Mike Preston

A dove-tail provision will cure most of the issues being raised in this thread. If you sprinkle a little fairy dust on top (the ability of the Plan Administrator to resolve ambiguities) my guess is that 99.44%+ of the issues can be dealt with.

Some pre-approved plans now have a dove-tail provision that incorporates the terms of the CBA as a specifier to the allocations.

Forget this sub-comment, I was basing it on an ultra-conservative (and fictitious) body count.

You are correct about plan 3 as long as the transaction satisfied 410(b)(6)(C). When I say employer wide I mean taking into consideration all employees of all three entities. However, now that there is a 410(b)(6)(C) transaction to consider, the focus for testing shifts from current year counts to a determination on a specific date (see below). One little discussed consequence of a transaction that satisfies 410(b)(6)(C) is that all plans are treated as passing coverage during the transition period. So, if the testing required to satisfy 410(b)(6)(C) pans out all three of your plans are treated as passing coverage for 2016 and 2017. In that case, you would look to the satisfaction of 410(b) "immediately before" the transaction for Plans 1 and 2 [testing based on those two employers comprising a controlled group at that time] and also for Plan 3 [testing based on that employer and whatever other employers comprised a controlled or affiliated group at that time]. This stuff can get real messy, real quick.

Plan 1 is fine. Plan 2 is safe if you pass the Average Benefits Test (which involves employer wide data). Plan 3 is safe if you pass the Average Benefits Test (which involves employer wide data) and satisfy an easy-to-satisfy facts and circumstances test (which it goes without saying that I think you will probably satisfy). Can you define your role? Are you a plan sponsor?

JPete, you misunderstand. Using the information from the controlled group does not mean doing one test. It means doing three tests. Post some body counts (HCE's, NHCE's) and somebody will show you how it works.

No way to tell. Have to test coverage using the information from the aggregated controlled group unless they qualified and registered as SLOB's (separate lines of business).

In a non-qualified plan, sure. But in a qualified plan?

Huh?

There are a lot of them, do you have any particular subset in mind?

Seriously?

In the precious few plans that do not allow distribution prior to a specific age one might find the age lower upon disability.

I can't tell you how strongly I disagree with the last two posts. Life is difficult enough without us creating adverse rules that don't exist. Fact: Identified individual is an NHCE. Fact: An amendment which loosens eligibility in order to advantage an NHCE is not discriminatory. Fact: The natural operation of plan provisions (such as those that turn the NHCE into an HCE in a future year) can not be discriminatory. Strong letter to follow.

It is a bit more nuanced. I think the determination of employed status is definitely a settlor function. But it isn't unchallengeable by the Plan Administrator if it is obvious that the employer is doing something wonky (defined in ERISA Section 999.9). And, yes, that same person who is both ER and PA needs to put on the PA hat and blow the whistle if the ER is, in fact, doing something wonky. Of course, that isn't likely to happen. The structure of thinigs is such that the PA is exposed in a fiduciary sense for not taking action, while the employer is only exposed if the wonky action has employee/employer consequences outside of ERISA.

In our little corner of ERISA, the employer is also the plan administrator.

I think OP is confusing DFVCP (a DOL program) with Rev. Proc. 2015-32 relief (an IRS program). 1) Correct 2) 2016 3) 1/1/2016 - 12/31/2016 4) See this page: https://www.irs.gov/retirement-plans/penalty-relief-program-for-form-5500-ez-late-filers

How much was in the plan on 12/31/2016?

Yes.

See 1.416-1-Q&A T-39 for a similar example dealing with top-heavy calculations.

It is fair not to run the General Test because you know it will pass. That does not make it a Design-Based Safe Harbor.

It is fair not to run the General Test because you know it will pass. That does not make it a Design-Based Safe Harbor.

We use 541990.

What is Hacienda?

$162k (it is comp paid during the fiscal year).

MoJo, what part of the process that I described is objectionable? Are you saying that your system has no way to implement a separate interest account for an alternate payee?