Mike Preston

Unless a plan is established by 10/1 in a calendar year, it cannot satisfy the criteria set down by the IRS to be a safe-harbor plan. If, on the other hand, the plan year is established as a March 31, then establishing it by 12/31 would work. Not recommended for the faint at heart, though, as off-calendar 401(k) plans can be somewhat of a challenge.

The answer is that there is no answer possible based on the information that you have given. In some circumstances, yes. In others, no. But there is no prohibition against having both a SEP and a 401(k). For example, the SEP might have the box checked that requires one to work in at least 3 of the last 5 years. In that case, maybe the owner is the only one that satisfies that criteria. The 401(k) will have, at maximum, 1 year eligibility and therefore some of the employees would be eligible for the 401(k) but not for the SEP. It is also possible that the SEP is supposed to cover everybody and the client is just blissfully unaware of the requirement that a SEP cover everybody who satisfies the eligibility criteria.

I agree that if the plan is solely a safe-harbor plan then the plan does not need to provide the top-heavy contribution. Archimage, I think you and I are saying the same thing. It is just that you characterized the top-heavy minimum as encompassing the SH 3%, while I was saying that the SH 3% is different from the top-heavy total. But I think we agree on the result.

What WDIK is saying is that there are requirements to establishing a safe-harbor plan that go beyond checking the box. Specifically, they must issue a notice and the length of the plan year after which it is turned into a 401(k) must be at least 90 days. There is no way to satisfy either requirement. As far as the marketers go, just tell them that they need to follow the rules or it isn't a safe harbor plan. Somewhere, compliance should publish what those rules are. You should be able to say to the marketers: "Look at memo such and such. See requirement x? Was that met? See requirement y? Was that met?" If nobody is publishing, then there is a huge disconnect.

I don't think the result is the same. If the plan is top heavy the top-heavy minimum, which is likely to be 3%, is calculated on the basis of full year compensation.

They can set 'em up. They can even check the box that says it is a safe harbor plan. But it isn't a safe harbor plan in the sense that the ADP test must be run because, well, it isn't a safe harbor plan!

It depends on the number of participants that are defined as "affected". Once you have that number, you then need to determine the percentage of the total they represent. If less than 10%, it is pretty clear that there is no partial termination and there would be no need to vest. If greater than 30%, no question but that it is a partial termination. Anywhere in between is a judgment call that requires discussion with plan counsel. In fact, discussion with plan counsel sounds like a good idea even if the numbers are fairly clear. For example, some circuits, I believe, have come up with the astounding interpretation that the percentages determined should be based only on those that are not already 100% vested. That would significantly change the percentages. In short, there is nothing that can be determined from the information you have given and plan counsel is your best bet.

The posted URL merely confirms that there is an extension to January 31, 2004. It doesn't indicate that there is a subsequent extension, does it?

If the two companies are part of a controlled group, then you can do anything you want provided you could do the same thing if there was only one company and you wanted two separate plans. Think of it that way and you will be able to tell what you can do and what you can't do.

I have always felt exactly as mbozek describes. However, EPCRS seems to have a "theory" that moneys should never be paid out of the plan, even for the reason stated here. I wonder whether the IRS would really do anything to a plan that refunded monies in this situation? I somehow doubt it. However, with that said, since the EPCRS revenue procedure seems pretty clear on the issue, I think a plan Administrator would be well advised to get specific advice from counsel before taking this route without an EPCRS filing of some sort.

