Mike Preston

It is highly unlikely that the spouse of the deceased child would be entitled to anything. However, if the deceased child had a child, then you would need to determine whether the original beneficiary designation was "per stirpes" or "per capita." My document indicates that "per stirpes" is the default election, but that it can be over-ridden via a beneficiary designation that states "per capita." If the beneficiary designation was "per stirpes" (whether by election or by automatic provision of the underlying document) and the deceased child had an heir (what is legally known as the child's "issue" - child or grandchild, etc.) then the deceased child's share would be payable to the heir (or heirs) of that individual. Here's a website with some iinformation on the terms: http://www.recer.com/bwh/dostirpe.htm

"We received advice to send a letter to employee and prior sponsor that until we received either legal direction or agreement from the employee and prior employer that the rollover is incorrect, that we'd hold onto the entire rollover. " You might want to get a second opinion on that advice. Maybe it works, maybe it doesn't. As a participant, if I rolled over money, I know that I wouldn't want a letter from a third party indicating that the rollover was incorrect to potentially interfere with my ERISA rights in that plan. I received advice from a lawyer once in a similar situation where I was advised to write to the party claiming that the amount was "too much" and tell them that in the absence of a court order, my firm would have no choice but to follow the terms of the plan, including honoring any participant withdrawal requests. The other party then had a chance to decide how it wanted to proceed. Thankfully for us, they decided to have their attorney deal directly with the participant. I believe, had the participant not been willing to deal with them, they would have gone to court to force the participant to do what they thought was right. I don't know what the attorney that gave us the advice would have told us to do had the other plan decided to include our client's plan in a court order of some sort. If it was a state court order, knowing that attorney as I do, I'm pretty sure he would have suggested that the order be resisted. Of course, if it was a federal court order, then he would no doubt have suggested that it be followed. At the end of the day, though, in the absence of a valid court order, the advice no doubt would have been: "follow the terms of your plan, as you really have no choice."

The sanctions are imposed on the plan sponsor, not the document provider. The only way the document provider pays is through a claim made by the plan sponsor that the document provider is somehow liable for the failure. It is precisely those clients that do not keep the records necessary to complete an accurate plan document that will tend to blame others.

Well, it may be unfortunate, but the fact is that the IRS considers inclusion of the testing method in the plan to be a qualification issue. So, when you say that there may be a fine, you are right. It appears the IRS, on audit, would put you into Audit CAP under EPCRS. Last I checked, the penalties weren't as severe as they were before the last version of the IRS correction procedure. But they were still bad enough that it makes absolutely no sense to consciously play audit roulette.

Ah, so _that's_ the question! I don't think I have a magic bullet for you, but I strongly believe that the flexibility of pure class-based allocations is worth its weight in gold. There are certain circumstances where it is unneceesary, such as the one HCE plan or the multiple HCE plan where everybody knows, in advance, that the HCE's want to max out. In those situations a formula based solution, like a super-integrated plan, makes a lot of sense. Or, where it is impossible, such as where a money purchase plan is required in order to allow a contribution in excess of 15% of pay. Of course, this last reason disappears in 2002 for all states that have conformed to EGTRRA. However, even in those cases, things get a little sticky if the business model changes. I think the back-and-forth is limited, although not eliminated, if the allocation for the year begins with an estimate that is performed based on the client's preference. In the absence of guidance, I use the allocation parameters from the prior year. But, with guidance, if it works, there is no back and forth. Sort of puts a little meat behind the expression: "Well, how much do you want it to be?" Assuming the client says: "25% to Groups A and B, 10% to group C, 3% to groups D & E" and the test passes, there is nothing left to do. Assuming the client says: "$50,000 into the plan, maximize group A, minimize group E, and make Groups B, C and D equal" we can put tht together, too. In the event that the test fails at the desired client contribution levels, we typically can identify two methods of making it work: increase the allocation to the groups necessary to make it pass, or design a point-and-shoot -11g amendment. Here is where the judgement comes in. One can spend quite a bit of time trying to figure out the best mechanism for lifting a failed test into passing territory. I wish there was a good way to have the client determine, in advance, whether they want us to contemplate all the basic testing methods, such as accrued to date with permitted disparity, or not. And whether they want us to move into restructuring (amazing how that helps sometimes). But all of that takes extra time. Does that help?

