I have a client which failed the 70% average benefit prong of the general test for the 2003 plan year, which is being calculated on a cross-tested basis. The failure is attributable primarily to one HCE who deferred the $12,000 401k max on comp. of only $33,000 ... which produced a very high EBAR for 2003.
In an effort to salvage the situation, I came across the three-year averaging provision of 1.410(b)-5(e)(5), which seems to permit an EBAR to be determined based on the average of the employee's EBARs for the the current year (2003) and the immediate preceding one or two testing periods (2002 and 2001).
This "problem HCE" had a very low EBAR for 2002 and was not a participant in 2001.
Questions:
1. Can I perform the average benefit prong of the general test for 2003 based on the average of her 2003 EBAR (16%) and her 2002 EBAR (1.5%), which would result in an EBAR of 8.75% (16% +1.5% = 17.5% / 2 = 8.75%)?
2. Can I assign an EBAR to her based on the average of 16%, 1.5% and 0% when, as stated, she was not a participant in 2001?
3. If I calculate this one person's EBAR based on a 2 or 3 period average, am I required to determine all participant EBARs based on an average also? and, if so, how do I do this for those who were not participants during the entire period?
4. Is there any authority and/or guidance in addition to the scant language in the cited Reg?
Thanks for the help.