I guess it comes down to what is meant by "most valuable". I think Q&A 16 of 1.401(a)-20, in the case of a married participant, is about as clear as it can be: "In the case of a married participant, the QJSA must be at least as valuable as any other optional form of benefit payable under the plan at the same time. " I find a lump sum to be an optional form of benefit. Transparency is what these new regulations are all about. And if the lump sum is based on 417(e) rates of, 4.37% (check out the June 2003 GATT rates at http://www.irs.gov/retirement/article/0,,id=96450,00.html) it will be the most valuable benefit in every plan I've seen. That is, I've never seen a plan with more generous actuarial equivalence provisions, whether tabular or explicit. But I rarely see a plan offer to a terminating participant the immediate commencement of a QJSA that is actuarially equivalent to the, in this case anyway, "more valuable" lump sum. You say that you don't see any reason to adjust the QJSA based on anything going on with the lump sum. And I agree that your position is the norm. I've just never thought it to be compliant with the regulation in the case of a married participant. Not without more stretching of the plain language of the regulation (that is, interpreting how one determines "most valuable") in a way that is more likely to make the sound of a rubber band snapping than the 404a7 discussion in another thread. All, IMO, of course. I know my position is unpopular and I'm pretty sure that the intent of the regulation was not to cause the subsidized lump sum to increase the annuity required under the QJSA. But the reg was drafted long before 417(e) went into effect and the IRS could have, had they decided to, amended 1.401(a)-20 to clarify the issue by now. They haven't. Maybe the "intent" of the new regulations, which I perceive to be transparency for the benefit of the participant, will push the IRS to make a change to 1.401(a)-20 so that it is clear there is, in fact, no reason to adjust a QJSA based on anything going on with the lump sum. That would be fine. About the only way I can surmise that compliance with the reg is perceived is by stating that the lump sum is to be compared to the QJSA on the basis of the factors that are used to determine the lump sum. If that is the logic that is used, by definition, it must be the "same" value and therefore wouldn't violate the requirement for the annuity to be the most valuable. I just don't see it as being that clear. In fact, I think transparency would argue that it is unreasonable to assume that 4.37%/94GAR (aka 417(e)) rates are reasonable at the same time rates of say, 6%/83GAM-U, are being used as reasonable rates for other disclosures. Certainly a participant has the right to know that the lump sum based on 4.37%/94GAR is more valuable than an annuity that is actuarially reduced based on 6%/83GAM-U.

Well, certainly gets good credit. But Tom gets the prize! Or maybe Lorraine gets to share it for asking Tom to put it together.

Why do you have to consider the NHCE if the NHCE terminated before the beginning of the year?

[WARNING - THESE POSTS ARE GETTING LONG] I appreciate you helping me through this. You bring up the relationship between the most valuable benefit and the lump sum and that this relationship is specifically not addressed in the regulations, although there does seem to be one comment that lends support to your theory (if I understand your theory, which admittedly I might not). I'll address that first, and then get back to some specifics on your post. I am on record as saying that the 1.401(a)-20 regulation does, indeed, require the QJSA to be the most valuable benefit in all circumstances. Hence, if the 417(e) rates force a current lump sum to be increased above what it would otherwise be based on the plan's definition of actuarial equivalence, then there must be a corresponding increase in the immediately payable QJSA. Many people dispute this. I think your comment disputes it. I have heard, in support of what I think your comment is saying, that this was NOT the intent of the regulation, and I don't doubt that at all. But I've never seen anything from the IRS that makes it clear. And as I read the regulation it does not leave much wiggle room. In the newly published regulation, in the background section, there is one sentence that lends support to the position I think your comment espouses, but doesn't come out and say it: "Further, the anti-forfeiture rules of section 411(a) prohibit a participant's benefit under a defined benefit plan from being satisfied through payment of a form of benefit that is actuarially less valuable than the value of the participant's accrued benefit expressed in the form of an annual benefit commencing at normal retirement age." No mention of 417(e) there at all. In fact, this "tension", as it was referred to in the ASPA ASAP that was just published on this issue (author Barry Kozak of Chicago Consulting Actuaries) is specifically not addressed (directly, anyway) and the "Explanation of Provisions" section of the new regulation highlights this by saying: "Several commentators raised questions concerning whether the methods used in disclosing relative value of a plan's optional forms of benefit in accordance with these regulations affect the application of the requirement at Section 1.401(a)-20, Q&A 16, that the QJSA for married participants be at least as valuable as any other optional form of benefit under the plan. While this issue is not addressed in these final regulations, there is no requirement, or implication, that the same actuarial assumptions used by a plan for purposes of disclosing relative value in accordance with these regulations must be applied for purposes of the requirement in Section 1.401(a)(-20, Q&A -16, that the QJSA for married participants be at least as valuable as any other optional form of benefit under the plan." Not even I would think that these new sets of assumptions, which we are going to need solely for disclosure under the new regulations, would cause the actual benefits under the plan to vary. The statement from the reg. makes it clear that you can use different assumptions from those that are specified in the plan when developing disclosures. But it leads to a serious problem with communication if you intend to use the “example” approach. If I’m going to show that $1,000 of single life annuity is, as a lump sum, worth something other than $172,306 at age 55 (based on 94GAR/5.14%) and yet show that this person, who might have a single life annuity benefit of $1,000, is entitled to a lump sum of $172,306 my forms are going to look very funny indeed. I’m not sure what the above means. Are you saying it shouldn’t end up higher because of this reason, but it might end up higher for another reason? Or are you saying that it won’t end up higher because you are in my camp and believe that the QJSA must be the most valuable benefit so if the lump sum has a subsidy over the otherwise payable life annuity, that subsidy cascades to the QJSA? Example: $1,000 life annuity at age 65. Plan rates are 6%/GATT. Lump sum at 65 = $127,756.20. Participant is age 50, so plan rates lump sum (no preretirement mortality) is $53,308.20. However, 417(e) for a benefit distributable in January, 2004, based on 94GAR and interest of 5.12% (two month lookback – best example I could come up with because December rates still not published) require lump sum to be $66,249. Now, the actuarial equivalent annuity at age 50 using 6%/GATT would be $320.81 if I ignore 417(e). But if the increase in the lump sum from $53,308.20 to $66,249 must be recognized in the QJSA, it looks to me like I use 94GAR/5.12% at age 50 and come up with an annuity of $356.34. Or, of course, it might be neither of the two. I know this is not directly on point with respect to the new regulations, but it does highlight the circular nature of these disclosures. Can you give a short numeric example to see if we are talking about the same thing? But if there are other alternatives under the plan, if you use the lump sum rates for purposes of determining this particular optional form’s relative value (which I admit will be 100% of the life annuity because, by definition, if you use the same rates for converting one to another and then back again, you have to end up at 100%), won’t this just create a situation where the other optional forms now end up being something other than 100%? If I’m understanding this correctly, you are saying that it is ok to use different sets of actuarial assumptions for comparison of different optional forms. I’m not sure I see where in the regulation it allows this. If it does, then there is virtually no disclosure required at all in the case I posit: the plan that has no early retirement subsidies and all annuity options are actuarially equivalent to the life annuity payable. Talk about loopholes! Let’s work backwards. The new regulation appears to want disclosure of the LACK of subsidy built into a lump sum. Theoretically, this is so a participant doesn’t accept the lump sum when the annuity would clearly be more valuable. But if the EXISTENCE of a subsidy in the lump sum can be masked, aren’t we just moving the problem around on the table and putting the participant in a position where they might accept an annuity even though the lump sum is clearly more valuable? I guess to simplify, I’d like to know what is thought the necessary disclosures are in the numeric case I posited above: terminee age 50 with a benefit payable at age 65 of $1,000 month, and the plan has a lump sum option where the actuarial factors are clearly not as generous as the 417(e) rules require.

