Hello,

I have a situation where a key employee (who is past NRA) did not work for the year, but also did not retire either. That employee is getting an actuarial increase to her DB benefit, but because of no compensation earned during the year, her HCE rate for the minimum gateway is being computed at 1,000,000% by our system. That is obviously wrong!! But, how should I treat her for the test with a comp of 0, considering she owns the company and is essentially getting a benefit accrual? Some NCEs are getting benefits in both plans.

Your help would be most appreciated!!

Thanks.

If the person has no compensation, one argument is that such employee should not be in the test. "No compensation. Not an employee. Not in the test. Any test" I heard Jim Holland state a few years ago. That is what I have followed. Of course he or someone else may have changed their mind, but I think that approach makes sense.

Thanks! My concern was, even if that person is an owner?

I agree with Andy that Holland has said that, but I would be wary of applying that concept too literally. After all, I could create a non-safe harbor DB plan basing benefits on past Hi-3 compensation. An HCE could conceivably get a very large accrual without any compensation. I wouldn't want to just exclude him from the test in that scenario. I think Holland's comments apply better to DC tests.

For your situation I would recommend testing on average compensation. Then you have compensation to test and you aren't simply ignoring the accrual.

That is a good point! I will look into that. Thanks!

Agreed. Good points as usual by the wet one. I'd still adhere to the rule I cited unless it causes some manipulation of results.

Remember also that if you test accrued to date that only people benefitting sould have NARs in the test. Is this person benefitting? Is there an hour requirement for an accrual? A little messy there it seems to me.

The benefit payable at the new assumed retirement age is equivalent in value to the benefit earned at the old assumed retirement age. Now that the person is assumed to have a new retirement age, you must adjust the prior accrued benefit to its equivalent value, and treat this as the benefit earned as of the beginning of the current year. Since this is the same benefit you already computed in your example, there is no new accrual.

I guess it pays to read the orginal post carefully. I agree if the increase is only the actuarial increase for late retirement, there is no accrual.