I thought about this some more.
the whole thing smells. and smells real bad.
oh, an owner quit and now we want to retroactively amend the document to conform what took place in which the owner was 'accidently' provided a contribution.
The company doesn't run the cross testing, it is the TPA or whomever, and apparently it was discovered that the owner won't get anything, so now lets go ahead and give him something. (instead of rerunning the test how it should.)
and let's call it a correction under -11g and to be fair we will give to the nhce terminees as well. we never gave terminated NHCEs in the past, but then we never had a terminated owner either.
arguing they can pass all the mathematical testing reminds me of the comment made by the IRS in regards to cross testing in general
Although these designs may allow the plan to satisfy the vesting or numeric general tests for nondiscrimination and the associated regulations, they don’t satisfy Treas. Reg. Section 1.401(a)(4)-1(c)(2), which requires that the provisions of Sections 1.401(a)(4)-1 through 1.401(a)(4)-13 be reasonably interpreted to prevent discrimination in favor of HCEs.
so is it a reasonable interpretation to say this flies? lets see, we give an owner $40,000 or whatever and give terminees who probably have minimal comp a 5% contribution and because mathematically it works out, it is ok.
dang, might as well have a document that reads
in no year terminees will receive except in a year in which it is an owner and then all terminees receive. (Could you at least have said, oh I didn't see that the owner is past NRA and those folks do actually get a contribution)
bet you would have problems getting that document approved.
or taking it a step further if it was a MP plan at 5% would you say it is ok to retroactively amend to 10% to all and call it a corrective amendment because that is how we ran the allocation?