Rate Group Testing

[Fred Payne]

If, for example, a plan has 10 HCEs benefitting, is there automatically 10 Rate Groups?

What if three of the HCEs all have the same Equivalent Benefit Rates (EBRs), are there then only 8 Rate Groups?

The Ratio Percentage Tests and the Average Benefits Ratio Tests of the three separate Rate Groups of the HCEs with the same EBRs will produce identical results. So does it make any real difference (to the client or the IRS) whether results are presented as 8 Rate Groups vs 10 Rate Groups?

I've seen such test results presented one way vs. the other by different TPAs.

[rcline46]

The regs say 1 rate group per HCE, so you have 10 rate groups. The fact that 3 of them are identical and so may be shown as one is a matter of preference. I don't think the IRS would quibble.

[AndyH]

Fred, let's slow down. Three HCEs with identical EBARs for 401(a)(4) do not necessarily have the same Average Benefit Percentages because the latter includes items that the former does not. So, let's not confuse that.

For programming purposes, I would think it would be easier to have one rate group per HCE, because, due to testing methodology, testing one way may produce two identical EBARs for two HCEs, but testing another way will produce different EBARS. This may occur when inputing permitted disparity, for example.

So, I think it would be technically and practically correct to have one rate group per non-excludable HCE.

[Earl]

this may not be different but.... don't the regs say that each HCE defines a rate group? Then you test each rate group for cov & non-disc. So.....

if different HCEs do in fact have the same EBAR, they would all be in the numerator for the ratio percentage test of that rate. (the denominator would count them as individuals for all rates). So, how could you show separately? If 3 of 10 have a BAR of 3.65% that rate group covers 3/10ths of the HCEs. Not 1/10 individually....

For Nondisc. class test they would again be lumped for the fractions at each rate.

For ave benef test they would be totalled individually.

Maybe you are looking at benefiting schedules listing each person that are then extracted for the test results?