Coverage Testing and Fail Safe Language

[justatester]

I have a group of plans that are part of a controlled group...

Plan A: QACA SH, but requires ACP test since match above limit-uses prior year

Plan B: QACA SH-No ADP/ACP testing requirement

Plan C: QACA SH, but requires ACP test-uses current year

Plan D: Not SH

Plan B is on a standardize prototype document with "Fail Safe" language for coverage failures (ie: 1000 hrs/last day).

Of course, Plan B does not pass the ratio test. Plan B does not have a last day/1000 hours requirement. Since the plan is not failing coverage because of this, is it possible to run an ABT in order to pass coverage for prototypes?

Another question...Can I NOT apply the under 21/less than 1 YOS option to group B for all testing and apply it to all other plans? Of course in each case the denominator would be the entire controlled group (A, B, C & D)

Example: Plan B coverage denominator would all employees. Plan A's test would be divided in to 2 "groups", those over 21/1 yos & those under 21/1 yos and the denominator would have all of the other plans divide as well.

Any thoughts?

[Tom Poje]

my understanding

I have a document for Plan B with fail safe language.

Therefore, it says if I fail coverage I bring in the next person on the list and keep bringing in people until I pass.

I don't think there is anything is such language that says

"But if the plan is part of a controlled group I can use the avg ben test to pass coverage."

.........

my understanding if you are not aggregating plans then yes, you could test one plan on an otherwise excludable basis, while you don't do so for the other plans.

[ETA Consulting LLC]

I have a group of plans that are part of a controlled group...

Plan A: QACA SH, but requires ACP test since match above limit-uses prior year

Plan B: QACA SH-No ADP/ACP testing requirement

Plan C: QACA SH, but requires ACP test-uses current year

Plan D: Not SH

Plan B is on a standardize prototype document with "Fail Safe" language for coverage failures (ie: 1000 hrs/last day).

Is it a "standardized" or "non-standardized". There is a huge difference with major implications.

Of course, Plan B does not pass the ratio test. Plan B does not have a last day/1000 hours requirement. Since the plan is not failing coverage because of this, is it possible to run an ABT in order to pass coverage for prototypes?

Okay, which is it?

It's hard to overlook this as your terminology appears to be inconsistent. If it is a standardized prototype, then that document doesn't contain a provision to exclude employees of other members of a controlled group.

Operationally, as a rule, you are allowed to include a plan that passes coverage and non-discrimination on its own in a group to help another plan pass. You can do this by Permissively aggregating two plans without having to perform the average benefits test. I'm still unclear on the fact pattern.

Good Luck!

[justatester]

ok...so here is the actual language...

If the Plan is a non-standarized plan, and would otherwise fail to statisfy the coverage requirements of Code 410(b) for a Plan Year because ofa requirment that a Participant complete a specified number of (greater than 501) hours of Service during the Plan Year or be an Employee on the last day of the PY in order to receive an Employer Contribution, then the Hours of Sevice requirement will be redued by one, and one again (but not below 501) until a sufficient number of individuals receive a contribution to pass the coverage requirements of Code 410(b). If the Plan fails to statisfy coverage after such process, then the last day requirement will be applied by treating all individuals wo were Employees on the day before the last day of the Plan Year, and the day before again, as having been employeed on the las day of the Plan Year until a sufficient number of individuals receive a contribution to pass the coverage requirement.

It is a non-standarized prototype.

I don't believe I have any options to aggregate since they are SH plans. As mentioned, some have an ACP testing requirement. Are you suggesting I could aggregate Plan B & C together for ACP testing, butt then I would need to run BRF?

Hope this adds some clarity...

[justatester]

Tom--thanks for the information about the otherwise excludable option and applying to only one member of the controlled group...Do you know if the regulation that would back that up.

Also, is it just our prototype document that has "fail safe" language or is the typically of prototype documents?

In Sal's book, it seems to suggest if after you have "activated" the fail safe language and you still fail...ABT might be available? Any thoughts?

[ETA Consulting LLC]

Okay, you are saying that "B" is failing the coverage ratio test, but not because of any accrual requirements. They are failing because of "class" exclusions; not covering the employees of "A", "C", and "D".

Now, I see what you are asking.

I am assuming, however, that A, C, and D does pass 410(b) as stand-alone plans; leaving plan "B" as your only problem.

Notice from the document language you included "because of a failure to complete a specified number of hours". This is clearly not the case, so the fail safe would not apply.

