Mike Preston

Andy, check out 1.404(a)-14©. It is the cite that lays out the 3 alternatives and confirms that a change requires IRS approval. But keep in mind that this is the determination of maximum deductible, not the determination of minimum funding. It is possible that the plans you have seen use inconsistent treatment have been taking deductions which were not based on the absolute maximum, or were taking advantage of 1.404(a)-14(e)(1), dealing with "includible contributions".

Confusion? Why not just go with 2/1 through 1/31? Of course, since this is post 12/31/01 there are no choices, are there?

General rule: don't argue with plan counsel. It does not nourish the soul.

pax, does it matter? Salary is individually determined. The basis for that determination needn't be applied to all employees in the same manner.

Whether the employer chooses to "blame" the salary decrease on the top-heavy contribution or not, the fact is that an employer is free to set your salary at any level they think is appropriate. Aren't you free to quit? If you don't like the salary level or the combination of salary and benefits, go eslewhere. At a salary of more than $80,000, you would think you would have a better grasp of the dynamics of the workplace. Benefits have always been and always will be a part of total compensation. Unless you have a compensation agreement with your employer which ties the employer's hands in this type of situation, and as long as your employment is characterised, I think, as "at will", then you are free to reject your employer's offer by finding another job. And if you do it early enough, you get the satisfaction of receiving the additional top-heavy contribution for last year, without having worked very long for your employer at a compensation and benefits level that the employer actually expected. How wonderful for you! [Oops, I just noticed that your employer is adjusting your 2002 compensation for a contribution that you will not earn unless you are employed on 12/31/2002, as top-heavy minimums aren't typically paid unless one is employed on the last day of the year. So, you don't even get the satisfaction of sticking your employer with the extra compensation. Sorry, no wonderful for you.] Somebody else will probably respond in kinder and gentler terms.

I think you are confusing apples with oranges. Go the simple route, have the employer execute two certifications, one for each plan, on or before 2/28/2002. The purpose of the language you have cited is solely to determine the filing deadline associated with the executed certification. It doesn't create a certification (and hence an extention) for a plan that doesn't execute a certification. I can understand the confusion, though, because the original language of the "sample" certification doesn't identify the plan. In the more recent version, published in the IRS newsletter, they display a certification that identifies the plan. Now, in retrosepect, upon submission, or on audit, if I run into one of these, I'll most assuredly pull out the original notice and try to get the IRS to buy off on the concept that a single certification applied to all plans of the employer. In advance, though, it is just as simple to avoid the issue.

Look at your GUST document. Does it contain EGTRRA materials? The answer is no. If you amend and restate a plan onto your GUST document, everything that went before it is thrown out the window, except those things you specifically mention in your document that you are keeping. So, I repeat, look in your GUST document? Does it "keep" anything with respect to EGTRRA? The answer is no. So, whoever told you that you needed to adopt the EGTRRA amendment again was correct.

My volume DB plan received its letter on 1/30/2002.

It is basically the same sections already cited. Note that in determining the ABT testing group, one had to have an exception to the exception. Without that second exception, the 401(k), 401(m), ESOP, et. al. are excluded from aggregation under 410(B). QED.

dmj, sorry but I don't agree with the logic. If one is an HCE and is allowed to contribute 15%, they will almost definitely exceed the 402(g) limit, along with the catch up. It is only the NHCE's that will benefit from a provision where the cap is increased. Also, most plans don't even allow for anything other than deferrals and match. I would call that a traditional plan. A plan that allows for after-tax contributions these days is anything but traditional.

Only the IRS could have come up with the "curvy" logic. With that said, what you are looking for is very specifically defined in 1.410(B)-7(e). In that section, you will find the definition of testing group. Testing group includes all plans that COULD be permissively aggregated as defined in 1.410(B)-7(d). -7(d)(2) basically says you use the disaggregation rules of 1.410(B)-7©. -7© says you can't aggregate 401(k) plans or ESOP's (amongst a few others). So, if that is as far as one reads, one will come to the wrong conclusion. At the end of 1.410(B)-7(e) there are three exceptions noted. The third exception is the one that gets us where we want to go. It says that we apply the rules of -d(2) by ignoring the disaggregattion rules of ©(1) [401(k) plans] and ©(2) [ ESOPS]. Does that help?

