permissable correction method?

[AndyH]

I'm looking at a takeover "super integrated" non-standardized prototype which allows for several selectable correction methods in the event of 401(a)(4) general test failure.

One of the choices, which was in fact selected, is what I'd call the "Robin Hood" method, that is reduce the allocation to the HCE in a failing rate group, and give that amount to the highest NHCE below the failing rate group to enlarge the failing rate group. Repeat with the next NHCE until it passes.

The test I'm looking at fails badly, so a correction is needed. I have an internal opinion that the correction method in the document should be ignored and the plan should be amended within the 401(a)(4) correction period to adopt a method which allocates additional $.

Clearly a corrective amendment cannot provide for a takeaway, but the existing language provides for such a takeaway.

Any dissenting opinions, i.e. anybody think the takeaway method is acceptable if it is in the document?

[Guest Doug Goelz]

I've seen some plans like that. The ones I've seen get around the take-away issue by specific plan language that declared that no contribution was deemed accrued to any participant until the total contribution was allocated among all groups (and the total contribution was not considered allocated until all testing was passed).

This language was drafted and supported by a pension attorney who really knew 401(a)(4) issues. Its been a few years since I actually did DC work, so I don't know if anything has caused him to have a different opinion.