AndyH

Somebody told me that AIG used to provide this product, but will no longer provide it. Does anyone else?

Many thanks again Mike.

How are timely 401(k) refunds treated under the Average Benefits Percentage Test component of the 401(a)(4) test? I have one with refunds under the ADP test and the 402(g) limits. I think the last time I looked this up the answer was "unclear". Opinions? Thanks.

I find the 410(B) aspect of this interesting. Mike, extending what you are saying, which does look right to me, it would seem that if you had a group of people who you didn't want to give the gateway to, and they were 30% or less of the nonexcludable group, then you could exclude them by class and give them their own plan, with a lower contribution rate, therefore bypassing the gateway, a slam dunk if none are HCEs? Is this right?

The 105.93 (I'd use 105.9347) matches our systems for 1983 IAM (Male) at 8.50%.

Someone in my office just asked Jim Holland if this is permitted and he said yes. There was a 1984 ruling or release of some type which is still "operational". So I guess my suspicions were unfounded.

What I'm saying is that a representative of the DOL told me that the DOL's position is that a 990 is not a "Federal income tax return". They argued that a non profit does not file a "Federal income tax return".

Right, but what is that date for a non-profit? A DOL official told me that it is the DOL's position that a non-profit does not file a federal income tax return. The context was a 5500 extension where a non-profit has an extension period beyond the 9.5 months to file a 990 and the 5500 instructions say the 5500 is extended to the due date, including extension, for filing a federal tax return.

Pax, are you sure this went though without a hitch? I get the impression this wasn't your idea. I understand the tax issues, but my situation is a large manufacturing company, which could be in a loss position, so the 20% might be all that is due. And the transfer could be used to reduce current company cost of funding a 401(k) match, so that is a clear savings. I have to wonder if the exclusive benefit rule is a problem. I assumed there were other clearcut problems, but we haven't found any yet. I remember when the termination application used to ask whether or not another db plan would be established within one year. Why did they ask that?

What prevents the sponsor of an overfunded DB plan from terminating a plan, taking a partial reversion, transferring a portion of the surplus to a qualified replacement plan, paying the reduced excise tax, and then re-establishing a DB plan covering the same people and providing comparable benefits? There must be something preventing this (other than a permanency issue), perhaps the exclusive benefit rule. Is that it, or is there something else?

ASPA sent an email to members yesterday saying that the methodology was under reconsideration and that the February rates would be issued in about 2 weeks; thereafter the rates should be released more timely.

Sounds to me like it's being tested on a contibutions basis with permitted disparity being imputed. Might pass (I don't have my cheat sheets with me to be sure). The 5% gateway would not apply because it's not cross tested. p.s. Yes, that's right, depending upon the comp levels, imputing permitted disparity can make the allocation rates for the NHCEs at 23.5% higher than the HCEs at 25%, so it may well work.

How is ASC considered for 401(a)(4) testing, i.e. cross tested plans? Can anyone compare it to the others for this purpose?

At last year's ASPA annual conference (maybe it was 2000?-I don't have my Q&As with me), Jim Holland said a person with no compensation should not be in any nondiscrimination tests (by any I mean a(4),410(B), ADP, ACP). The words I recall were "Not a zero. Not in the test".

We're cross testing a profit sharing plan and it fails badly. Plan has liberal eligibility and some of those "otherwise excludables" are HCEs, so we want to try separately testing otherwise excludables. Rate group testing for the non-excludables (I'll call them "statutory participants") seems straight forward enough. Ignore the excludables as if they don't exist. Right? Is rate group testing needed for the otherwise excludables who are benefitting? I assume so. How are the statutory participants treated? Is the concentration percentage calculated by ignoring them? Are they in the rate groups as non-benefitting, non-excludables? If all rate groups are not at 70%, and we go to average benefits, are both groups lumped together? I think not; I assume you ignore the excludables when testing the statutory participants, but is the reverse also true, i.e. ignore the statutory eligibles when testing the otherwise excludables? Lots of questions, I know. I did a search and couldn't find anything that addressed these specifics. The ERISA Outline Book comments seemed vague to me. Help would be appreciated!

If I may request elaboration on this discussion, and I know this is a 401(k) forum, but .... Assume you want to set up a new, small, DB plan for a principal age 62 who planned to work 5 more years. Would anything prohibit the following: NRA is 67, ERA (providing full vesting) as 65, and the benefit is 50% of pay, payable at age 67. This is as opposed to defining NRA as 65&5P and having the principal's benefit payable at age 67 and everybody else's as payable at age 65, or alternatively defining NRA as 65 and using an assumed retirement age of 67, in which case again the employees would get more than needed if the NRA were 67. Is my question clear?

That certainly would explain the plan I described; and I'd guess that it was what Mike was alluding to. Very interesting. Thank you.

