If an employer with a 401(k) plan with a discretionary match decides to make no matching contributions in 1999, does that mean the allowable HCE match limit for 2000 is zero? Any known exceptions to this?

------------------

I think you are correct.

The only "exceptions" I can think of are the first plan year rule (which would only apply in the first year the plan provides matching contributions) and the plan coverage changes rule (which requires you to use a weighted average of the ACPs for the prior year subgroups for certain plan coverage changes that occur during the year).

In the facts you describe, the ACP for HCEs would be zero.

the only other possibilities I could think of would be
1. if Govt extends the deadline for amending plans (we know they have never done that before!)
2. if you have a lot of 'extra' NHCE deferrals in the prior year that you can 'shift' to use in the ACP test

Your question points out that using the prior plan year method for NHCE percentages is a bad idea for plans where the matching amount might decrease.

Just to make one of your assumptions explicit: we're all assuming that there are no employee after-tax contributions to this plan, because they also will affect ACP testing.

[This message has been edited by MWeddell (edited 10-14-1999).]

If the situation does occur so that no HCE match is allowed for 2000, what is the best solution to be sure HCE match is again allowed for 2001?

The best solution based on what we know is to switch to the current year ADP/ACP testing method if it's at all possible.

Is it OK to use prior year testing for ADP and current year testing for ACP and vice versa?

Yes, if you aren't using a protype.
regardless of that, please remember to keep track of whatever method you use.
when you finally amend, your document MUST state testing method used...
e.g. 1997 ADP current ACP current, 1998 ADP Prior ACP current, etc.