Different levels of matching contributions

[Guest dlm]

An controlled group of companies is part of a 401k plan. Each member of the controlled group would like to put in a discretionary match. The match per controlled group may be different. For example one may match 50% up to 4% deferred and the other may match 25% up to 5% deferred.

If 401(m) passes on a whole, then would 401(a)(4) testing (rate groups etc) be needed?

Are there any reg cites that give examples?

[Tom Poje]

I don't know about reg cite examples, but there are a few examples.

(You do not do 401(a)(4) testing, its 410(B), though the process is close to rate groups, its just not on an individual HCE test.)

The concept, or whole idea involved is similar to the idea of forfeiting match $ attributed to corrective distributions involving return of deferrals. For example, suppose my match for one group was greater than the match for another, and this group consisted of HCEs only. That would fail the 'rate of matching' contributions test. Yes, I know, how do you pass ACP test and yet fail the rate test, but it is possible.

Remember, not all HCEs might have deferred, but they are still in the higher rate band.

anyway, examples:

see ERISA Outline Book by Sal Tripodi

1998 edition is p. 11.191 Nonuniform match formula (under the section on 401(k) and 401(m) testing.

or, of all places, try The 403(B) answer Book by Panel publishers. Q 12-17 (second edition)

I believe I posted a similar example awhile back as well under this section.

[Guest Harry O]

Each rate of matching contribution is treated as a separate benefit, right or feature. See Reg. 1.401(a)(4)-4. Thus, as long as each rate of matching contribution is available to a group of employees satisfying section 410(B), you should be o.k. (Of course, you still need to do ACP testing . . .)

[MWeddell]

The 401(m) portion of the plan (match and any after-tax contributions) is disaggregated for 410(B) purposes, so you'll have to test it separately. To the extent that the rate of match available to employees differs from one group to another (as in your situation), then you must do benefits, rights, and features testing under 1.401(a)(4)-4 for each rate of match:

1) You're testing the matching rate that was available to employees, so you look at all eligible employees, not just those who make sufficient contributions to actually receive the maximum match.

2) One can't cross-test the match on a benefits or age-weighted basis. Also, there's no imputing permitted disparity. This makes the test fairly easy to do but harder to pass often than if you were testing nonmatching contributions.

3) Good news (if you track through all of the cross references in the regulation) is that you only need to pass the nondiscriminatory classification test (which requires a ratio percentage between 20.75% and 50% depending on the employee population), not the regular 70% ratio percentage test and not the average benefit percentage test.

All of this is in addition to your 401(m) test. Hope that helps. Good luck.

[Guest Norma Snapp]

Do you know of any non-standardized prototype document that will handle different levels of matching contributions for members of the controlled group?

[Guest Beth N]

I'd like to resurrect this topic to see if anyone has any additional thoughts after the passage of time. It seems you made three points. First, that 401(a)(4) will apply to each match separately. Second, that 410(B) will apply to each match separately. Third, that the ACP test will apply to all of the matching contributions combined and applied against the universe of employees.

Tom Poje (if you're still around) - I have no idea what you're talking about. Can you elaborate?

[Richard Anderson]

Beth N,

Starting with your 3 points, bottom up.

3. You are correct, the ACP is done exactly the same as if the match was uniform.

2. The 410(B) coverage test is also done exactly the same as if the match was uniform. Any participant eligible to receive a match is considered to be benefiting. For example if the match rate is 100% for division A and 50% for division B, a participant is considered benefiting if he or she is eligible for either the 50% or 100% match. You do not test each match separately for coverage.

1. Each rate of match must be tested to see if it is available on a nondiscriminatory basis. This is done similar to rate group testing using the general test to satisfy the 401(a)(4) nondiscrimination requirement for the amount of contributions or benefits. The difference is that instead of having a rate group for each HCE, there is a group for each rate of match. Each different rate of match must have an availibility ratio (% of NHCEs benefiting divided by % of HCEs benefiting) which meets or exceeds the safe harbor percentage for the plan.

Tom Poje points out where to find examples of how to do the required testing.

[Guest]

my brain hurts, but here goes anyway...

your situation is not that much different than a cross tested allocation by classes. e.g. division 1, division 2, etc.

allocation will be made to each division, comp to comp.

this is fine as long as you designate the $ amount to be allocated to each group.

I would conclude you have the same scenario, as long as you designate the $ amount to be allocated to match and profit sharing.

I don't know, it was an IRS directive issued sometime in 1998.