I've got a top heavy plan. Basic 4 step integration formula. Plan has a last day/1000 hour requirement. I've got 14 NHCEs & 3 HCEs. 3 NHCEs are nonbenefitting, nonexcludable. 2 NHCEs receive top heavy only. I know that the 2 NHCEs are benefitting for coverage. In the basic integrated plan I get the pass on 401(a)(4), but since this is top heavy is this a multiple formula situation? If so I've got to pass 410(b) assuming the 2 don't benefit or I don't get the free pass on 401(a)(4); correct?

coverage: how many NHCEs are covered as compared to HCES covered. doesn't matter how much an individual gets.
therefore, 11 / 14 = 78.57%, passes ratio percentage test, goes on the sched T.

.........
401(a)(4) - now it matters how much one gets. There is an exception to the multiple formula rule. You can treat the top heavy only people as neot benefitting.
now you have 9 / 14 = 64.28% fails ratio % but will pass safe harbor for nondiscrimination classification test.
If you can pass average benefits % test, you would be deemed to have a safe harbor formula and therefore not have to perform rate group testing.
My initial reaction would be you would want to pass using the allocation method. I believe, based on the population, if this is a non 401(k), if the plan is fully integrated at the TWB, as long as the base % < or = to 9.34% you would pass. That was a quick test on an excel sheet, imputing disparity.
If you were to test using accrual method, you have cross tested and I would think you are stuck having to provide gateway minimum. but their is a rule under broadly available rates 1.401(a)(4)-8(b)(1)(i)(B)(3)(iii) [I think I typed that correctly] - two alloaction rates can be aggregated and treated as broadly available if each group can satisfy 410(b) without regard to the avg ben % test...differences due to permitted disparity can be disregarded.
Thus it would appear you pass this and are ok.

I was afraid it would be considered multiple formulas. Now I'm somewhat lost. We handle some fairly simple cross-tested plans, but nothing 2 difficult.

1. Contribution formula is integrated at 60% of TWB. So does that mean when I do the average benefits test I adjust allocation rates for imputed disparity?

2. Assuming #1 is yes, nobody has compensation in excess of TWB, all allocation rates are greater than 5.7%. I should increase everyones allocation rate by 5.7%?

3. Plan has semiannual entry dates. Comp. prior to entry excluded. Obviously top heavy done on full year comp. When doing the average benefits test, for participants entering 7/1 can I use only comp. from 7/1-12/31.

Thanks in advance for any guidance

mentally I can't think, especially on a Friday afternoon.
If you were to test a fully integrated plan on an allocation basis (and impute disparity) an NHCE's E-Bar would be
assuming a base % of 6%
base 6% plus 5.7% imputing disparity = 11.7%
I don't care how much the HCE makes, no matter what comp he will also end up with 11.7%.
e.g. comp = 100,000 TWB =87,000
so HCE received 6% or $6000 plus 5.7%(13000)=741 total 6741.
his e-bar is the lesser of
his C rate is 6741/ (100,000-43500)=11.93
D rate is [6741 + (5.7%*87000)]/100000 = 11.7
try any comp figure greater than the integration level. you always get 11.7%

on the other hand

at 60% of the TWB (or 52,200) he receives 4.3% (47,800) = 2055.4 extra

thus his adjusted rates would be
C rate 8055.4/(100000-43500)= 14.26
D rate 8055.4 + (5.7%*87000)]/100000 = 13.014
which is less than the NHCE of 11.7

so using a level less than 100% integration will produce a larger e-bar and thus increase the chances of failure if testing by allocation method.

so, answer to #1 is NO you don't change how you perform the test
#2 becomes a non existant question.
#3 - the regs say use 414s comp when testing. this can be from date of entry despite the fact the allocation was based on full year so you get a gift in that situation - it should help your testing.

I can't pass using allocations, so I gotta use accrual. A couple more questions.

1. I'm kinda slow, I understand the response in the last post to my question regarding imputing disparity, but I'm still lost. I guess my question was (& really still is) if the document has an integrated formula, must I impute disparity to adjust EBARs in my average benfit test. (In other words can I use 6% instead of 11.7%)

2. If I do have to impute disparity to adjust EBARs, it is my undersatnding that I now have to follow DB rules since I'm using accrual method. Is this correct? Where can I find a good explanation of how to do that? (Relius is just adding .65% to the EBAR, but I'm not going to use anything I don't understand & that I can't manually do on Excel.)

You should bite the bullet and read the A, B, C, D rules of 1.401(a)(4)-7. Put it into Excel once, and it's not that bad at all. I cannot memorize it; my brain automatically rejects any requests for storage of that and the affiliated service rules, among other things.

Tom, is there any other way to think?

I am sure a wife whose husband cheated on her would answer differently.

The plan document formula has nothing to do with the testing under the ABT. Nothing. Whether the formula is comp-to-comp, or integrated or new comparability doesn't matter. You are free to run the ABT on the basis that you want to, whether that includes imputing permitted disparity or not.

If you impute permitted disparity, you must do so in a manner that is consistent with the test. If you are testing on contributions you would use the x% up to 5.7% rules associated with a defined contribution plan. If you are testing on benefits you would use the permitted disparity factors associated with the retirement age (testing age, actually), which in all likelihood is indeed 0.65%. But just like the DC rules, don't forget the "no more than 2 to 1" rule, so if an individual has a DB accrual rate of 0.5%, the permitted disparity factor can't be 0.65%, but instead would be limited to 0.5%.

The Butler:
generally, it will only help your testing to impute disparity, though that is an assumption and you wouldn't have to. Quite possibly, a 'poorly' written document will list the 'assumptions' and you are stuck. Quite an oxymoron having 'required assumptions' but I have seen it.
I suppose if you had an NHCE with comp greater than the TWB or covered comp if might hurt to impute, or if you had a bunch of NHCEs with less than 5.7 (or an EBAR less than .65) you would not want to impute disparity.

pax: obviously you are not a sports fan. practically every football weekend I hear the following on Monday: "The coach couldn't have been using his brain when he thought up that play", or at least something similar.