Maybe I'm the eternal optimist, but let's not beat them up too badly. They might actually be getting some good advice on the cash balance plan. I think the cash balance plan can be designed in such a way that it covers 40% or 50 bodies, but also provides for an offset of certain benefits from the other plans, such that the partners will receive a large majority, if not all, of the cash balance plan benefits, while the others get a larger piece in one of the DC plans. Before anybody jumps on the prior benefit structure bandwagon and what is, or what is not, a meaningful benefit, let me say that I think the IRS has a say in that. Therefore if designed in advance and put to the IRS on an LOD application before one implements the program, I think one might be somewhat pleased with the result. Of course, YMMV. On the PT issue, with 20 partners it is possible that there will be enough assets in the plan that they could get an exemption. I have heard that the DOL will approve up to 25% of the plan's assets if the not insignificant rules with respect to independent fiduciary, etc. are followed. Andy, if you can believe it, I've seen people ask: "By the way, what does BTW stand for?" FWIW Is there an acronym reference somewhere on here or do we have to start giving a url to some outside page? Such as: http://www.slack.net/~thundt/abbrevs.htm

"i'm just as worried as having to pass the acp test now that more people will be spilling over into the after tax contribs." Huh?

I agree with Tom. In answer to your second question, just do the Demo 6 and, if necessary, display the results of the ABT, but don't refer to it as Demo 5. The industry went round and round with the IRS on Demo 5/Demo 6 stuff a while back and the IRS now accepts it this way. At least, they did the last time I submitted one! But you never know whether you will get a reviewer that may not be up to date on this particular issue, so be prepared to respond to a request for clarification or a request for Demo 5. A brief conversation should provide the clarification.

Your amendment to cure the 410(B) failure is governed by 1.401(a)(4)-11(g). It is my understanding that you can craft your amendment any way you want, as long as it meets the rules therein. The most important rule is that it must have economic substance. This generally means that you can't provide benefits solely to those who have no vesting percentage. But you can certainly state that the benefits being provided pursuant to the amendment are specific to named individuals. So the answer to your first question is yes. There are no regs that require the use of total comp in the second situation because there are regs that specifically allow you to use comp from date of entry when testing under the annual method, unless the plan has some strange requirement that limits your options. AndyH has put a great message up which goes through the regs chapter and verse on this issue in another thread. Trust him! Here is a link: http://benefitslink.com/boards/index.php?showtopic=44931

Averaging the rates might help. It might also hurt. Or, better I should say, if it hurts, you just don't do it that way. The gateway is defined in terms of the "allocation" rate. If the DB plans provides for an annuity of 1% of pay per year of service, payable at NRA, you convert that to an equivalent "allocation" in order to test the gateway. You would definitely NOT just use the "accrual rate" of 1% of pay! Think of it as the reverse of what we normally do under cross-testing. But it is still cross-testing.

This is 401(k) plan thread, so the likelihood is that this is a distribution from a 401(k) plan and not an IRA. If that is truly the case, and there is a controversy, the plan can not charge the beneficiaries directly for legal fees incurred.

The simple answer is that only those who otherwise are considered as benefitting under 410(B) need be considered in determining whether the plan satisfies the gateway percentage of 1.401(a)(4)-9(B)(2)(v)(D). But, keep in mind the following: 1) 7.5% is the required "total" of the allocation rates between the DB and the DC plan, not the amount that must be provided solely from the DC plan. 2) If an NHCE is benefitting not only under the DC plan, but also under the DB plan, the plan sponsor can optionally use the average of the NHCEs' DB allocation rates for all NHCEs who participate under the DB plan.

I don't think I used "and" or "or". I would reword it slightly to say that one excludes those in both categories. That is, if you answer true to either of the following questions, that person would be in what I've defined as Group 2: 1) Did that person have less than 1 year of service as of the end of the plan year being tested? 2) Was that person under age 21 as of the end of the plan year being tested? While we are in the process of being picky (which I think is generally a good thing), I should point out that there is a difference between statutory eligibility associated with general 410(B)/401(a)(4) testing and statutory eligibility under these sections of 401(k)(3) and 401(m)(5). In the case of 401(k) and 401(m) one tests for statutory eligibility as of the last day of the plan year without consideration of entry dates. In the case of 410(B)/401(a)(4) one tests for statutory eligibility as of the last day of the plan year with consideration of entry dates. Hence, in a calendar year plan, testing for 2001 would put somebody full-time hired on 7/2/2000 in Group 2 in a 410(B)/401(a)(4) test but they are in Group 1 for a test under 401(k)/401(m).

