Dual eligibility & 401(a)(4) testimg

[SMB]

401(k) plan with dual eligibility/entry - immediate for 401(k) deferrals, 1 Year of Service/semi-annual entry for employer discretionary (no matching). Employer discretionary contribution is cross-tested class allocation.

I am "assuming" (always dangerous!) that for purposes of the 401(a)(4) non-discrimination testing on the employer discretionary contribution that employees who have not met the 1 Year of Service requirement are excluded from testing, even though they are "participants" by virtue of being eligible to participate in the 401(k) component.

This is my first experience with this particular situation and would certainly appreciate any and all input from those of you who have "been there, done that"!

Thanks!

[Mike Preston]

This design can get thorny, rather quickly. In general, you are correct. However, if the plan is top-heavy, you have to include everybody in the group that gets a top-heavy minimum. If the plan uses the average benefits test to prove that the rate groups satisfy the rules, there are questions at the highest levels as to whether the average benefits test should be run without considering the under21/1 participants.

[Guest merlin]

Mike, could you clarify your answer a bit ? For a t/h plan the non-keys who have

[Mike Preston]

Choices, choices, choices.

410(B) testing, 401(a)(4) testing, 401(k) testing and 401(m) testing all allow for separate testing of excludables and non-exculdables. There is a slight difference in definition between who is a non-excludable under 410(B)/401(a)(4) than under 401(k)/401(m), but that is a complexity that doesn't always matter.

Who is included in what test is defined first by the selection one makes under the optional rules. So, you can test 410(B)/401(a)(4) separately for those who satisfy the definition of non-excludable. But you don't have to. Hence, choices.

But, assuming you do choose to separate them for 410(B)/401(a)(4) testing, there is still the average benefits test to consider. Some people, including some at the IRS, have been known to say that there is one, and only one, average benefits test, the results of which are then used to determine whether one is stuck with the 70% test in the required rate-group tests. There are some, however, that read the code to allow (in fact, force) two separate average benefit tests when the optional election under 410(B)/401(a)(4) is made to test the excludables separately from the non-excludables.

So, in answer to your questions, it is an optional thing to test separately. If you decide not to, then all are included in a single rate group analysis. If you decide to, then, well, you test the groups separately.