Restructure the Plan

[Logan401]

Cross-testing allows you to restructure the plan into 2 or more component plans.

The key words I am interested in are "or more."

Can you split the plan into 3 component plans?

A plan consists of 4 HCEs, ages 64, 55, 43 & 29.

It also consists on 16 NHCEs of various ages.

Can we test based on HCE1 & HCE2 in one compnent, HCE 3 in a 2nd, and HCE 4 in a 3rd?

Also, does it matter how you divvy up the NHCEs into the different component plans?

Thank you!

[Tom Poje]

a more important question is whether it will do any good

1.401(a)(4)-9( c)(1) a plan may be treated...as consisting of two or more component plans

1.401(a)(4)-9( c)(2) ...the employer may select the group of employees in any manner, and the composition may be changed from year to year. every employee MUST be included in one and only one component plan...

1.401(a)(4)-9( c)(3)(ii) ...can.t be used for gateway purposes

If using Relius, it can be 'tricked' into producing reports if need be.

It makes a big difference how you 'divvy' things up. I assume that is one of those technical terms.

usually young NHCEs are used in the component plan tested on an accrual basis.

so lets suppose I have a plan, we will call plan O

I split the plan into different component plans (in some ways I am simply pretending I have 4 plans instead of 1

HCE 1 is in Component Plan 1 of Plan O (or C1PO) - test on accrual basis with some young NHCEs, all other ees includable and not benefiting

HCE 2 in C2PO

HCE 3 in C3PO (may the force be with you getting that one to pass!)

HCE 4 in C4PO

[Logan401]

Thank you Tom.

So, let's assume HCE1 that is owner's son and 29. He can be placed in his own component plan & be tested on a contributions basis with maybe 1/2 of the NHCEs.

Can two of the HCEs be accrual tested in one group of NHCEs, and the remaining HCE accrual tested in his own separate group with another group of random NHCEs?

I am assuming the number of NHCEs in each component plan does not matter?

[Tom Poje]

correct. it doesn't matter how you do things as long as no one has his contributions counted in more than one group.

and yes, that is common, the owners son ends up in a group tested on an allocation basis.

[Logan401]

Very good to know, thank you Tom!

[chc93]

Also... coverage testing for each component plan has to each pass the 70% ratio test, since the way you split the HCE's and NHCE's will most likely not be a reasonable classification so the ABPT for coverage cannot be used.

[Logan401]

A pro rata allocation automatically passes the cross-testing rules. I would think that the same would apply to an Integrated allocation, but would like confirmation on that..

[Tom Poje]

cross testing implies you are taking a dc allocation and 'crossing' over and testing on an accrual basis (though I guess you could do the same with a db accrual and cross over and test on an allocation basis)

so no a pro rate allocation does not automatically pass cross testing rules.

e.g. if I give a young HCE 3% and an old NHCE 3% and cross test I would fail.

but a pro rate allocation is considered a safe harbor allocation (not to be confused with safe harbor 401k) so no testing is needed, same with an integrated plan. but you could test if you want on an allocation basis and lo and behold you would pass (of course!!!!) conceivably, if the plan is integrated at 5.4% and you impute disparity you could actually fail testing on an allocation basis, but the regs say don't sweat it.

if only 60% of the NHCE benefitted, yes arguably you would fail the rate group test, but then you would fail coverage as well so it is something of a moot point.

[Logan401]

That is what I was trying to convey. A pro rata allocation will pass as long as it is tested on a contributions basis. So, what you are saying for Integrated is the same. An automatic pass as long as you are testing on a contributions basis.

Similar as we implied above, if we assume there are 2 HCEs and 10 NHCEs:

HCE1= 20% allocation

HCE2= 5% allocation

10 NHCEs= 5% allocation

We can test the HCE2 with 9 of the NHCEs on a contributions basis, and HCE1 with 1 of the NHCEs that has an equal or higher EBAR on a benefits basis.

Is it as easy as this?

[Tom Poje]

you are forgetting the rule 'treat everyone not in the component plan as includable and not benefiting'

so you have component plan accrual basis (assuming the NHCE in the rate group)

1/10 NHCE ratio

1/2 HCE ratio

this is only 20% for the rate group test.

you have 10 /12 for nhce concentration percent or 83.33 always round down, so 83%

so a midpoint of 27.75% and at that point the big fog horn goes off

in addition, even if the ratio % passes, you need to pass avg ben pct test as a whole, since your ratio % is going to be less than 70%.

