MWeddell

No, there is no exception to the multiple use limit for the reason you describe. The easiest solution is probably to switch to the current year method for 2000. It's widely expected that the SBJPA remedial amendment period will be extended to the end of the 2001 plan year. If so, this means that one may switch back from current year to prior year in 2001. All one has to do is amend the document to record this. Note that switching back to prior year becomes much less advantageous if one is using QNECs, switching QMACs from the (m) test to the (k) test, or switching elective deferrals from the (k) test to the (m). Tread more carefully if this describes your situation. Obviously if the client expects to vary the match in future years, the current year testing method will be more suitable to avoid a repeat of your current situation.

Thanks, JWBrown. That's what I was trying to say.

Perhaps the citation you want is Treas. Reg. 1.401(k)-1(B)(2)(i)(flush language after subparagraph (B)). It basically says one doesn't fail the ADP test when there are only highly compensated employees eligible. I'm sure the 401(m) regulations contain a parallel provision for the ACP test.

If the plans pass coverage testing separately, then the ADP/ACP tests are also run separately. (There's a special rule that says HCE contributions in both plans count in either plan's test, so testing separately has some problems if HCEs switch plans partway through the plan year.) Because ADP/ACP tests are run based only on employees eligible for the plan, employees eligible for the second plan won't impact the first plan's test. Assuming the plan documents give you this flexibility, you can aggregate the plans (i.e. treat them as if they were a single combined plan) for testing purposes and run one set of ADP/ACP tests. Frequently, one has benefits, rights, and features that differ between the two plans, so one must still pass the 1.401(a)(4)-4 test for each of the two employee groups, which through a series of cross-references leads one to the nondiscriminatory classification test in the 410(B) regulations.

If the plan is drafted to clearly match only on a payroll period basis, then the true-up isn't required. Beth's original post stated that the plan's match was based on annual compensation, so we've proceeded based on that assumption, which requires a true-up calculation. One can true up every payroll period or do it at the end of the year. The same amount of contribution dollars will be computed in either case. I agree with beth that there is a danger of creating a discriminatory benefit, right, or feature if one trues up every payroll period for those who hit the $10,500 limit but waits until after the end of the plan year to true up for everyone else. It's an aggressive position, but not clearly a problem.

Yes, one can amend the plan to reduce the matching formula, but it typically can't be done retroactively.

Yes, I am aware of a few employers who compute the matching contribution each payroll period on a year-to-date basis, in essence doing a true-up every payroll period. There is no need to estimate future contributions this way and no risk that one has overpaid match (assuming that the calculation is programmed correctly).

There's nothing in the regulations allowing you to reverse the corrective distributions. However, I think you'll have to correct the distributions using the EPCRS because now that the ADP test is run correctly with the correct HCE determinations, there is no distributable event allowing for the distributions to HCEs.

In general, collectively bargained employees are not subject to ACP testing. I suppose a poorly drafted plan document could subject the union employees to ACP testing even though regulations don't require it. If you want to find other exceptions, the cross-references in the regulation you cited will eventually lead to Treas. Reg. 1.410(B)-6©(2), where you'll find other exceptions. For example, the union could have professional employees or retirement benefits were not the subject of good faith bargaining. [This message has been edited by MWeddell (edited 05-17-2000).]

I've never used the recharacterization method and of course it does no good at all if the problem is a multiple use limit violation, not just an ADP test failure. I've discussed it with clients when a collective bargaining unit (with a few employees earning enough to be classified as HCEs) fails the ADP test. Because those plans aren't subject to ACP testing, recharacterization doesn't lead to other testing challenges.

Chances are there is communication gap somewhere, between the other firm and the client or between the client and you. Of course, it's possible that the other firm is giving out the wrong answer too! Code Section 4979(a) clearly puts the 10% excise tax on the excess contributions and excess aggregate contributions, not on any investment gains or losses. This is the opposite of what the client says another consultant is claiming.

The change would be transparent to participants because it could continue to be communicated as a 10% MP contribution with the current vesting schedule even though the plan document would separate it into a 3% QNEC and a 7% contribution with the odd vesting schedule mentioned in my earlier post. I don't think it's much of a hassle considering one avoids the ADP/ACP tests and no extra cost to the plan sponsor other than some small implementation costs.

