AndyH

I just read lkshipman's comments and I think they're terrific. Bravo!

Thank you for your comments. Very interesting. In the situation prompting these questions, we're on the plan side, but it's the CFO of the plan sponsor who is getting divorced and asking these questions. There seems to be a shortage of attorneys well versed in these matters. Our legal staff could not see a way to separate the QPSA and QJSA consents (i.e. one but not the other) That is not to say they reject it, just are having trouble accepting the ability to break it apart as a result of the language in the Code. The idea of restoring the participant's benefit is fascinating. One thing I find odd is that the DOL/IRS' model language has optional language for the alternate payee being treated or not treated as the spouse, but not the same option for the participant. I have followed your comments on similar questions on these boards with great interest, so I appreciate the feedback.

1. In a plan with a REA-only death benefit, can a QDRO name the participant as the spouse for purposes of a QPSA benefits of the ex-spouse in a Separate Interest QDRO (i.e. 50 % of accrued benefit as of date x is assigned to ex-spouse, 50% remains benefit of participant)? 2. Can a participant name the (ex-spouse) alternate payee to be treated as the spouse for purposes of the participant's remaining (50%) share for the QPSA only, but not the QJSA, again assuming a Separate Interest QDRO. I ask this wondering how to avoid the need for the participant to ask the spouse 20 years or so from the divorce date to give spousal consent to a form other than a QJSA (with ex-spouse as beneficiary), but to preserve some death benefit in event of premature death while divorced (not re-married). 3. Can a Separate Interest QDRO stipulate that the ex-spouse be treated as the spouse for QPSA and QJSA but only up until such time as the participant remarries?

I need to correct my earlier comments on the amendment. It appears to me that Revenue Procedure 99-23 does deem a 415(e) amendment to be a disqualifying provision. It seems to say that you have until the end of the GUST remedial amendment period to retroactively amend to repeal the 415(e) application, retroactive to the first day of the first year that the repeal was applied in operation. So, the amendment due date appears to be 12/31/2001 for 2000, although it does not have to be repealed at all, or can be repealed effective later provided the plan satisfies the nondiscrimination rules, which could include the 401(a)(4) general test if NHCE's are affected. I am finding conflicting analysis of this, however, so I'd suggest reading RP 99-23 and Notice 99-44 for yourself.

I'm not sure it has to be removed at all. If NHCE's are affected, however, it may make the plan subject to general testing. But, you can't take advantage of the repeal for a particular year unless the plan is amended in that year. You do not have the remedial amendment period to amend this retroactively because it is not a disqualifying provision. [Actually, I better hedge on that. That's my understanding-someone please correct me if that last statement is incorrect. I know it is correct for a DB plan.]

Agree. This has been referred to as "self destructing" language.

Tom, thank you for your comments. I'll start with the 403(B) answer book.

Thank you, MWeddell. Very helpful.

Now the prospect in my earlier post wants the cross tested employer contribution to be in an existing 403(B) plan, or plans. Are the employer contribution non-discrimination testing rules the same for a 403(B) plan that doesn't satisfy the 403(B) safe harbor nondiscrimination rules? Admittedly, I haven't researched this yet. Anyone done this? Are the 403(B) deferrals included in an average benefits test in the same manner as a cross tested DC 401(a)(4) plan with a 401(k) provision if rate groups are not all at 70%? Are the testing aggregation rules the same for commonly controlled organizations if such organizations are deemed to be under common control? I'm not asking about how the controlled groups are established, just if they are deemed controlled, are they subject to aggregation for the average benefit test component (if there is one) in the same manner as a plan tested under 401(a)(4)? The one I'm dealing with has maybe 5 pseudo-related organizations, so I'm trying to figure out the "what if's".

True of course. Also, I think there would be complications to the hours requirement if a short year were to somehow be created. I recall reading that in the case of vesting a short year would result in the same hour computation (e.g. 1000 hours), but for the 12 months ending in the short year. So, what I'm trying to say is that somehow running a short year may require that the 501 hours be deemed satisfied. I'm not 100% certain of this, but I do seem to recall reading that somewhere, and it merits caution.

Why would that be better than freezing the plan and having no contribution for 2001?

You cannot do it retroactively to 2000. That would be an impermissable cutback in benefits. The consequences would be that the contribution would still be owed, but the plan would be disqualified. A decrease in the formula cannot be effective prior to 15 days after the plan has been amended and participants have been formally notified. The 3/15/01 date is probably the date that participants have earned 501 hours and the contribution would be locked in for 2001, so it may have the same result as being retroactive to 1/1/2001, but technically it is not.

