Testing Comp

[Fred Payne]

The Relius recordkeeping software included a participant's full-year comp in the calculation of the Average Benefits Test even though the participant didn't enter the Plan until 7/1. Relius' calculation of this participant's EBAR differs from mine since I had only used comp while a participant (as the Plan docs stipulate) for calculating both the dollar allocation and the EBAR.

Did I err in my calculation of the allocation and EBAR, or do I possibly have a setting incorrect in Relius?

[Tom Poje]

The regs require you to use on of the definitions that satisfy 414(s) - and the typical document will define 414(s) comp as being from date of participation, so that is what you have to use.

If you are processing the plan only once a year on Relius, then the system has to be 'told' what comp to use. there are a couple of ways to do this, otherwise the system will use whatever comp (e.g. ytd comp) that you plug in census.

[Mike Preston]

Just a small clarification. The document definition of compensation (whether it is a 414(s) definition of compensation or not) is not the only compensation that can be used when doing non-discrimination testing. Unless the document specifies all of the non-discrimination methodology (like we used to do back when the IRS was insisting on such language, but we no longer typically do because they aren't), one can test for non-discrimination against any definition of compensation that satisfies 414(s).

I don't use the system in question, but I would be shocked if there wasn't a mechanism for pointing the non-discrimination testing at a definition of compensation that is independent of the other definitions of compensation under the plan.