BRF correction?

[Tinman]

Plan has discretionary match. Formula being used is dollar-for-dollar up to a max $40/week, based on hours worked but capped for deferrals made. Examples:

Joe works 36 hours and contributes $50 in deferrals. He receives a match of $36.

Bob works 15 hours and contributes $10 in deferrals. He received a match of $10.

Don works 45 hours and contributes $100 in deferrals. He received a match of $40.

Performing a BRF test on this match on a weekly basis - some weeks pass, some weeks do not. Document is silent as to how to correct this - would it be as simple as providing additional match to those who did not receive as much as the HCE in the same group? Or could we look at this differently and do it on an average basis instead of looking at each week individually? (HCE A averages 32 hours/week for the year - they compare that to the weekly average for the NHCEs)

Opinions, please!

[K2retire]

Why are you testing on a weekly basis rather than an annual basis?

[Tinman]

The match is calculated based on hours worked weekly - so we were looking at each employee, each week in order to perform the BRF testing.