Notice 98-1 states in Section VII "Accordingly, a plan is permitted to change from the current year testing method to the prior year testing method in any of the following situations: ... 4. The change occurs during the plan's remedial amendment period for the SBJPA changes (see Rev. Proc. 97-41). " Since the RAP has been extended through at least 2/28/2002, and in most cases until at least 12/31/2002, and in some cases maybe sometime in 2003, I think a calendar year plan that amends during its RAP can select whatever they want for any year up through the year that includes the end of its RAP. I seem to recall that the IRS takes the position that a plan that updates for GUST before its deadline is deemed to have closed its RAP. So, if a plan has adopted a GUST-approved plan before the end of 2001, I'm not sure that the general RAP extensions referenced above would allow such a plan to change methods in 2002. Unless, of course, the IRS has updated 98-1 specifically with respect to this issue such that there is a restriction of some sort. I'm not aware of any such restriction, but I haven't researched it, either. One thing that I've not seen clarified is when changes may be made effective. If one wants to change methods it is clear from 98-1 that an amendment must be made to the plan. So, assume that a plan sponsor wishes to change from prior year testing to current year testing for the 2004 year. Must that amendment be adopted before the 2004 year? During the 2004 year? During the regular RAP with respect to the 2004 year? Under EPCRS, sometime before the end of the second plan year after a failed ADP test, which in this case would be by 12/31/2006?

Yes. However, if the match was instituted in order to satisfy a particular provison of the Internal Revenue Code (such as a safe harbor match under 401(k)(12)(B)), then that provision would no longer be satisfied and the plan would lose whatever advantage it enjoyed by satisfying the provision.

This particular issue is somewhat of a thorny one. As I think I stated earlier, there is a general rule that a plan may not directly charge a specific participant who chooses to exercise an ERISA right that exists under the plan. There are a number of published citations on that and, I think, perhaps the best known is a DOL Advisory Opinion indicating that a plan may not charge an individual participant for expenses that are incurred by the plan in conjuction with the approval and processing of a QDRO. I don't have the exact cite at my fingertips but that is the jist of it. [Here's the cite: http://www.dol.gov/dol/pwba/public/program...ry94/94-32a.htm] The processing of a "normal" distribution certainly falls into that category and I'm sure the DOL would have problems with a plan charging a participant in order to get a "regular" distribution under the plan. So part of this revolves around whether or not this was a "regular" distribution or not. In that regard, it sounds like it wasn't. But just because it wasn't "regular" doesn't mean that the plan can charge excessive amounts to a plan participant Now, the work involved in having the plan process things in a way that is something other than "regular" can be minimal or it might be substantial. An example of a minimal extra expense might be a case of a distributioin offered by wire transfer in lieu of having a check mailed. I can see where a plan might be able to charge a participant for the expense involved in the wire transfer, because if the employee didn't want to pay the expense associated with the wire transfer, the participant would still have the right to the distribuiton in the form of a check. An example of a charge that might be substantial is where the plan is re-valued for all participants in order to enable a set of distributions (or a single distribution) earlier than what might otherwise take place. This could be thousands of dollars in some plans. I don't know whether such a charge could be passed on to the participant in the eyes of the DOL. I am somewhat doubtful. It sounds like yours is somewhere in the middle. So the question would be whether the additional charges that were imposed on the participant were the kind that the DOL would think should have been borne by the plan as a whole, rather than charged to the individual. Another fact intensive based inquiry, which could turn on a number of things. Does the plan offer this to everybody? Or was a special exception made? Is everybody charged the same amount? In the absence of agreeing to pay the additional charge in order to receive the distribution "early", would the participant be denied any of his/her ERISA rights? Is the charge so high that it serves to get in the way of ERISA rights available to most participants (whether or not this participant was willing to pay the charge)? Again, questions for the lawyer to try to get answered well enough to decide whether to include in the previously suggested letter. Or whether or not to ask the DOL for some support, if the plan is not reasonable with respect to the return of the disputed funds. Good luck.

Maybe the flush language following (B) serves to make the study notes, and you, correct. Thanks for sticking with this.