I have to amend my comments; I was partially incorrect. The plan in question does provide for full vesting at age 65. Oddly enough, NRD used to be defined as the first of the month following age 65, but it was amended 1/1/89 to change NRD to SSRA as I noted, but also to stipulate that someone is always vested at age 65. Further, it allows early retirement at age 60 (of the vested accrued benefit), so even though NRD is defined as later than the statutory date, benefits are always vested and always payable (although reduced) at age 65, so I suppose that's why it was approved. And I suspect that somebody made sure that nobody's accrued benefit would be reduced (the formula was radically changed).

Yes, I just double checked it. NRA is not defined. NRD is defined as the first of the month coincident with or next following the later of the fifth anniversary of participation or a table, which is a SSRA table. And I can understand your points about how such a thing could possibly work, (if benefits were always vested at 65+5 and allowed to be paid then) but in this case ERD is age 60 at which time you get your vested accrued benefit, so there isn't full vesting at age 65 if your SSRA is 66 or 67. For what it's worth, it was written by a law firm. I realize there is some uncertainty about SSRA being ok for testing age, but I thought you'd have to vest someone at 65+5 P. That is clearly not true in this plan. Maybe the IRS reviewers just missed this.

Well, now I'll admit to putting my foot in my mouth by remembering that I have a DB plan that defines NRA as SSRA! Yes, it's general tested. And, despite our failure to understand why, it keeps getting favorable determination letters!

But, Mike, I think you'll agree that the latest NRD can be is the earliest of a. the NRD in the plan, or b. the later of age 65 or the 5th anniversary of participation So, it can't be past 65 for somebody participating before age 60. Right?

Wasn't there specifically some transitional relief for the family aggregation change? Of course it would make perfect sense to issue transitional relief here, but it hasn't happened yet, and I don't get the sense that anybody is focusing on this issue. The only place I've seen it mentioned anywhere is Corbel's website, and if you read that Q&A you certainly can't come away with the impression that future relief will be forthcoming.

I have to believe there would be problems under 401(a)(26) or 410(B) with this "indexing". I haven't researched it; just my gut reaction-either it's frozen or it's not. Wouldn't these people at $200K be benefitting under 410(B) and/or 401(a)(26)? Also, what is the justification for giving selective credit for post-freeze changes?

I'd like some more feedback on this question. Here's the language in some documents in question. If this is a ps plan with no last day or 1000 requirement, is it a cutback to amend during this year for EGTRRA's $200K limit, or does this language incorporate 401(a)(17) by reference? "1.6 "COMPENSATION" shall mean the compensation paid to a Participant by the Employer for the Plan Year which is reportable on Form W-2, but exclusive of any program of deferred compensation or additional benefits payable other than in cash. Compensation shall include any amounts deferred under a salary reduction agreement in accordance with Section 4.1 or under a Code Section 125 plan maintained by the Employer. For purposes of determining who is a Highly Compensated Employee, Compensation shall mean compensation as defined in Code Section 414(q)(4). In addition to other applicable limitations set forth in the Plan, and notwithstanding any other provisions of the Plan to the contrary, the annual compensation of each Employee taken into account under the Plan shall not exceed the OBRA '93 annual compensation limit. The OBRA '93 annual compensation limit is $150,000, as adjusted by the Commissioner for increases in the cost-of-living in accordance with Section 401(a)(17)(B) of the Internal Revenue Code. The cost-of-living adjustment in effect for a calendar year applies to any period, not exceeding 12 months, over which compensation is determined (determination period) beginning in such calendar year. If a determination period consists of fewer than 12 months, the OBRA '93 annual compensation limit will be multiplied by a fraction, the numerator of which is the number of months in the determination period, and the denominator of which is 12. Any reference in this Plan to the limitation under Section 401(a)(17) of the Code shall mean the OBRA '93 annual compensation limit set forth in this provision. If compensation for any prior determination period is taken into account in determining an Employee's benefits accruing in the current Plan Year, the compensation for that prior determination period is subject to the OBRA '93 annual compensation limit in effect for that prior determination period. For this purpose, the OBRA '93 annual compensation limit is $150,000. "

Lori F., that's correct. Earnings on the balance already in existence at the start of the measurement date must be backed out, but you can, if you wish, include earnings attributable to contributions during the measurement period. Actuarysmith, yes, the examples you've cited are the most common uses for accrued to date, but any situation in which there have been changes, to compensation or contribution levels, for example, can make this method useful. And, when you're general testing a DB plan I find it often easier to use accrued to date, because the resulting accrual rate can be set to equal the formula in a unit accrual final pay plan, i.e. if the formula is 1.5% of average comp, then you can use 1.5% as your accrual rate if you define testing comp in the same manner as plan comp.