So, you could begin to permissively aggregate additional plans in order to get it to pass.

Does this begin to move you in the right direction?

[justatester]

Yes, exactly..Plan B is failing because other members of the controlled group do not participate in their plan. Yes, Plans A, C & D all pass coverage as stand alones.

I am not sure I have any aggregation options. Plan A uses prior year for ACP testing (they match to up to 9%), Plan B is both ADP & ACP Safe Harbor, Plan C usues current for ACP testin (they match up to 9% as well). Plan D is not a Safe Harbor plan.

Based on this, what aggregation options do I have?

[ETA Consulting LLC]

Now, we are getting somewhere; fact patterns are important.

We know that the fail safe option isn't relevant because of the aforementioned dicussion.

I would start by permissively aggregating B & C.

We know that when you aggegate for one, you must aggregate for all. At any rate, you are still aggegating with respect to each testing requirement.

After aggregation, the ADP should become a non-issue; since both share the exact provisions. When you aggregate for ACP, if a failure should exist, you could then distribute from both B & C since your safe harbor in B is invalid due to the failed coverage ratio test.

I'd get Poje to verify the logic, but it appears pretty straight forward.

Good Luck!

[justatester]

Ok...I get what you are saying...So if I aggregate B & C together (B only matches to 5% and C matches to 9%), I would then need a BRF.

I guess in theory I could aggregated A, B & C together for the deferral portion since all are SH for that. And then aggregate for coverage B&C for the match side.

Technically, I would only need to test the match portion from 5.01-9% in the ACP test for (A & C). But it is easier to test the entire match.

I just didn't think you could consider aggregating for coverage a plan that is fully SH (both pretax & match) vs one that is not?

Aggregating B&C passes ACP (total Match). Not sure about BRF, but should be ok since it has lower thresholds.

[ETA Consulting LLC]

I just didn't think you could consider aggregating for coverage a plan that is fully SH (both pretax & match) vs one that is not?

Well, technically, you are aggregating two plans that are safe harbor for deferrals and not safe harbor for match. The Safe Harbor for "B" is negated since "B" has failed the coverage ratio test. So, when you permissively aggregate with "C", you are effectively testing two like-kind plans.

BRF is actually a 'relatively' low threshold to pass.

So, you should be in good shape.

Good Luck!

[Tom Poje]

1.401(k)-1(b)(4)(iii)(B)

plans with inconsistent testsing methods...

may not aggregate a plan using the safe harbor provision and another plan that is using the ADP test, (or a plan that uses current year testing with a plan that uses prior year testing)

so it would be impossible to aggregate paln D with anything.

This is in regards to the ADP testing (both coverage and nondiscrim)

applying the same logic to the ACP test, I don't see how A and C could be aggregated since one is current and the other prior year testing. (1.401(m)-1(b)(4)(iii)(B)

but my understanding is that you could aggregate the ADP test one way and the ACP test another way

[ETA Consulting LLC]

1.401(k)-1(b)(4)(iii)(B)

plans with inconsistent testsing methods...

may not aggregate a plan using the safe harbor provision and another plan that is using the ADP test, (or a plan that uses current year testing with a plan that uses prior year testing)

so it would be impossible to aggregate paln D with anything.

This is in regards to the ADP testing (both coverage and nondiscrim)

applying the same logic to the ACP test, I don't see how A and C could be aggregated since one is current and the other prior year testing. (1.401(m)-1(b)(4)(iii)(B)

but my understanding is that you could aggregate the ADP test one way and the ACP test another way

The problem here, Tom, is that both plans have safe harbor for ADP; so ADP is not a problem. While "B" has a safe harbor provision for ACP, it doesn't pass 410(b) as a stand alone plan; so the safe harbor provision is rendered ineffective. You're testing it with "C" since the safe harbor is rendered ineffective.

The testing methods, then, are not inconsistent.

Good Luck!

[movedon]

Tom, ETK or anyone else...

In this situation where the (Relius) NS-proto adoption agreement has the failsafe box checked (evil), but the base document (4.3(m)) clearly says that the failsafe applies in the case of RPT failure due to a last day rule or hours requirement (and not, presumably, due to controlled group issues), then why wouldn't the average benefits test be your next step (rather than jumping straight to aggregating)? Am I correct in assuming that the ABT is on the table if the failed RPT has nothing to do with accrual requirements?