Controlled Group, multiple plans, controlled group status never reviewed

Company A owns 20 other companies at 100% ownership. All 21 (20 plus company A) entities have their own separate plans with different record keepers, TPA's and financial advisers. No one has performed combined coverage testing for the past 15 years.  The plans vary from safe harbor to traditional 401k with profit sharing. All plans are small plans so on their own have not had an audit requirement. All plan year ends are calendar year.

Question: Is this an obvious VCP submission requiring them to go back and run coverage testing for X years? Or can this be self corrected? 

Thank you

There's no chance of QSLOB status? 

I would do testing (maybe pick a few years as a starting sample) and see if it passes. If it passes - then there isn't anything to correct. 

If it fails - then look at VCP, cause most likely yes, it would be hard to categorize this as a insignificant failure. 

As JAA says if testing passes, noting to correct. If testing fails VCP is likely only option.

Agree with justanotheradmin and Lou S.

So we are clear about what is involved, each of the 21 plans will be tested for coverage for each year.  Of course, the deferral, match, and nonelective elements of those plans must be broken out and tested separately, particularly if there are different eligibility or allocation conditions.  For each plan, the numerator will consist of the employees who benefit from that specific plan.  The denominator will take into account all nonexcludable employees from all the employers (based on the eligibility conditions of the plan being tested).  If a plan does not pass the ratio percentage test, and you need to perform the average benefit test, then for purposes of the average benefit percentage test (only), you will take into account all nonexcludable employees and all plan maintained by the controlled group (even though you cannot permissively aggregate those plans for other purposes. 

And by the way, if one of those plans has nonelective contributions which require the general nondiscrimination test, you must rerun the nondiscrimination testing taking into account all the nonexcludable employees.

You don't need EPCRS to perform the testing correctly.  What you do need is time, data, and somebody willing to pay for the work.  And if every plan and subplan passes for every year (which it conceivably could if each company has a similar percentage of HCEs and NHCEs), you heave a huge sigh of relief.

And if one of the subplans for one of the years does not pass, you have a demographic failure for that plan.  Demographic failures can only be corrected under VCP.

Note that it is perfectly acceptable to have some plans ADP tested and some safe harbor, so long as each can separately pass coverage. But, if you have HCEs that participate in more than one plan, that complicates your testing.

Or, as an alternative, you could consider putting the whole shooting match under VCP and work out more realistic approach.  The choice is yours.

Thank you all, I really appreciate the expertise of those who responded.

Mr. Watson, I was actually reading a Q&A you wrote on a somewhat similar topic before I posted this. Thank you for the detailed clarification.

Great to hear from the great Derrin Watson!

As you start ding the testing, if you see some of them failing coverage, keep in mind that you might be able to aggregate some plans together.

For example, when we did this for four large plans and one small plan for five employers all owned 99% by a foreign company, two of the plans were current-year tested and could be aggregated (that helped us pass), two were prior-year tested and could be aggregated but did not need to be aggregated to pass, and the last small plan was safe harbor and had to pass by itself. That last plan just squeaked by using the average benefits test for coverage and by applying the OEE rule and some other uncommon testing options that are available. Of course, it's more complex than just that since you have up to three coverage tests for each plan as Derrin mentioned, but just keep aggregation in mind when you start getting these going (after they've engaged you, of course).