MGB, it looks like your interpretation is better than mine. I just naturally assumed that if the intent of the reg is to eliminate the uninformed participant that the existence of a future benefit entitlement would need to be disclosed. I see that might have been an expansion of the actual words in the regs. I'm only on my 18th time through them, so I'll reserve judgment on what they really mean (it takes me a long time to sift through the cross-references before I feel comfortable with something like this). It does seem like the focus is on those plans that provide two (or more) current options, one of which is a seriously unsubsidized benefit (typically the lump sum) and the other of which is subsidized (typically the QJSA) benefit. They want the participant to know that if they select the lump sum they are giving up a lot. In plans where the lump sum involves a subsidy, however, I'm struggling with just what they want to see disclosed. Seems like a simple statement along the lines of: "Gee, every benefit you can get from this plan is equivalent to the QJSA (which is typically the life annuity option), except the Lump Sum, which has an estimated relative value of X%" Where for this purpose, x% will generally exceed 100%. But since we can use "reasonable" rates for purposes of determining relative value, if the 417(e) rates are considered, by definition, to be reasonable, then relative value will always be 100% for the lump sum. I think you have to use the same reasonable rates for all comparisons, though, so if we use 417(e) rates as our reasonable rates, we can find that the other benefit options in the plan (such as an actuarially equivalent J&S) will now have different relative values because they might be based on, for example, 6%, 83-IAM-M, rather than 417(e) interest and 94GAR mortality. My preliminary conclusion is, therefore, to simplify the disclosures, that it would be better to use the plan rates as the reasonable rates for this purpose and then either add a statement that says the relative value of all benefits other than the lump sum is 100% and the relative value of the lump sum is x% (or, alternatively, to provide a chart that lists the relative value of the lump sum at representative ages - or at each age) and then at each place where a lump sum relative value is mentioned, also compute the lump sum for an immediate commencement $1000/month life annuity. But this will look very strange, because the immediate commencment lump sum at age 21 will be a very large number and it will decline each year as the chart shows higher ages. There has to be a better way. My focus is those plans that don't have early retirement or J&S subsidies and, instead, provide that all optional forms are actuarially equivalent to the single life annuity - with the exception of the lump sum, which due to 417(e), is the subsidized benefit. I would welcome more thoughts on the regs and how best to comply.

