Blinky the 3-eyed Fish

You already said that the prior methodology limited the UAL to 0, so you have to keep that. So your expected and actual UAL are both 0, thus no gain or loss. True, you do now have a balance equation that is "out of balance", but that is okay in this situation. What to watch for is the time in the future when you do have an UAL > 0. At that time you will have to set the experience gain/loss base to an amount that balances.

I assume you have a credit balance and no amortization bases?

Limiting the UAL to zero or not is a component of the funding method. You must continue the same methodology unless you apply for approval. What was used in the past?

Mike, I would never say you are off base. You certainly have much more experience and knowledge than me, so I always consider your opinion to be valuable. But how about this logic to argue against what you are saying? Say as an example we have the same situation as first presented, except the regular nonelective contribution is not cross-tested, but rather allocated pro rata on compensation, a safe harbor formula. Agree that the determination of 410(B) and 401(a)(4) would consist of the same people in the testing? Agree that for the formula to be considered a safe harbor you would consider those receiving a lesser contribution (i.e. only a top heavy minimum or a safe harbor nonelective contribution) as not benefiting? So, say we have many people who have only met the 1 YOS requirement. Can you see by including in the test all those who receive some form of nonelective contribution, you could have a situation where a safe harbor plan will be forced into general testing? There couldn't be a safe harbor 401(k) plan with a 2 YOS requirement for the regular profit sharing piece that would be safe from a potential general test. I am betting this scenario is allowable to be chosen in a standardized prototype. So, the above situation where an otherwise safe harbor plan design is being forced into a non-safe harbor design by the addition of the safe harbor 401(k) nonelective does not make sense to me. That is why I answered the way I did in the first place and am not convinced it is incorrect. Thoughts?

Yes, 410(B). You won't run afoul of 410(B) because you are manditorily disaggregating the 401(k) portion and the nonelective portion of the plan.

Yes. No. Just keep in mind that if the plan is top heavy, you will have to give 3% to those new hires.

Assuming you are going to amend the plan to say this, I don't see any problems with the idea.

Tom, to avoid inciting any pyromaniacs out there, I prefer dad gummit.

Ah, while not specifically stated, I see the idea was to avoid 404(a)(7). One other thought to make them aware of is that if one of the doctors leaves, because of his HCE status, his distribution may very well not be able to be made unless the plan is appropriately funded.

Mike, your question got me thinking a bit. Do you feel in determining the rate groups you need to: 1) include all that met the 1 YOS requirement; 2) just include all that met the 2 YOS requirement; or 3) you have your choice between 1) and 2)?

If you are including the only NHCE in the DB plan, why does it need to be aggregated with the DC plans to pass 410(B)?

First, I will assume that the safe harbor nonelective contribution is being used and not the match. That being said, yes, those with only 1 YOS do need to receive the gateway. Now when you cross-test the profit sharing allocation, you are only going to consider those that met the 2 YOS requirements for eligibility when determining the rate groups. The additional gateway for those with less that 2 YOS will only benefit you in the average benefits test.

I had always understood a predecessor plan to be a plan (any qualified plan that is) that terminated within 5 years before or after the newer plan was establised. I have never seen a similar interpretation of the "another such plan" language Jaemmons desribes and am curious what others think. Playing devil's advocate that the IRS would not consider the difference in plan types, there could be issues if you exclude years of service prior to the DB effective date IF the DC plan were to terminate within 5 years of the DB effective date. At that point the DC plan would become a predecessor plan. There might have been distributions that were done excluding some vesting service that should not have been excluded and some of the current participants might receive an increase in vesting years of service. But if you are sure the DC plan will not terminate within 5 years, then there is no issue that you can exclude YOS for vesting prior to the effective date.

It definitely is not a safe harbor and must be general tested by aggregating the plans. So the real issue is 401(a)(26) and I am not sure that it violates anything. None of my plans limit the offset amount. I suppose the argument could be made that there is a variation of the percentage of the profit sharing contribution being used for the offset. For those that receive a higher contribution, only a fraction of the contribution is effectively used versus those receiving a lower contribution. Honestly, I don't know if there is a problem with that being not considered "uniform and reasonable."

I have followed the .5%, sans other guidance. In one of my typical plans, the owner gets x% and the rest get .5%. In the end all but the owner's benefits are completely offset, but that is still fine according to 1.401(a)(26)-5(2)(i).

My understanding of the reasonable and uniform basis is it relates to whom is benefiting, not the level of benefits. We have plans with determination letters where the PS plan is cross-tested and the DB plan has a formula that benefits the owner at a higher percentage.

Why make extra work for yourself either way? It is easiest for me to just pay out the people under $200, so I do that. It sounds like in your recordkeeper's situation the opposite is easiest.

Brian is correct that there was a change in the definition of participant for premium payment purposes effective for 2001 premium payment years. The instructions to the forms clearly spell this out in the definition of participant or the what's new section of the 2001 forms. Again note that this change is only effective for determining the amount of the premium and has no bearing on determining coverage. I have a few plans with 40-50 lives, but only the owner has a net DB benefit. So far they've had to pay $19 per year.

Off the top of my head, it's 416(g)(2)(A)(i). Just kidding, but I did have it handy.

The top paid election only applies to determining HCE's via the compensation criteria. How about a little example. A - 50% owner, comp 50,000 B - 50% owner, comp 150,000 C,D,E,F - comp each 120,000 G,H - comp each 100,000 So, using your determination of 5 people in the top paid group, your HCE's would be as follows: A - owner B - owner & compensation C,D,E,F - compensation G,H - NHCE's because of the top paid election.

Okay, thanks to all. As far as the complexity, I haven't noticed it being a problem. To me the complexity would be no different than if I was running and end-of-the-year valuation.

Mike, I am not following. It's a NHCE that exceeded a plan imposed limit. I have always understood that exceeding a plan imposed limit is similar to a 402(g) violation in that the excess is distributed with earnings. I am unaware of a correction method that would allow this "excess" to stay in the plan. That's is why when going back to the original question, I would again liken it to a 402(g) violation and run the testing using a rate of 15% for the particular NHCE.

My opinion is that the statutorily excludables are not included in the ABT. Obviously, others have different opinions.

You might have to look at it two ways. If your plan has been amended for EGTRRA prior to those in 2002 accruing the right to a top heavy contribution (most likely by 12/31/02), then you will do the top heavy testing under the new rules. The new rules will consider only the 1-yr lookback period from 12/31/01 in your case. If your plan was not amended by 12/31/02 and your determination above proves the plan to not be top heavy, you will also have to run a test under the old rules and use the 5-yr lookback period. This is because EGTRRA provides no 411(d)(6) cutback relief.

It looks to be the case. An additional fault with this language is that it references non-keys are to receive the additional allocation, when it should reference nonhighlies.