Adding participating employer via 11(g) ?

So a small employer (1 HCE and 4 HCEs) has a SHM plan. we take over tpa work and discover they are a controlled group with another entity with no plan (1 HCE and 2 NHCEs). the hce is the same person in both scenarios. So now their nice and simple safe harbor plan fails coverage.  

we can pass coverage using ABT with some QNECs (this option is reserved to be used whenever needed). But, the fail-safe in the document is to add excluded otherwise eligible employees until it passes. Cannot really find anything definitive on whether it is permissible to add a participating employer in this scenario. 

the qnec to pass abt is the cheaper option but still good to know what the friday morning experts of benefitslink think.

thanks 

It is not entirely clear whether adding a member of a controlled group to an existing plan that is already sponsored by another member of the controlled group counts as an amendment or a plan adoption. If it is a plan adoption, then you should be able to do it retroactively under the new 401(b)(2) as added by the SECURE Act. If it is not a plan adoption, then it is an amendment, and it should be able to be adopted retroactively under the rules of 1.401(a)(4)-11(g) for purposes of correcting a failed coverage test. Either way I think you should be able to do it.

About the ABT, read your plan document carefully. Even with the failsafe language, you might not be precluded from satisfying the coverage test using the ABT. If the plan allows discretionary profit sharing contributions, then you may not even need to use a QNEC.

And if you do end up using the ABT, remember that it can be calculated using cross testing.

Isn't there a potential problem with using the ABT for COVERAGE testing, (but not a problem for nondiscrimination testing) due to the requirements of 1.410(b)-4(b), and the IRS stance that having each person in their own group (as many plans do) violates the "reasonable classification" requirement? The IRS takes the stance that this is tantamount to enumerating by name or having the same effect, and by definition it is not a reasonable classification? If your document is set up with actual groups, rather than everyone in their own group, then this shouldn't be a problem.

Who mentioned having anybody in their own group? But let's assume that's the case. He would only fail to be able to use the ABT if somebody was excluded by getting an allocation of zero?

No one mentioned having everyone in their own group. Hence my use of the word "potential." But please tell me what I'm missing - I'm sure there is something - if no one is getting allocation of  zero, why would you be using the ABT to pass coverage testing anyway? I'm missing some key point here, I'm sure. Thanks.

The excluded class here is the employees of the second employer, which is a member of the same controlled group as the plan sponsor but which has not adopted the plan. Is that a reasonable classification? I don't know for sure, but it doesn't seem unreasonable on its face.

Thanks for the input. I'd tend to agree, and that would be my argument if it were questioned. But I'm not certain either - seems a little gray...

Ms Ilene addressed this earlier this year.

thanks all. 

If separate businesses are not a reasonable classification then there is no such thing as a reasonable classification.

Ilene wrote what she wrote before the update to epcrs. She implies strongly that since the requirement to benefit everybody has been eliminated it very well may be possible to retroactively adopt as an 11g amendment or as an SCP amendment. I didn't check the effective date of the change. That might change things.