Gateway with ATD testing

[LarryDavid]

I have a DB/DC plan this is failing the minimum gateway test for 2010. The level of failure appears to be much worse when testing on an Annual basis (highest allocation rate = 25%) vs. Accrued to Date (highest HCE allocation rate = 13%). However, once I start analyzing the amount of contribution the plan has to make to correct the gateway, I think the contribution required is much higher under ATD due to having to multiply the contribution by total service. For example, in my test all NHCE's must have an allocation rate equal to at least 4.33% (1/3 of 13%). So if I have an NHCE with a current allocation rate of 4.00% (and let's say he has 10 years of service) does that mean the required contribution to satisfy the gateway is 0.33% x comp x SERVICE? So if comp was equal to $100K, instead of just having to put in $333 (0.33% x $100K), I'd have to put in $3,333, since the $333 would not be enough to get the overall allocation rate up to 4.33% once it is added to the total account balance and then divided by 10 years of service.

Am I getting all this correct? If so, should I just stick with the Annual method? The minimum for NHCE's would jump from 4.33% to 5.00%, but if my assumption with service is correct, it would be a much cheaper overall contribution to the plan.

And as a side question, if the highest HCE allocation rate is 26%, does the minimum required allocation for NHCE's jump from 5% to 6%, or is it just 26%/5 = 5.2%?

Much thanks in advance for anyone that can help.

[Tom Poje]

1.401(a)(4)-9(b)(v)(D) provides the min aggregation gateway requirements. the regs require the minimum to be at least 5% increased by 1% for each 5% increment (or portion thereof) by which the HCEs exceed 25%. e.g. 6% for an HCE that exceeds 25% but not 30%

as for using annual method vs accrued to date - you are never 'stuck' using one method or another, all you can do is run the numbers and see what happens. there are too many factors involved to say one is better than the other (besides the most import one, namely can your software do it correctly)

[LarryDavid]

1.401(a)(4)-9(b)(v)(D) provides the min aggregation gateway requirements. the regs require the minimum to be at least 5% increased by 1% for each 5% increment (or portion thereof) by which the HCEs exceed 25%. e.g. 6% for an HCE that exceeds 25% but not 30%

as for using annual method vs accrued to date - you are never 'stuck' using one method or another, all you can do is run the numbers and see what happens. there are too many factors involved to say one is better than the other (besides the most import one, namely can your software do it correctly)

Thanks Tom.

Follow-up questions/comments: So 6% is the required allocation rate for all NHCE's that benefit under the plan (since highest HCE is 26%). The average NHCE allocation rate in the DB plan is 5.66%, therefore each NHCE theoretically has to receive an additional .34% allocation to satisfy the gateway. My questions are as follows:

1.) Several NHCE's have higher allocation rates than 6%, therefore do they still have to receive an additional allocation to meet the gateway? I assume yes since I'm using the "average NHCE" approach (because several NHCE's have small allocation rates in the 1-2% range). So basically, by using the average NHCE approach to help reduce the required allocation for the lower NHCE's, I in turn create required contributions for the higher NHCE's that would otherwise not be needed. Is that correct?

2.) In my testing group, none of the DB participants receive allocations in a DC plan. The only reason we are aggregating the plans is that the DB plan does not satisfy nondiscrimination on its own. Therefore, when providing allocations to satisfy the gateway, how should this be done for the DB-only people? I see that QNEC's are not allowed. Does the plan sponsor need to create a DC account for these people?

3.) As mentioned above, I see that QNEC cannot be made simply to satisfy the gateway test. What exactly is the difference then, between a QNEC and an separate nonelective contribution made solely for purposes of satisfying the gateway? Is the only difference that a QNEC makes its way into the General Test, but a gateway contribution does not?

Sorry for all the questions. This stuff gets more confusing the more look into it!

[Tom Poje]

if the highest HCE receives a rate of 26%, then yes, the gateway is 6%.

one issue not discussed is top heavy, my experience is most combo plans are top heavy, you did not mention if that is a an issue. (and there is no 'avg' rule, its on an individual basis).

if you have an avg of 5.66% you have a pretty good cash balance formula for the NHCEs!

1. you are correct, if you use an average it doesn't matter if someone has an equivalent allocation rate that is higher, you still use the average.

2.has the dc plan even been updated to include gateway language? most documents I've seen simply say you provide each NHCE with enough to satisfy the gateway, so its like an addition allocation formula.

3. generally a QNEC is used to satisfy the ADP test. in that case it can't do double duty - be used to satisft the gateway as well, and in addition such QNECs force you to run 2 nondiscrim tests one with the QNEC and one without.

you are always permitted to put in a corrective amendment providing additional contributions to people to pass testing. it does not have to be a QNEC. corrective amendments have to be done before 9 1/2 months after plan year end.

[LarryDavid]

if the highest HCE receives a rate of 26%, then yes, the gateway is 6%.

one issue not discussed is top heavy, my experience is most combo plans are top heavy, you did not mention if that is a an issue. (and there is no 'avg' rule, its on an individual basis).

if you have an avg of 5.66% you have a pretty good cash balance formula for the NHCEs!

1. you are correct, if you use an average it doesn't matter if someone has an equivalent allocation rate that is higher, you still use the average.

2.has the dc plan even been updated to include gateway language? most documents I've seen simply say you provide each NHCE with enough to satisfy the gateway, so its like an addition allocation formula.

3. generally a QNEC is used to satisfy the ADP test. in that case it can't do double duty - be used to satisft the gateway as well, and in addition such QNECs force you to run 2 nondiscrim tests one with the QNEC and one without.

you are always permitted to put in a corrective amendment providing additional contributions to people to pass testing. it does not have to be a QNEC. corrective amendments have to be done before 9 1/2 months after plan year end.

Thanks Tom.

The DB plan has a grandfathered DB benefit with a traditional FAP formula. That is why the average allocation is relatively hight at 5.66%. The ultimate benefit is the greater of FAP and CB for the grandfathered people, so for the most part the CB doesn't even come into play (since the FAP almost always wins). For the non-grandfathered group, they've had their DB plan benefit frozen and they now receive a 5% PS contribution (so they all need to receive an additional .66% allocation; or in most cases more since PS is on base pay only).

Thanks for your help with all this. I think I have enough to give the client an accurate estimate of what the necessary gateway contribution will be. Next step will be to play around with all the different acceptible forms of comp that I can test on, to see if I can improve the results at all.