[Logan401]

So, for the accrual component plan, you must include all the NHCEs & HCEs in the respective denominator even though some are in the contributions component.

In this case, if we add an additional NHCE to the accrual component, we have:

2/10 NHCE ratio

1/2 HCE ratio

The result is 40%, so we are above the 27.75% and no big fog horn.

And, just to confirm my first question, Integrated is an automatic pass?

i read the "regs. say don't sweat it" part, but you also mentioned coverage testing.

Thank you!

[Logan401]

I should have stated 4/10 NHCEs so we do not have to bother with ABT.

And, just so I understand, when splitting groups of EEs into different component plans for testing, the groups themselves must be rate group tested using ratio test and ABT if necessary.

[Tom Poje]

think of it this way, instead of 1 plan being split into 2 component plans you have 2 plans

the plans will be tested separately (permissive aggregation not used)

one plan consists of all people tested on an allocation basis

the other plan consists of all people tested on an accrual basis.

for avg ben pct test, no matter whether you aggregate or not you would combine everything because there is one and only 1 test

when looking at the individual plan all members of the other plan are includable and not benefitting.

now you look at plan 1 and use whatever mean you can to pass testing

now you look at plan 2 and use whatever means to pass testing.

note: again, you can't avoid the gateway with the group tested on an allocation basis.

........

yes integration is considered a 'safe harbor' formula, assuming you are using the proper taxable wage base and proper excess %

[Logan401]

Okay, now you have me confused!

You stated previously the following:

you are forgetting the rule 'treat everyone not in the component plan as includable and not benefiting'

so you have component plan accrual basis (assuming the NHCE in the rate group)

1/10 NHCE ratio

1/2 HCE ratio

this is only 20% for the rate group test.

But, if only one HCE is in the one accrual basis plan, would it not be:

HCE1 tested on accrual basis with one NHCE:

1/1 NHCE ratio

1/1 HCE ratio

1/1= 100%

HCE2 tested on allocation basis with 9 of 10 NHCEs

9/9 NHCE ratio

1/1 HCE ratio

Result = 100%

Maybe the boat left without me?

[Tom Poje]

I guess I am unable to convey the concept and need to leave that to someone else that can explain better.

again

if you had 2 separate plans (or 1 plan split into component plans) you count ALL bodies in the denominator whether you aggregate the plans or not. thus your denominator will always have all bodies counted, not just the plan (or component plan you are looking at)

the numerator consists of just the plan (or plans) you are aggregating (or not aggregating)

[Logan401]

Hi Tom,

I did understand it the first time you explained it when you used the term rate group test. I understood that to be all employees included in the denominator within each component plan.

However, after your 4:13 post yesterday I interpreted it differently.

I believe I am grasping it very well now!

[Logan401]

Tom, for the numerator are you only using the HCEs & NHCEs that are actually in the component plan? So, in this example, it will only be one even though HCE2 may have a >= benefit than HCE1?

Just want to be clear on the numerator. Thanks!

[Tom Poje]

yes, because when you are looking at the component plan you are treating everyone not in the component as 0 so HCE2 can't have a benefit greater than HCE1 for that component plan

[Logan401]

Okay, think I got it. Don't give up on me.

Let me know if the fog horn sounds!:

Example:

HCE1: 20% allocation 5.8 EBAR

HCE2: 5% allocation 20.00 EBAR

10 NHCEs: 5 have EBARS >= 5.8, & each receive 5% allocation

Accrual component test

HCE1= ½

5 NHCEs with EBARs >= 5.8= 5/10=1/2

So, 1/2/1/2= 100%

Allocation (contribution) component test:

HCE2= ½

5 NHCEs with contributions >= 5%= 5/10= ½

So, 1/2/1/2= 100%

Both component plans pass

[Tom Poje]

yes, that sounds correct

go out a like a firework and celebrate.

but not so close here, too dry, I don't need the fire

[Logan401]

Tom, thank you so much for your input and showing me the way.

I will light a sparkler in your honor tomorrow!