I am trying to match results for a 401k safe harbor cross-tested plan between Datair and Relius.

I pass the Gateway Test and the Ratio Percentage Test.

On Datair the plan passes the Average Benefits Test by the Equivalent Annual Benefit Basis with PD. Relius also passes this test with the same percentage.

I am having problems with the 401(a)(4) test. When the test is done on Datair it passes based on both the Equivalent Accrual (Acc-to-Date) with and without permitted disparity. It does not pass the rate group based on the Equivalent Accrual (annual) with PD. Relius also does not pass the rate group based on the Equivalent Accrual (annual) with PD. On Relius under the General Nondiscrimination Reports, there is a DB tab which contains the options for the Actuarial Equivalent Accrued-to-Date option. Relius has told us that this option is not available for DC plans. It has been my understanding from conversations with Datair that you only have to pass 1 of the 6 rate group tests to pass the 401(a)(4) test. I am confused as Datair and Relius seem to be contradicting each other.

I'm hoping that someone can clarify what is really needed to pass the 401(a)(4) test.

Thanks,

this might boil down to a misunderstanding on what was said.
you can certainly test using accrued to date on a DC plan, however, that option is not available on the Relius system for DC plans . (It hasn't been programmed) hopefully that is what they meant.

Thank you. Our company just began using Relius this Spring so we are still trying to understand how to test different plan designs.

Tom, didn't you, er, test that system in a prior life?

many moons ago.
to test on accrued to date you would need an average comp. they did just recently add historical comps to the system a year or so ago, so I suppose it would be possible - but in the past that simply didn't exist. you also need historical gains - I doubt many plans on the system have that value. One could, if I understand things correctly, use a fresh start date - but with gateway minimum requirements and stuff, I'm not sure if it is worth the extra time and effort to base stuff on a fresh start date.
If I read the stuff for the latest update they did just program the system to test DB on an allocation basis, and do some combo work with cash balance plans. of course, I think Andy can can do those things in his head, unlike the rest of us!

Interesting. The need to test on average comp (and maintain and calculate it, etc.) is certainly a show stopper for a DC-only system. That puts a plus in the spreadsheet column IMHO.

well, remember you are sort of treating things like a db, therefore you have to use at the minimum a hi 3 average or something like that to determine the e-bar.