Actually seems like it might be a pretty simple transaction. If there are no due diligence issues then a simple merger seems like it will work just fine. Still have to deal with contributions for period before merger but within plan year. Strangely, not too many notification issues involved when PS plans with pooled investments merge. Participants will be new to a plan and therefore due an SPD, that sort of stuff.

Well, the least would be the 3% of compensation earned while a participant. However, it depends on your document. It may define the least as being 3% of compensation for the whole year. Also, if the plan is being used in a cross-tested plan, the minimum might be 5% (or more). So, you need to check the plan document to see what it says and then make sure that nothing else gets broken in the process.

You can't. But the QJSA you have today is not necessarily the QJSA you have in 5 years! Let's say you have an early retirement subsidy that is payable only once an individual attains early retirement age, but they can quit before early retirement age and just wait to get their early retirement benefit. Assume somebody quits 5 years before their early retirement benefit is payable with the early retirement subsidy. They can, if they choose, take a current lump sum. As you point out, if they can get a current lump sum, they also have to get the right to start an annuity immediately, although it won't have the early retirement subsidy. I think that is the issue the reg cite was trying to address.

Lots more. How are the monies invested? If there are individually directed accounts, will there be any blackouts? BE VERY CAREFUL ABOUT SOX NOTICES. What do the financial institutions where the monies are invested have to say about all this? How will the earnings be calculated? Who are the Trustees? What do they say? Any loans outstanding? How will they be administered before, during and after the transition? Does Company A know that they are taking on the warts of Company B's plan? Any due diligence been done? Any lawyers involved? This stuff is so simple. Yeah, right.

Who's Mr. Preston? Oh, yeah. That's my Dad! At least he's been to the upper floors of 1111 Constitution (get the tape from my session at ASPA a couple of months ago to hear how he got there). Anyway, I've already given my thoughts on this so I'll try to fade into the background.

Oooh, can we tell it is near the end of the year? We probably did blow by a26 a bit too quickly. Partially because that is a discussion that can drag on and on, as well. The definition of "reasonable and uniform" in the a26 regs has always caused me consternation because of something I heard in the early 90's which indicated that the design that Mike S. speaks of can work while other things I've heard more recently indicate that it doesn't. So I decided to ignore it....for now. We've also got the Paul Schulz memo to contend with. No matter, I'd prefer to focus on the issue that Mike S. has brought up to see whether there are any cites. This is purely selfish on my part, because I have a plan where the client would implement the floor in a New York second if it could, but doesn't want to lose the 404a7 exemption. While the NHCE's would be fully offset for the next few years, the client would like to see the DB provide a floor, just in case. To date, I've said: No can do. But I suspect the focus I prefer will no longer be possible. Que will be, sera. ;-)

It is not a multi-employer plan unless the benefits are subject to collective bargaining. I think you are talking about a multiple-employer plan, right? Yes, they can merge. Lots of i's to dot and t's to cross, but shouldn't be too bad.

Whether I agree or disagree is not the issue. If we can't pummel a point to death and argue mercilessly about angels, pins, dancing and things of that nature, Judge Learned Hand would be mighty disappointed with us. Remember, MGB, we are talking here about the ability to increase benefits for certain non-highly compensated participants. Yes, the same logic can be used to generate additional benefits for HCE's, but the sword cuts both ways (and must do so under our tax system). So, while I don't agree with it at the moment, I'm actually rooting for Mike to convince me that a cite does exist somewhere so I can layer a DC plan for NHCE's on top of a DB plan that is more generous than will fit right now under the 25% limit. Of course, I could suggest that the design of the DB plan be modified to reduce the benefits for the owners and managers. I'm sure you can tell where that suggestion will end up. These things really do cut both ways. I just want to make sure that, if at all possible, no blood is spilled by the edges of whatever happens to be sharp. Have I mixed enough metaphors for one day?

I think I hear the sound of a rubber band being stretched mighty tight. Those things hurt when they break, you know. I think that having the floor available satisfies the definition of being a beneficiary in the Webster sense. I'd be very surprised if counsel for your client didn't. But as far as whether it rises to the level of satisfying the definition of beneficiary as intended under 404a7 and the regulations thereunder, that is entirely up in the air!

Bet you can figure out how to get the next year's stuff, too!