In general, the answer is no. Deferrals must be made from compensation received as an employee. While the IRS guidance could be more clear about it, in general allowing deferrals from any pay that continues after the employee's termination date should not be permitted.

Assuming of course that you look at the plans carefully to check eligibility, definition of compensation used for the MP plan contribution, etc., I like this idea. If the plan currently has a 20% vesting schedule the first year, then you'd only need to increase it to 30%, not 34% unless I'm missing something. I do think that you'd have to change the plan document even though the change would be invisible to participants. You'd have to break it apart into two contributions, a 3% contribution that is fully vested and a 7% contribution with a vested schedule of 0% after one year 14.29% after two years 42.86% after three years 71.43% after four years 100% after five years I would tend to want the recordkeeper to track these dollars separately even though they'd be lumped together on participant statements, but I guess that's not legally required if the plan document doesn't mandate that. Incidentally, I've done this sort of thing with the traditional 401(k) testing rules to convert 1/5th of an employer nonmatching contribution with a 1-5 year graded vesting schedule and a one year eligibility period into a QNEC that substantially helps the 401(k) test.

Often loans may be available on a more favorable basis from a salary-deferral only 403(B) plan. That type of plan is unlikely to be subject to ERISA and hence the DOL loan regulations don't apply. For example, a participant with a $10,000 balance might be able to borrow the whole $10,000 in a non-ERISA plan but probably will be restricted to only $5,000 in an ERISA plan.

If you have a copy of the IRS regulations and like to read primary material, you'll find the multple use limit in Treas. Reg. 1.401(m)-2. Many secondary sources will summarize this as well.

For 2000, use the prior year NHCE percentage computed in the manner that was used for the 1999 test. Because the 1999 test was run excluding those under age 21 and under 1 year of service, one uses that figure (after excluding those employees) in the 2000 test.

Using the pre-SBJPA method of computing refunds is not allowed for plan years beginning after 12/31/1996, regardless of whether the document has been updated.

I don't think one can quote only one sentence from the IRS response. While a portion of the IRS response does seem to say that reemployment negates the distribution event, another portion seems to imply that it depends on whether the participant has had a reasonable time frame in which to elect to take a distribution prior to reemployment and yet another portion of the IRS response seems to say that the issue is ambiguous and is within the Plan Administrator's power to interpret the plan document.

Here's the text of Q&A 44 from the 2000 Enrolled Actuaries Meeting Questions to IRS/Treasury and Summary of Their Responses (the "gray book"). It's a pretty muddy answer and not an official cite, but may help anyone still interested in the topic raised by this thread: "A 401(k) participant incurs a bona fide separation from service. Two years later she returns to employment with the sponsor before having elected a termination distribution of her 401(k) account. Can the plan allow her to take a distribution of her pre-separation balance before she again separates, attains age 59 1/2 or encounters a "hardship"? If the plan is silent, may she demand a right to withdraw those amounts? RESPONSE A distribution may be made once a bonafide triggering event has occurred. However, if the person does not elect the distribution within a reasonable time frame, it is interpreted that the person chose to leave the value of her account on deposit. Once reemployed, the individual would need a second triggering event in order to be able to receive a distribution. The plan could provide that the distribution is not available if requested while the plan sponsor currently employs the individual. If the plan is not clear, the Plan Administrator would be responsible for the interpretation of the plan."

There are no restrictions on doing this. Whether there is adequate security is measured at the time the loan is taken, not afterward. Furthermore, the security is the half of the account from which the participant borrowed (which is now invested in a promissory note), not the remaining half of the account.

The 10% excise tax is on just the contributions being refunded, not any associated investment gains or losses. The Form 5330 is very simple to complete, more like a transmittal form than anything else.

That's right, the refunded amounts over the 415 limits are not subject to the early distribution 10% excise tax under IRC 72(t). The cite is Rev. Proc. 92-93. There might be a more recent authority, but not one I could quickly find.

As the previous post said, he's a HCE for 1999. Whether he's included in the test depends on whether he was eligible to make elective deferrals into the plan at any point during 1999. While the IRS frowns on allowing employees to defer from severance pay made after the employee has ceased to provide services, if your plan document clearly allows the employee to make elective deferrals and has a determination letter, it might push you toward including him in the test.

I'd keep digging for possible answers in the prototype plan document and the adoption agreement, instead of the SPD.