Employer with ERISA-covered 403(B) plan wants to set up cross tested profit sharing plan. The 403(B) had both employer money and employee deferrals. Still has 403(B) deferrals. For the average benefits test component of the 401(a)(4) test, do the employee deferrals need to be included? Would the answer be different if the 403(B) plan were non-ERISA employee deferrals only (on the basis that it was sponsored by the employee, not the employer?) POSTSCRIPT: I think I found the answer, which is tht the 403(B) is not aggregated with the 401(a) plan under any circumstances for purposes of testing the 401(a) plan.

I think that is an accurate statement. Good luck.

Sorry for the delay in responding. It's that time of year. First, a REA type death benefit is very cheap. We're talking about maybe 40% on average of the pension value in the event of death. I'd first suggest that you have the actuary handling the case cost out a "similar" (REA-like) death benefit. For example, the beneficiary of an unmarried participant gets 45% of the present value of the participant's accrued benefit. You'll be surprised by how low the cost is. This assumes that the actuary has software that can handle this calculation. If not, they can provide an estimate. You will be surprised how low the cost is. The conditions you've added will save a small percent of a small amount. I wouldn't bother. As to insurance, the benefit you're talking about is small. Any benefit which can be provided by insurance can be self insured. Outsourcing it to an insurance company will add the element of administrative costs and profits. This assumes a perfect insurer (I'm being kind to them). I'm not an insurance expert, but common sense tells me that the dollars you are talking about could be easily handled from self insuring, thus avoiding margin, profits, commissions. I would only outsource this if the trust could not easily absorb a couple of deaths. But, you are talking about small dollars. P.S. costs are amounts participant must include in their taxable income as a result of the "economic benefit" they derive from insured death benefits. The easiest way to think of this is that the participants pay taxes on the insurance premiums. The actual figures are different. I recently had a client with group term insurance whose P.S. 58 costs were actually 3 times the premium! The figures come from IRS tables. In place of the table, you can use figures from the insurer which is based upon a 1 year term product which is commercially available (or some technical language like that-I don't have it with me). This definition isn't technically precise, but the point is that participants need to pay taxes on insured death benefits. If you provide a death benefit which is a function of the accrued benefit, I don't think that applies. Surely someone will point out that there is a tax benefit derived from paying the PS 58 costs, in that the death benefit is then free from taxation. You can look into this, but my advice remains that it is not worth it. (I'm not certain where the line is drawn for when PS 58 costs apply in a self insured environment, maybe a board member can help me with this). Hope this helps for now. Think in terms of profits for insurers and check out PS 58 cost tables as issued by the IRS. I think that whoever told you term insurance is the way to go is either not experienced in such matters or has an "economic benefit" derived from such advice.

I would also sign up for the course in S-Corp 101, if there is someone out there willing to teach it here.

Sure it could be more generous. Why would you think not? Many plans covered by ERISA have much more generous death benefits. The main limit is that the death benefit be incidental to the retirement benefit. The limit would be multiples of the REA death benefit. Since it's a non-electing Church plan, it may not even be subject to the incidental death benefit limits (off hand, I'm not sure). I don't, nor I trust would many others in this field agree with your suggestion that the client would find term life insurance more cost effective than providing that benefit through the plan, for a number of reasons, one being that the insurance company needs to make money, and another that it would subject the participant to PS-58 costs.

Wouldn't he be able to amend for 2001 if the plan is standardized (by that I mean get contribution unless either employed on last day or worked 501 hours), unless of course he worked 501 hours in 2001 before the amendment?

I don't see any problem with this. I don't see any common control or affiliation issues. Unless I'm missing something, sounds like a go.

I don't think so. Why would you suspect that?

Yes, I agree with each of these comments. This reinforces the problematic nature of this. I have to believe that others have tackled this before.

Anybody have a cite for the due date for a profit sharing plan sponsored by a non-profit organization?

Thank you both for the comments. We do know that you can recognize a post-valuation change for ERISA, and are required to (provided it is significant) for FAS. The question is, if you decide to recognize the amendment for ERISA, how are you supposed to adjust the liabilities, based upon some theoretical assumption (with no real basis), or should you use actual experience, which you don't in theory know as of the valuation date. We're looking for the technically correct or proper answer, trying to do it "by the book".

Tom, thanks much.

Tom, sorry but I'm a little late re-focusing on this. Can I try to make sure I understand, since in my world, the ADP test is done first, by someone else, then the 401(a)(4) test later, which I get involved in. But, now, the K testing people are asking me if they should consider their testing method in light of step 2, the 401(a)(4) test. I think what you're saying is that whether or not the ADP or ACP test was done using permissive disaggregation (separately testing excludables) has no impact on the method used to test for 401(a)(4) on the profit sharing component. In other words, this does not mandate component plan testing. It would, however, require that the same methodology be used for the 410(B) test of the K plan component. Similarly, excluding "otherwise excludable" NHCE employees from the ADP test has no effect on the 401(a)(4) test of a 401(k)/profit sharing plan subject to general testing. Do I have that right? Anybody disagree? Thanks.