General Nondiscrimination Test (7.1)

[Guest Jeff Underwood]

I am attempting to perform cross-testing on a plan that defines normal retirement age as the later of age 65 or the 5th anniversary of plan entry. The normal retirement date is the anniversary date coinciding with or following attainment of the above age. Relius appears to be "rounding" the "Tst Ann Age" and the "Tst Age". All participants born during the first six months of the year have values of 65, while all participants born in the second six months of the year have values of 66. The rounding does not appear to apply to "Cur Age".

This yields some interesting results. For example, one participant born in March 1935 shows "Cur Age" of 66, "Nrm Ret Age" of 65 and 67 for "Tst Ann Age" and "Tst Age".

After reading the nondiscrimination thread from February, I changed the plan specs to "Date of Event" to remove the Normal Retirement Date from the equation. I got exactly the same results for the participant mentioned above. It did change the "Tst Ann Age" and "Tst Age" to 65 for participants who had not already exceeded age 65, however.

Has anyone else encountered this problem? If so, does anyone have a workaround?

[Tom Poje]

Jeff -

It sounds like you are saying the problem with people who have not exceeded age 65 is fixed by switching to 'date of event' for normal retirement age.

as for the people who are past age 65 - are you talking people with less than 5 years participation or more than 5 years participation?

If they have more than 5 years participation, and accumulation factor = 1.000, and APR = the value for the persons current age , then I think you are ok, no matter what values show for test age, test ann age - although maybe it matters if you are imputing disparity. If you are easily passing, then it probably doesn't matter, as participants past age 65 generally have small E-Bars. (On the other hand, I am big into calcs, and numbers and stuff, and like things to be as correct as possible.)

[Guest Jeff Underwood]

The person in question has more than five years of participation, so his normal retirement age should equal his current age. The problem seems to be that Relius is rounding his age up to the next year for "Tst Ann Age" and "Tst Age". It is using the APR for the "rounded" age, not his current age. In addition, it is not showing an accumulation factor of 1, it is showing 1.085 for some reason. I am not imputing disparity, so I'm not sure what the problem is. Also, the gentleman in question is an HCE, so it would be nice if I could get an accurate result for him. Any thoughts?

[Tom Poje]

yes, maybe.

The fact the accumulation factor is showing 1.085 and that his age is 1 more than what you expect at least makes sense from a consistency point of view.

I would check to see if

the system says he has 1 year of future service, or

manually change his retirement date. Hopefully that will work.

don't rerun transaction, just manually change and see if that fixes the problem (for now). There may be a bug involved, who knows, but if you can work around maybe that will suffice.

[Guest Jeff Underwood]

He has no future service according to census. I tried changing his retirement date last night, but that doesn't change any of the test results for him. It does "fix" participants younger than 65, however.

[Tom Poje]

in plan specs, (plan entry requirements) try changing

age definition

(e.g. nearest / last)

otherwise I am out of ideas.

[Guest Jeff Underwood]

Unfortunately, the plan specs already have age defined as "last".

[Guest Jeff Underwood]

Corbel issued a "fix" today for certain general non-discrimination problems, but it did not resolve the testing age issue. Below is the workaround that I came up with.

Change normal retirement date to "date of event". This will cure the problem for all participants less than age 65.

For participants age 65 and older, change their date of birth (if necessary) to a date after June 30. You will also need to change the normal retirement date accordingly.