Well, I stand by my original interpretation. The language says that folks have to pay excise taxes on non-deductible contributions except for a certain amount of non-deductible contributions. You determine that certain amount by adding the dollar values determined under (A) to the dollar values determined under (B). The dollar values under (A) are only applicable if the plan has more than 100 participants. The dollar values under (B) are applicable whether the plan has more than 100 participants or not. You add up (A) and (B). If (A) is zero, you still get (B). Whatever you get, that amount is not subject to excise taxes. The linkage you identify by the 'AND" between (A) and (B) is the same as the plus sign in the equation: X = ND - [(A) + (B)] where X is the amount subject to excise taxes; ND is the total non-deductible contributions (A) is contributions identified with reference to subclause (A) (B) is contributions identified with reference to subclause (B).

I think the recordkeepers or the plan's fiduciaries may be on very thin ice here. If the plan or the recordkeepers really charged the individual participant for work required to allow the participant to exercise an ERISA "right" then the DOL might be very interested in this case. While it isn't 100% clear that this is what happened, it might get somebody at DOL interested, especially if the bullying can be documented. I have heard of plans charging participants for all sorts of things that the DOL thinks should not be charged directly to them. The DOL takes the position that such charges, if borne by the plan, should be treated as general expenses of the plan and allocated to all participants' accounts. While processing individual distributions in "normal" fashion seems clearly against the DOL's wishes, I've never heard an opinion on whether an "interim valuation" that is requested by a participant can legitimately give rise to a charge directly to the participant. That's an interesting question that the DOL may have an opinion on that I just haven't read about. The fact that there were special efforts taken on behalf of this individual means that the individual would, in my opinion, have very little reason to doubt the accuracy of the monies forwarded at the time, unless there were materials provided at the time that made it clear an overdistribution had taken place. So, you have a bit more you can add to that letter from the lawyer. But I reiterate that it is a very fact intensive situation that could turn on such things as what was really disclosed to the participant at or near the time of the distribution and whether such disclosures push some or all of the responsibility to the participant. Complicating this is that there may be two paths that the participant will need to traverse, if things get really, really sticky: federal issues and state issues. If the plan won't be reasonable in the eyes of the participant, then maybe the DOL is contacted for the federal issues and small claims court is contemplated for the state issues? Better have that lawyer I mentioned provide some guidance here!

Maybe I'm speaking (typing?) with my heart here, and not my head, but I think the participant should ask a lawyer (pre-paid legal services? brother-in-law?) to forward the participant's check for $1,000 (or whatever the participant believes is "left") to the plan as repayment of the over-distribution. In that letter, the lawyer might gently explain that if the Plan accepts the payment, then the participant will not ask for damages that may result from the participant potentially being liable for taxes on the overdistribution and the resulting excise taxes with respect to depositing (and not withdrawing on a timely basis) funds which were initially not eligible for rollover. Maybe I'm being too kind on the plan, here. In fact, after reading Q&A 130, maybe the participant would be stuck with the taxes on $4,000 so the participant will just keep the $1,000 (thereby making the participant liabile for taxes on $5,000) and turn around and ask the plan for additional funds to make the participant whole! If the participant can't get it from the plan, maybe the participant can get it from the service providers to the plan, who processed the overdistribution to begin with. Basic fairness (not a particularly strong legal theory, especially in the ERISA area) says that a participant should not be damaged by the plan's error (or the plan's service providers' errors - plural intended) if the participant had little reason to suspect there was an error. Of course, if there are extenuating circumstances not disclosed so far, like the fact that the participant was in possession of materials that made it clear that the overdistribution had taken place near the time of the distribution, rather than 2 years later, then the participant has nobody to blame but his or herself and should accept the loss and be thankful that there are no cases supporting criminal prosecution for not returning funds known to be in excess of one's entitlement. By its nature, this is a fact-intensive situation that is not likely to be fully understood here. The participant needs to determine if they were trying to get away with something, or if it could be construed that way by an independent reviewer. If so, the Scott case from Texas is fair warning enough that the penalties for non-cooperation can be steep. But if the participant is not the guilty party here, I am disgusted that the plan would ask for the full $5,000 back Strong letter to follow.

Blinky, my Code doesn't agree with you. The 100 participant requirement is specific to 4972©(6)(A) and doesn't have any bearing on 4972©(6)b). Is my copy of the Code incorrect?

I would encourage you to make sure your files are well documented as to the availability of the k plan, both as to deferrals and match, to the NHCE's. There is something that makes me think that, should the IRS look at this plan after it has been in operation 2 years, if there are truly no NHCE contributions, they will at least question things. But with that said, even if there is a required contribution to the SH match, and even though it won't give rise to a deduction, as long as the contribution is less than 6% of pay, at least the employer is not stuck with the excise tax on non-deductible contribuitons. See 4972©(6)(B).

