I have 4 companies (A, B, C, D) all owned by the same 4 people (Owner 1, 2, 3, 4) in differing percentages.

Owner 2 is the daughter (age 30) of Owner 1. Owner 1 owns more than 50% of each company (looking at each company separately). Is my reasoning correct, that the Owner 1 'owns' Owner 2's shares since he owns more than 50% of the shares of the company?

If my reasoning is ok so far. Now, I am down to three owners to do determine if a controlled group exists. Owner 3 does not own any of Company C, so he is dropped from the testing when I am testing for a CG with all 4 companies together. There is still a greater than 80% controlling interest when I remove Onwer 3.

If I pass the 80% 'controlling interest' test, and then pass the 50% 'effective interest' test, then I have a Controlled Group consisting of all 4 companies.
Then am I right that only 1 coverage test needs to be performed OR do I need to look at other sub Controlled Groups that might exist and perform other coverage tests?

I, at least, cannot venture an opinion without more specific information on the exact percentage of each corporation for each owner, and the specific relationship, if any, between each of the owners. I'm also assuming that there is no other attribution from options, that there is no exluded stock, that all stock is the same value and that there is no non-voting stock, etc...

I agree that you always have to consider whether there are any CG's that do not encompass all four corporations and perform testing accordingly.

I have a great dislike for CG/ASG situations, as I find they offer unparalleled opportunities to look like a chump, (which I can do very nicely on my own!) particularly since the clients always neglect to tell you some little fact that changes the whole scenario. We always advise them to seek experienced ERISA counsel.

What I was trying to get at was: If Companies A, B, C, D comprise one controlled group, can I perform only one coverage test?

For example, company A and B are a controlled group as well. Do I need to perform a coverage test for this grouping as well? And if Company B and C meet the requirements to be a controlled group, do I need to test them as well?

Bel -- What do you make of IRC Section 1563(b)(4) and Reg. Section 1563-1(c) (http://frwebgate4.access.gpo.gov/cgi-bin/P...action=retrieve)? Does the component member rule have any application to qualified plans or, because it relates to making an election, does that mean that it is only applicable to consolidated return matters (the primary purpose of Section 1563)?

Sieve - not considering myself any sort of CG expert, this is dangerous territory, and I'm probably misinterpreting your question. But I've interpreted 1.414(b)-1(a) to essentially ignore the component member rules for a CG determination (which we don't make anyway, but tell the client to consult ERISA counsel.) So while the component member rules you mention can make a difference for corporate income tax filing, it seems to me that CG status for retirement plan purposes ignores 1563(b).

Bel --

I've never consulted the component member rules for controlled group determinations, and now you've reminded me why (by reference to the 414 reg.)! You also pointed to the authority (last sentence of the 414(b)-1(a) reg.) why multiple controlled groups are treated independently.

I seem always to forget that we get to the 1563 rules for corporate controlled groups through 414(b), and therefore a starting point for the anlayis should be 414(b) & its regs.!

So we know that Company A, B, C, and D comprise one CG. There is no other company relationship with any other companies aside from the four listed. When you stated "I agree that you always have to consider whether there are any CG's that do not encompass all four corporations and perform testing accordingly.", are you saying that if Company A and B comprise their own CG, that needs to be tested? And if A and C comprise their own CG, you need to test them separately? Or is it enough that if A,B,C,D are a CG, I don't care about the subgroups that might exist?

If A, B, C & D are in one controlled group because of common ownership by 4 individuals, you do not have a separate controlled group just because A, B, C & D are also a controlled group just considerering 3 of the owners. However, if, by considering 3 owners, there's an A, B & C controlled group, and considering 2 owners there's a B/C controlled group, then you have 3 controlled groups: A, B & C; A, B, C, & D; and B & C. Each controlled group must pass appropriate t
esting.