Mike Preston

Correct on all counts, including the "it isn't backloading" one.

I suppose this is a plan where each participant has their own investment account and the deposits you mentioned were made directly into those accounts. That is, there is *not* a pooled investment account, right? If so, about the only thing that might be doable is to redirect some of the HCE contributions to NHCE's. My guess is that the IRS may not like this approach, so you would want to confirm with ERISA counsel before going down this path. One thing to layer on top of this is that an employer/plan sponsor, when dealing with a plan where everybody is in individual allocation groups is, according to the IRS, supposed to make a contemporaneous writing with every contribution saying who it belongs to. Was that done?

Yes. Yes.

I recommend "Password Safe", an open source free program that fits "just right" for things like this. You can set up a master password that is a strong, yet easy to type, and have complete confidence that the information stored is protected (of course it is encrypted). I have nearly 300 entries in mine and it covers everything from logons at various websites (link benefitslink) to social security numbers of family members.

One other cautionary note is that in case you end up using zero as the percentage for a specific participant, the 410(b) test must use 70%, as it has the effect of excluding a participant by name.

In most cases, employees are blissfully unaware of the Social Security taxes (I have a hard time believing they are excise taxes, because excise taxes are not deductible). So, the "savings" to the employee would only by 1/2 of what you describe, and if they are under 59 and 1/2, by pulling the money out they would be in the hole. But your concept is absolutely correct in theory and many employers who have employees with discretionary income do exactly what you describe. There is a human element, though, that tends to upset this particular apple cart. Over time, an employee is likely to compare their NET compensation to a job offer. When that happens they will frequently scratch their head and wonder whether they have been woefully underpaid. Of course, they haven't, because they should look to the combination of their take home pay *and* their retirement/profit sharing. But they, at the least, have to be reminded frequently if the employer has implemented a plan that takes that into account. Also, keep in mind that there is an employer limit of 25% of pay that would also preclude taking too much advantage of this concept.

Possibly, but unlikely.

Which is SoCal's way of saying: 404a7 doesn't apply any more, so you should be fine.

I agree, disagreement is OK! I think we both understand the issues. The ultimate question is whether the IRS or DOL could/would claim a disguised service requirement in excess of that which is statutorily/regulatorily allowed.

I suspect that the best and most accurate answer to your question will leave you with more questions than you had before you posted. Suffice it to say that the rules on who can be excluded, and when they can be excluded, are the subject of many hours of continuing education for those who deal with this stuff on a daily basis. But, to answer your question, the rules are contained in the Internal Revenue Code (assuming your employer is not a governmental agency or religious organization) under sections 410(b) and 401(a)(26). Each 412(i) plan has to satisfy both of those sections. And while the rules are labyrinthian, somebody familiar with them should be able to explain, in a few words, how a plan satisfies both definitions. However, the real question is what would you do with the answer? Even if you could substantiate a claim that the plan in question failed one of the requirements, the plan sponsor would not be required by law to fix the failure. Fixing the failure is completely optional. There are tax considerations and cost considerations that go into the decision. Hope this helps a bit.

You could post questions here, but don't include identifying information. You could try: http://www.actuary.org/palprogram.asp Good luck. BTW, it is not uncommon to have a 412(i) program roll through a series of advisors. This has nothing to do with the quality of the service firms, but everything to do with the relationships between the advisors. For example, if a 412(i) plan sponsor decides it is necessary to change legal advisors, that legal advisor might have an ongoing relationship with a different administrative firm. And, if you do any research about 412(i) plans, you will easily find that the IRS has audited many of them in the last few years and many clients have therefore changed legal advisors along the way.

Then we will have to disagree. I think the funnel you are describing does not need to be all inclusive at the front end (you describe a situation which is violative where all employees first must be interns or probationary). I think all that is required is that the funnel be all inclusive at the back end (I describe a situation where all, or most, interns become employees). Read the example in the regs again. Clearly they are talking about part-time employees and there is no requirement that all employees first hired be hired as part-time employees. At least that is how I see it.