I disagree with the assertion that one combines 401(k) plans for testing ADP and ACP results solely because there is a single entity, such as a controlled group, that maintains the plans. There are two situtations where aggregations is required. The first is IF the plans are combined to satisfy 410(B). If they are, then you combine them into one and test as if there were only one. However, if the plans independently satisfy 410(B), one does not need to aggregate them. Just to round out this paragraph, if the plans indeed stand on their own, you CAN still combine them voluntarily under 410(B). In any event, if they are combined under 410(B), you test them as one. In the case where you have separate plans that each satisfy 410(B) one is still required to count all deferrals and compensation to all plans for each HCE in each of the plans where the HCE participates. This is a participant/employee level aggregation, not a plan level aggregation. Perhaps a simple numeric example will make it clear why the IRS decided to impose this rule. Assume you have 1 HCE and 4 NHCE's. Assume you set up 4 401(k) plans, each of which has one NHCE in it and each of which allows the HCE to participate. For now, let's assume that each plan satisfies 410(B) [it can be done under the right circumstances]. Now, imagine that each NHCE puts in 1%. If the HCE were allowed to avoid aggregation, the HCE could put in 2% in each plan. This would provide the HCE with an ability to defer 8% of pay. As you can see, if there were only one plan, the HCE would be limited to 2%. In this example, in each plan, the HCE's compensation and deferrals are aggregated and tested against the NHCE's percentages. In this manner, the HCE would be limited to 2%, in the aggregate, in order for all of the plans to pass. You can still get a little (and I really mean the word "little" here) mileage by having multiple plans. For example, if you know that there will be a significant group that will always defer a very small percentage of their pay, it is best to have them in a separate plan so that the HCE's would be supported by the NHCE group that, hopefully, defers a larger share of their income. Again, going to the example with 4 NHCE's, if 2 of them were known to be people who would defer 0.50% of pay, and the other two were known to be people who would defer 3% of pay, if you had two plans you could put the HCE in the latter plan and he/she could defer 5%. If there were only one plan, the HCE would be limited to 3.5% of pay. I think you will find that the advantage one gains by multiple plans is more imaginary than real. While it is theoretically possible to construct these examples where the HCE's are better off by having multiple plans, in real life the advantages are few and far between.

Andy, no, I didn't mean to imply that at all. Here is what I said on that issue: "Note that in this particular case, each plan's ratio percentage exceeds 70%. Hence, one never needs to cross the threshold as to whether or not the average benefits test is passed. So therefore one never needs to determine whether the component plans satisfy the reasonable business classification rule. But let's go there anyway. Let's assume that, while the average benefits test is passed, the reasonable business classification test is not. That is, we cherry picked who is in each plan. Now, the 410(B) testing isn't impacted, because each restructured plan passes the ratio percentage test of 70%. " The corresponding statement, which I didn't make, is that if one of the restructured plans failed the 70% test, then the fact that the reasonable business classification test is not satisfied would definitely cause the plan to fail 410(B). If this isn't clear, let me know.

My interpretation of the reg is that you count the number of employees on December 31, 2001 and call that X(0), you count the number of employees on December 30, 2001 and call that x(1).... you count the number of employees on January 1, 1997 and call that x(1825). You use the result of max(x(0)......x(1825)). mike

Thank you, Andy, that is very kind.

I couldn't tell from your original posting whether this was a new plan for 2001 or 2002. If it is a new plan for 2002, then I think you are correct, although I'd have to pull the law to confirm. On the other hand, if it is a new plan for 2001, then you would use the four year lookback.

I'm with Andy on this one. I would love to see the cite or source for saying that the denominators for the rate groups in any of the restructured plans should be based solely on the HCE's and NHCE's in the restructured plan being tested. That is not the way one tests if one has two separate plans (say, one for salaried and one for hourly) and I don't believe that is the way one tests a restructured plan.

It is highly unlikely that you would do anything other than count all hours with the employer when determining who is eligible. However, it _is_ a document issue, so you need to ensure that the document doesn't indicate that only hours while in an eligible class are counted towards the hours requirement. The same issue applies to counting salary. Here, there is no pattern that I've been able to discern. Some plans count all salary, some don't. Weighing in without looking at the plan document seems futile.

