AndyH

Don't forget that you can test this by carving out the groupies who have less than 1 year of service and have not attained the age of consent.

This year you are correct, except you don't need to do discrimination testing if you have only HCEs. Next year you would have the need to satisfy the gateway, but then you also have to pass the 401(a)(4) general test. Satisfying the gateway only allows you to proceed to do the test on a benefits basis if you choose that method. If you can pass on a contributions basis, you don't need to satisfy the gateway. But if you are maxing out 1 HCE and you have only 1 NHCE then you will not pass on a contributions basis.

The IRS takes the position that it is an impermissable cutback to amend the plan to change the formula once the allocation requirements have been satisfied. So, for example, in a standardized plan, the allocation requirements would be satisfied when somebody works 501 hours. If the plan has no requirements for an allocation, then the amendment must be done before the year starts. If the plan requires employment on the last day, then it can be amended at any time before the end of the year.

Any 414(s) comp. If you pass the 414(s) test by excluding bonus, then that comp will be 414(s) comp. If not, you need to test using 414(s) comp, gross wages being one option.

Tom, I think your second paragraph is true only if the criteria for eligibility for the plan is unreasonable, not the criteria for classification of contribution levels. Agree? For example, assume a controlled group of two companies in separate geographic regions each with 20 employees each. One company has 1 HCE and the other has 3. The company with 3 HCEs has a plan covering employees of that company only who are age 21 with 1 YOS. Then, for allocation purposes, there are two classes, one for people with blond hair, and the other for anyone else. Blond hair people get 5% of pay and everyone else gets 2%. I think the plan can use either ratio percentage (which fails) or average benefits for coverage purposes since the criteria for eligibility is reasonable, even though the criteria for allocations is not. Agree?

Two of you in the DB? 25K and 25K? Now we're getting somewhere.

You are dealing with a life insurance salesman, not an actuary. A plan doesn't have to be a 412(i) plan to be loaded with life insurance. Beware, the numbers do not work. Sounds like you need to get a stronger brand of shark repellant.

You can call them whatever you want that is not ambiguous, by name if you wish. How about "Majority Owners and Spouses of Majority Owners" or "Majority Owners including Majority Owners by attribution", but you don't want to do that if a child becomes employed.

And tell him the board said it's actually two retirement plans he's proposing, one for you and one for him. (Thanks again MWyatt for that line).

Understood. But the only way you can put away more than 40k is if it is actuarially required in a DB. And you need to be older for it to be actuarially required, at least under mainstream approaches. So the math doesn't work.

My opinion is that you are too young for a DB plan. Set up a profit sharing plan and put away $40,000 per year. Look at a DB plan when you reach age 40.

Maybe I'll be breaking new ground, so I'll ask the question carefully: What are market surcharges for administration of cross tested ps/401(k) plans, as opposed to, for example, integrated profit sharing plans, i.e. what might a typical testing fee be? I've seen $250 to $750. I'm not asking what you charge, just what charges you have seen or heard.

Yes, that is good advice, but if you would tell us your age and approximate income we might be able to rule in or out a particular type of plan, i.e. point you in the right direction.

Any 414(s) comp can be used for the 1/3 test. The 5% must be based on 415©(3) comp. I don't know for sure if severance pay is included in either, but I don't think so.

Right on point, David. Thank you. I do wish he had a cite or footnote, but it could not be more on point.

oops. My edition is 2002 (copyright 2001), so that probably explains why I didn't see Blinky's language, nor have I found yours, Tom, yet.

Blinky, I'm not sure it says that, at least not on that page of the 2001 edition. If you're using that edition, it says, "To determine the NHC concentration percentage, divide the number of NHCs by the number of all employees........" It says nothing about participants. Then it says "disregard employees ..... who are excludable under 1.410(B)-6", which I think means excludable under the ABPT. And the Nonhighly Compensated Employee Concentration Percentage is defined in 1.410(B)(4)©(4)(iii) as "the percentage of all the employees of the employer who are nonhighly compensated employees. Employees who are excludable employees for purposes of the average benefit test are not taken into account." I don't think this supports counting only people in the rate group test.

Tom, I just stumbled on my notes about the Concentration Percentage calculation. And it was from Carol's session. She said to include any person eligible for any plan sponsored by the employer or any member of the controlled group, so that would be a much wider group than those in the rate group test. She did worshops 28 and 31. I'm not sure which one I was in. I think it was 28.

Thanks for the clarification, MGB. Now I have to add that this lunacy is another reason why I will never refer to the IRS as "the service".

I don't see a realistic alternative to using the old rules (or forcing a retirement election) until this mess is straightened out.

I read something today that said this was going to be extended until Fall 2003 within the next few days.

Thanks for the comments, Merlin. In my situation, the plan must be general tested anyway, so if avoiding (a)(4) testing is the most compelling reason to amend for EGTRRA retroactively now, I don't see that doing me any good. This plan has about 40 participants, and with the stock market tanking, together with the new (up to current liability) deduction rules, and the application of EGTRRA prospectively, the contribution is through the roof anyway, so I'd prefer to avoid retroactive application of the $200,000 limit as long as I'm not permanently forgoing an opportunity. The formula happens to a creative Ed Burrows design, equal to a percentage multiplied by compensation per year, kind of like an annual accrual plan, so applying EGTRRA retroactively has a huge impact on the contribution.

The sponsor of a general tested plan, actually a DB/DC combo, may wish to defer applying the EGTRRA comp limit increase retroactively for now due to the cost being too much to handle along with the other stuff, such as poor asset performance. What are the issues if in the future if it is amended, for example, in 2004 to apply the $200,000 retroactively? I've read comments about a plan no longer being a safe harbor, but it is not now, so that would not appear to be a problem. There could be an issue of past service credit exceeding 5 years, but I can't see how that could be a problem if the only people over the comp limit are HCEs. Clearly there will be testing issues if the annual method is used, but that is quantifyable. Are there any other pitfalls to be considered?

So, the second plan concept is obviously dependent upon separate testing for 401(a)(4) and 410(B), and presumably satisfying top heavy in a second plan does not somehow require you to aggregate for 401(a)(4) and 410(B). I guess that is right.

I must be the only idiot that can't figure out what SHUSH stands for! Someone please, for the benefit of the class .........