I'm not sure what you are asking for. Profit sharing plans have individual accounts (whether or not each participant has a physically separate account) and allocation dates. On each plan allocation date, the value of each account in the plan is updated to reflect changes since the last allocation date. In some plans, allocation dates occur every day. In others, once a year. Changes include the allocation of new contributions, the allocation of forfeitures and most importantly to your question, the allocation of gains or losses in the underlying investments of the plan. You haven't given specific dates or amounts, so I'll just use an example. Assume that the plan has one allocation date a year, on December 31, and that the date of death was 18 months ago, or sometime in August of 2000. At that time, if one were to look at the value of the participant's account, they would find that it would be the 12/31/1999 value. Let's assume that value was $25,000. On 12/31/2000, the next plan allocation date, the participant's account would be credited with contributions, if any, for the 2000 year. For argument's sake, let's say there weren't any contributions (or foefeitures) for that year. Then the account would be increased (or decreased) to reflect the gains (or losses) in the plan's investiments since the last allocation date (in this case, 12/31/1999) By way of example, let's say that the underlying investments went down by 10% in that 12 month period. This participant's account value on 12/31/2000 would then be $22,500. We follow the same logic through to the end of 2001. Again, let's say that the underlying investments went down by another 10% in the 12 month period ending 12/31/2001. This participant's account value on 12/31/2001 would then be $20,250. If a distribution is made after 12/31/2001 the plan would be required to pay the widow $20,250. No more. No less. If the allocation dates under the plan are more frequent, the account values change more frequently. But the basic pattern still holds true: if the underlying investments in the plan go down, so does the account value for this participant's widow. The fact that the monies were not paid out until a significant time after the date of death does not change the requirement that the amount to be paid is based on the account value at date of distribution, not the account value at date of death. About the only way that the widow could claim that she should receive more money is if she claimed that the people who were responsible for investing the funds (the fiduciaries) violated the rules on investments (breached their fiduciary duties). In the last 24 months, nearly all pension funds have consistently lost money. To claim that the fiduciaries were in violation of the rules solely because the underlying investments went down in the last couple of years will be a mighty tough argument to make. If you are looking for citations, the "accrued benefit" in a profit sharing plan is defined at Section 3(23) of ERISA and in the Internal Revenue Code under Section 411(a)(7). Sorry to be so negative, but it looks to me like the plan is doing everything right.

Good point. It always helps to remember the context of the original question, doesn't it? In this case, I fell into the trap of assuming the follow-ups were dealing with a 401(k) plan, rather than a 403(B) arrangement. So, the answer is no, they aren't (at the moment). But they are subject to 401(m). So, just change my references from 401(k)(3)(F) to 401(m)(5)©. Same result, though.

I think it is a fairer description to say that you include "highly compensated" but that you don't include "Highly Compensated"! Confusing, I know, but that is the way it is. Or, at least, that is the way it is NOW. You should be aware that you don't "have" to exclude "Highly Compensated" employees, either. The provision in question, 401(k)(3)(F) is permissive, not required, and ends up allowing the plan sponsor three choices as to ADP testing: 1) Normal. Just test everybody. 2) Apply 410(B)(4)(B) the way it existed before 401(k)(3)F) was added. If the employer makes this choice, then you ignore whether someone is a highly compensated employee or not. You test two separate groups. Group 1 consists of those that meet statutory eligibility. Group 2 consists of those that don't. If you happen to have an HCE in Group 2 (which we've already discovered will usually only happen if one is a 5% owner), then you include that HCE in Group 2's test. Of course, the results for Group 2 aren't likely to work out too well, so Congress gave us..... 3) Finally, one can not only apply 410(B)(4)(B), but one can use the special language of 401(k)(3)(F). If the employer makes this choice, you move the HCE's that are in Group 2 above and test them with those in Group 1. In this manner, even if you happen to have a 5% owner, you have the ability to test that person with the people who meet the 21/1 requirement.

You may very well be right. But I haven't seen any case on point, or anything from the IRS that addresses what to do if the recipient plan develops a doubt post-reasonable-acceptance. And, in this case, I believe that Marie indicated that the participant specifically thinks the amount is the correct amount. A bit of a difference from the situation where the sending plan indicates an amount is too much and the participant merely states that they have no reason to doubt that the initial distribution is correct. Any votes for interpleader?

I suppose there are facts that might bear on the issue, too. If the timing was almost immediate, then I can see paying some sort of attention to the sending plan. However, in another recent thread, an individual was notified by the sending plan (in that case it was an IRA, not a receiving plan) that the amount was "too high" more than two years later. By that time, the funds had gone down due to investment selections that didn't turn out too well. In this case, the Plan has the right, under the IRS regs to rely on reasonable statements that the monies were, in fact, rollable. It used to be that conservative plans wouldn't accept rollovers unless the sending plan provided an LOD. Thankfully, that is no longer the case. And that principle still holds. If the plan acted reasonably to accept the monies, the receiving plan should be immune to IRS attack regarding those funds. We don't know, in the case, what "later" means in marie's initial message. We don't know whether the participant's accounts have individual direction. We don't know whether the participant's acount value has declined. I would be very reluctant to send money back to the sending plan without authorization from the plan participant. Maybe, if the plan participant elects a distribution of the rolled over monies, the plan can interplead the funds?

Can you provide specifics of who might be hired in 2001 and be an HCE? I think you will find that unless someone is a 5% owner, they can't be an HCE in 2001 if they weren't employed in 2000. Does that solve your problem?