Hi, Lynn. Whether by attribution or not, you have 4 HCE's in your plan. If one of them doesn't get a benefit, you have 3 out of 4 benefitting and all of your percentages go down accordingly.

See Q&A T-14 of the 416 regs. The relevant quote is: "The number of employees an employer ... has for the plan year containing the determination date is the greatest number of employees it had during that plan year or any of the four preceding plan years. For purposes of this Q & A, employees include only those individuals who perform services for the employer durint a plan year." So, it ain't that easy! For 2001, in your first case, the number is 60. In your second case, the number is still 60, because you never had more than 60 employed at any one time. But you need to go back and get four prior years' worth of data to determine the maximum number of employees. Fun stuff, huh?

Would anybody mind if I weighed in on this? The inclusion of the 3rd HCE in either of the plans is a red-herring, I think. The HCE isn't getting any benefit, so he isn't counted for 410(B), whether he's technically included or excluded. That is my interpretation of what Richard meant when he said "assigned." The ratio percentage of each plan is (2/4)/(1/3) = 50%/33.3% = 150%. This pretty obviously passes the ratio percentage test, without having to move to the average benefits test. So, both restructured "plans" satisfy 410(B). We then move on to 401(a)(4). When testing plan one, there is one and only one, rate group. That rate group has 1 out of 3 HCE's, so the HCE percentage in that rate group is 33.3%. The question is how many NHCE's are also in that rate group? The only people eligible to be in that rate group are the two NHCE's in this plan. You have one of three scenarios. Either 0, 1 or 2 are in the rate group. Zero will obviously fail. Going to the other end, the ratio percentage if two are in the rate group would be identical to the 410(B) testing percentage, so that must pass. The question is whether having one in the rate group would pass. The percentage calculation would be (1/4) / (1/3) = 25% / 33% = 75%. That will also pass. Let's review. By having separate plans, we essentially are able to assign a single NHCE as support for each HCE. If we chose not to restructure, there would be 2 HCE's in the single plan being tested. Hence there would be one rate group with both HCE's in it. That rate group would need at least 2 NHCE's in it for support. Restructuring serves to reduce the number needed for each HCE to one. So, let's see what can be done with this. Since the restructured plans can be tested on different bases, you can test one on a cross-tested basis, and the other on contributions. Or one on annual and the other on accrued to date. My favorite subject in college was Linear Algebra and this is stuff that is tailor made for techniques that are useful in Linear Algebra - of which "trial and error" is perhaps the best known. Note that in this particular case, each plan's ratio percentage exceeds 70%. Hence, one never needs to cross the threshold as to whether or not the average benefits test is passed. So therefore one never needs to determine whether the component plans satisfy the reasonable business classification rule. But let's go there anyway. Let's assume that, while the average benefits test is passed, the reasonable business classification test is not. That is, we cherry picked who is in each plan. Now, the 410(B) testing isn't impacted, because each restructured plan passes the ratio percentage test of 70%. But, if we were to increase the numbers of each, such that, let's say there were 30, rather than 3 HCE's. And 40, rather than 4 NHCE's, and we exclude 10 HCE's (rather than 1) while including 10 HCE's in each of the two restructured plans, where each restructured plan has 20 NHCE's, we get the same result as far as 410(B) goes. But looking at 401(a)(4) and the rate groups under each plan, we need to know how many NHCE's are needed to support each HCE. Again, let's assume that all 10 HCE's in each plan are identical in age, compensation and testing benefit. So, we have 1 rate group. How many NHCE's are needed to be in that rate group so that the plan will pass 401(a)(4)? If one takes the position that the reasonable classification test has meaning such that if it isn't passed, the ratio percentage must be 70%, then we must find X such that (X/40) / (10/30) = 70%. I think you will find that the number is 9.333, which means 10. Which is exactly where we were before. If one takes the position that the reasonable classification test has no meaning in this context (that is, even if it isn't passed, one can still use the midpoint of the safe and unsafe harbor (or, if lower, the ratio percentage of the plan) - which, by the way is what I believe), then, instead of 70%, the rate group would only need 33.75% of the NHCE's. Thus, our formula changes to finding X such that (X/40) / (10/30) = 33.75% and I think you will find that X is 4.5, which means you need only 5. Clear as mud, huh?

I don't think there is anything precluding you from doing that at all. But, as I said, I think it limits Plan Sponsor flexibility and this particular case seems to really make sense only if everyone in the HCE category would be deferring the "maximum", if given the opportunity.