Defined Benefit contribution

[jkdoll2]

Two Owners, One employee - Employee is Highly Compensated - does he have to be included in the plan or can we carve him out since he is Highly Compensated?

Thanks

[rcline46]

You should be able to answer your own question:

1. Does the plan pass 401(a)(26) if it is a DB plan?

2. Does the plan pass the Ratio Percentage test?

3. if 2 is no, does the plan pass the Average Benefit Percentage Test?

See, now you know the answer!

[Andy the Actuary]

While you must include the owner, there is nothing to say that you have to provide the same level of benefits. You can always discriminate against an HCE.

[SoCalActuary]

While you must include the owner, there is nothing to say that you have to provide the same level of benefits. You can always discriminate against an HCE.

Provided that you give at least the top-heavy minimum. Easier to just exclude the 3rd wheel HCE.

[rcline46]

SoCal - you forgot the original post - only 3 eligibles - two owners and one non-owner - all HCEs. Lucky all are HCEs, so you can pass ABPT with non-owner excluded. Just pray they never become an NCE!

[SoCalActuary]

SoCal - you forgot the original post - only 3 eligibles - two owners and one non-owner - all HCEs. Lucky all are HCEs, so you can pass ABPT with non-owner excluded. Just pray they never become an NCE!

I was responding to Andy, expressing my preference that the third-wheel HCE be excluded. If excluded, then you don't have to give them a top-heavy minimum benefit, which can be very expensive. My review of TH min in a DB plan is that the cost can be 15% of pay for an older non-key.

[frizzyguy]

While you must include the owner, there is nothing to say that you have to provide the same level of benefits. You can always discriminate against an HCE.

And try and be creative when you put them in groups. Don't name the groups directly or by age. Say something "group 1 consists of owner HCEs hired before 1980, group 2 consists of HCEs hired after 1979 and group 3 consists of non-owner HCEs". Silly symantics but the IRS doesn't like if you name participants directly in their own group.

[John Feldt ERPA CPC QPA]

"the IRS doesn't like if you name participants directly in their own group"

What have they cited when they've reviewed plans with this - are they concerned about coverage, 410(b), or over testing of benefits, 401(a)(4) - or both?

[frizzyguy]

"the IRS doesn't like if you name participants directly in their own group"

What have they cited when they've reviewed plans with this - are they concerned about coverage, 410(b), or over testing of benefits, 401(a)(4) - or both?

Off the hip I can't find anything. This is only for DB plans mind you, in a DC plan it doesn't matter. I asked around and if I find something written down I'll pass it along.

I know I have read it and heard it several times. I work plans drafted by attorneys that do it, though I've seen attorneys do things I would NEVER do with plan design.

I know it's silly that they tell you not to do it but don't care if you create a category with the obvious intention of creating their own group.

Does anyone else have an actual cite? I'll keep looking.

[jkdoll2]

"the IRS doesn't like if you name participants directly in their own group"

What have they cited when they've reviewed plans with this - are they concerned about coverage, 410(b), or over testing of benefits, 401(a)(4) - or both?

Off the hip I can't find anything. This is only for DB plans mind you, in a DC plan it doesn't matter. I asked around and if I find something written down I'll pass it along.

I know I have read it and heard it several times. I work plans drafted by attorneys that do it, though I've seen attorneys do things I would NEVER do with plan design.

I know it's silly that they tell you not to do it but don't care if you create a category with the obvious intention of creating their own group.

Does anyone else have an actual cite? I'll keep looking.

This is just a proposal plan I was running. I will do one including the participant, and one with excluding him since he is an HCE (tests pass)

If we exclude him I will put in the document "HCE - non-owner excluded". I will not be naming him.

Thanks

[DMcGovern]

The IRS LRMs that I'm familar with relate to DC cross-tested plans, but maybe the thought process behind them would follow for DB plans. The naming issue relates to the ABT for coverage and specifically the "Reasonable classification test", which requires that the classification be reasonable and established under objective business criteria. The IRS has taken the position that naming a person specifically in a class is not a reasonable classification.

A work-around for DC plans is to always pass coverage using the ratio test. Maybe your attorneys, frizzy, have already taken this into consideration?

[frizzyguy]

Maybe your attorneys, frizzy, have already taken this into consideration?

I don't think they have based on the testing!

[John Feldt ERPA CPC QPA]

As long as the plan (or combined plans if aggregated for coverage) provides a penny of benefits to enough NHCEs so the 70% ratio test passes for 410(b) purposes, then naming people should not be an issue, since it appears to only matter in the coverage test. Under the nondiscrimination test, 401(a)(4), naming names for benefit accruals or having individual allocation classes is not in and of itself going to be a problem, but giving enough someones a zero could be an issue.

[frizzyguy]

As long as the plan (or combined plans if aggregated for coverage) provides a penny of benefits to enough NHCEs so the 70% ratio test passes for 410(b) purposes, then naming people should not be an issue, since it appears to only matter in the coverage test. Under the nondiscrimination test, 401(a)(4), naming names for benefit accruals or having individual allocation classes is not in and of itself going to be a problem, but giving enough someones a zero could be an issue.

I think I had always known it as a rule of thumb, never directly name participants but that makes sense.

[AndyH]

As long as the plan (or combined plans if aggregated for coverage) provides a penny of benefits to enough NHCEs so the 70% ratio test passes for 410(b) purposes, then naming people should not be an issue, since it appears to only matter in the coverage test. Under the nondiscrimination test, 401(a)(4), naming names for benefit accruals or having individual allocation classes is not in and of itself going to be a problem, but giving enough someones a zero could be an issue.

Agree, well summarized. And I'd add that the only real no-no is that it is not a good idea to categorize based on age, especially if it can be demonstrated that older participants get less. IMHO.