC plan. If they insist on self-direction then mention the potential for benefits to fluctuate.

With all the 401(a)(26) and other ambiguity, we decided a few years ago to instead just recommend combo plans without offsets. 1/2% of pay per year to EE's in the DB and at least 7.5% to EE's in the DC.

One of the biggest problems with $0 net benefits to EEs in an offset plan is internal. Many employers hate to distribute benefit statements and other reports to employees. In fact, I am sure some employers dont even distribute them.

[Blinky the 3-eyed Fish]

I echo some of the earlier comments. I will very rarely consider a floor offset unless those with net benefits in the DB plan are limited to the owner or the owner and spouse or if every owner's net benefit is at the 415 limit. I have found there is just too much potential for the offset to cause havoc amongst owners. I do have one plan where they pooled the offset balance and that does work fine.

One other wrinkle for which I don't have a great answer to is if an offset plan fails the nondiscrimination testing. The solution of course is to provide higher benefits to NHCEs. Normally, giving a higher DC contribution to one or two will pass the tests. But if you do that in an offset plan, you run the risk of not satisfying the "uniform benefits" of (a)(26)-5. I have heard that the IRS has blessed some plans that caps the offset amounts (like only contributions up to 7.5% of pay count toward the offset), but I have heard that is not uniformly accepted.

I too don't like the uncertaintly that the offset balance cannot be frozen, and so even if future DB accruals are ceased, the offset rages on. I don't necessarily agree, but it's another potential trouble point.

[Effen]

I don't do any floor offsets for a multitude of reasons, most have already been stated.

I do work with attorneys who draft them and they are telling it has gotten very difficult for them to get a floor offset document through IRS approval process. So much so that some documents that were approved on the last go round aren't getting approved this time. The IRS has even forced them to retroactively "fix" documents they previously approved.

I'm just throwing it out there. To me, floor offsets just aren't worth the aggravation.

[John Feldt ERPA CPC QPA]

With all the 401(a)(26) and other ambiguity, we decided a few years ago to instead just recommend combo plans without offsets.

Ditto. I'll be happy when the last few offset plans we have are gone, or when the IRS provides really usable guidance to clears things up for these plans - I wonder which will be first...

[AndyH]

Regarding the investment performance issue, I'm a bit puzzled. Let's say there is an owner 60 years old, and he gets a DB formula of 12% of pay per year and everybody else gets .5%, all offset by the act equiv of a 7.50% PS contribution. If the doc says so, don't you project the PS forward at, say, 8%, and the DB becomes clearly zero?

And, for testing purposes, if the measurement period is the current year, can't you ignore investment earnings, in fact, aren't you required to ignore investment earnings attributable to prior contributions?

So, unless I'm missing something, my concerns would be focused on issues other than the self directed asset part. But maybe I'm missing something, or the self-directed issue may be more prevalent in a safe harbor floor offset??

[Everett Moreland]

404© protection is not available for an offsetting defined contribution account the investment earnings of which can affect a defined benefit. See the preamble to 29 C.F.R. § 2550.404c-1, 57 Fed. Reg. 46906, 46907 & n.6 (October 13, 1992); K Brown, Specialized Qualified Plans--Cash Balance, Target, Age-Weighted and Hybrids, 352-4th Tax Management Portfolio A-76 (2009).

So far I have not had trouble getting a determination letter for a non-safe harbor floor offset plan, maybe because the language for how to calculate the offset is so detailed no IRS reviewer can bear to read it, or maybe because I've been lucky.

[AndyH]

Interesting, thank you. Maybe I can scare away people with the 404© issue.

[Blinky the 3-eyed Fish]

Regarding the investment performance issue, I'm a bit puzzled. Let's say there is an owner 60 years old, and he gets a DB formula of 12% of pay per year and everybody else gets .5%, all offset by the act equiv of a 7.50% PS contribution. If the doc says so, don't you project the PS forward at, say, 8%, and the DB becomes clearly zero?

Yes, the DB benefit is offset by the AE of the PS contribution. Actually, it doesn't have to be AE and could be some other means to convert it to a benefit, but AE is most popular. I don't understand the last question though. Converting the PS to an AE benefit involves projecting forward the balance as part of the calculation.

And, for testing purposes, if the measurement period is the current year, can't you ignore investment earnings, in fact, aren't you required to ignore investment earnings attributable to prior contributions?

I have seen the testing done either way with approval from the IRS, either ignoring the current year offset earnings or not ignoring,

I personally don't understand how ignoring the earnings can be correct since the increase in the DB net benefit during the year is affected by the earnings of the DC balance. It's part of the equation and should be accounted for.

[AndyH]

I see your point about the earnings. Don't know what the answer is, but I see the illogic.

[SoCalActuary]

My experience is that the offset is not always able to offset even a 0.5% accrual, since some employees are in their late 50's or early 60's. This is made worse when the DC assets take a 30% dump.

This makes for some surprised clients when they find out that there is an accrual, and PBGC coverage, because someone went over the DC offset value.

As to the testing, I see Andy's point, where you are only testing the current accrual. The value of the prior year benefits (including any change in their investments) is ignored for current year testing.

I am curious if the detailed language mentioned by EM is used to define the current year accruals for testing purposes.

[Everett Moreland]

I am curious if the detailed language mentioned by EM is used to define the current year accruals for testing purposes.

SoCalActuary:

The language defines the accrued benefit after the offset. It does not expressly define the current year accruals for testing purposes.