That is the difficulty. Each "side" of the policy has its own set of expenses, fees and costs. Separating them is almost impossible. And the very rough justice approach of the IRS saying that 50% of a whole life policy's premium is to be treated as the cost of insurance will rarely, if ever, correspond to what one would come up with if a building block approach was attempted. You do realize that if you want to use a building block approach you are looking to essentially re-write 40 or so years of IRS practice? I'm not saying that what you want to do is impossible, just that it is revolutionary. That is not necessarily a bad thing, either, of course. All new theories, whether revolutionary or evolutionary, have to start somewhere. But if you are going to go down that path, you need to recognize that the IRS is not likely to go gently into the good night on this particular issue, given that the logical conclusion of your approach is to (greatly) increase the amount of insurance that could be considered incidental in the case of most participants. I suppose as the age of the participant increases, the opposite might be true.

I don't disagree. Did my post imply that I do? Sorry. Maybe I should have made it clear that the interns in question may involve repeat interns, such that if they weren't "interns" they would have, at some point, entered the plan. That is the concern, where somebody, classified as an intern, earns enough hours of service to become a participant, and is held out due to being in the excluded class. I think we both agree that if the pattern for an employer is to provide internships such that individuals (such as college freshmen) go through four years of said internships and that virtually every one of the interns becomes a regular employee when the internship ends that you have a pretty obvious 410(a) violation.

I think you would have a hard time becoming an insurance commissioner. Either that, or insurance companies would have a hard time meeting long term liabilities. So, let me ask you a question: where does the concept of "reserves" fit into your equation?

I would go the other direction. How would you treat an employee who terminated in the sixth month of the plan year and was paid a lump sum before 12/31?

Thanks for the links, masteff. Very informative. As to the above, I agree with the slippery slope being what kind of crossover is allowed before it becomes a disguised service requirement in excess of 1 yr. Let's say the firm historically provides for 4 internships and typically hires 1 of those 4 as permanent employees. Certainly that would be a potential problem. But let's say that in most years, the 4 internships run their course and those interns do not become regular employees of the company. After 4 years or so of this program, in the fifth year, the company hires one of the interns. I would think this would not be a problem. At issue is where one draws the line. I'd suggest conservatively.

Probably? Seems like precisely to me. I'm not familiar with rules regarding interns. UNPAID interns sounds like a violation of some law or another.

Huh? Did you mean to say "shouldn't"? If so, why not? Isn't the case you have defined precisely what the example is meant to preclude?

The documentation is a series of Revenue Rulings (Rev. Rul. 60-83, Rev. Rul. 66-143, Rev. Rul. 74-307, Rev. Rul. 83-53, Rev. Rul. 85-15 and there are others referenced within the bodies of those rulings), which taken together state that the old principle of allowing 50% of the contribution to purchase whole life and 25% of the cost to purchase term or variable with respect to profit sharing plans is extended to db plans only if you calculate said 50% or 25% with reference to the ILP cost of the naked retirement benefit (not with reference to the actual contribution to the plan). Proof of this is in the LRM's at http://www.irs.gov/pub/irs-tege/db_lrm.pdf. See page 125. Blinky did a calculation here: http://benefitslink.com/boards/index.php?s...st&p=107816

Yes and yes, providing it is not a disguised 410(a) violation. See the regs for an example whereby part timers are excluded by class and the result is a 410(a) violation.

Yeah, you should read the next sentence in the reg.

This is probably going to sound basic, but when you go back to them, make sure they spell out the entire year when they cite something, rather than just the actual year they describe it with. For example, above, your "2007" can be either the plan year beginning 7/1/2007 and ending 6/30/2008 or the calendar year 2007 or maybe 7/1/2006 through 6/30/2007. I know, I know. But it never hurts to be safe.

I'm still not sure the extent to which these "crossover" deferrals are being used. In general, though, deferrals made in one plan year are not allowed to be carried over to the next plan year. There is a possibility, however, that your payroll periods don't match your plan years. In which case the first payroll that falls within a given plan year will involve compensation, and deferrals, that could, in some circumstances, be attributed to the prior year. Many plans, however, determine their plan year deferrals on the basis of the payroll date and would thus automatically attribute the entire deferral from that first payroll to the subsequent plan year. Is that what they are doing?

What would you propose? It seems pretty clear to me that unless there was some sort of abuse with respect to the fact that the monies were paid out of the plan (which is a bit hard to imagine) the "correction" is to tell the IRS that too much was paid out.