401(a)4 Failure

My software is coming up with "401(a)(4) Failed; both the gateway as well as the ABPT passed.

How is this not passing 401(a)(4)

Is the coverage ratio for each rate group equal to at least the midpoint between the safe harbor and unsafe harbor percentages?

C.B. probably hit the nail on  the head. You probably have at least one HCE with a high EBAR and no or very few NHCEs in that HCE's rate group. It's often but not always the youngest HCE. Once you've identified the, you can come up with a solution.

On what system are you running the tests?

The Gateway and Average Benefit Test are just two components of "New Comparability Testing".  Personally, I hate that term since it is somewhat misleading.  "Rate Group Testing" coordinates with the testing concepts found in the posts by CB Zeller and Lou S., and is the term I prefer.  Semantics aside, this type of plan testing is not simply passing the Gateway and the Average Benefits Test.  I explain the testing for these plans when needed as you must first pass the Gateway, if you are using projected benefits in testing.  (Current rates are used in what is sometimes called an "Age Neutral Plans", and that plan is not subject to the Gateway.)  Contrary to one seminar that has since been removed from circulation, giving a 5% of pay benefit to all NCE does not let you do whatever you want.  Next, you must satisfy a special form of coverage testing for each rate group where each HCE represents a rate group.  A plan level (all sources) average benefit test is needed if any of the rate groups fail the special coverage testing; which assume that no group is classified as below the unsafe harbor percentage. Anticipating negative commentary toward my grossly oversimplified explanation, I know that I am leaving a whole lot out, and don't intend for this post to be a comprehensive explanation.  I strongly suggest that to gain the understanding that is required for this type of testing, get yourself a copy of the Coverage and Nondiscrimination Answer Book, which has one of its authors Mr. Thomas E. Poje.  Personally, I have found commentary of Mr. Poje to be a great resource.  His explanations, including guidance found in "his book" to be among the best explanations of this very complicated topic.  (I apologize to the Frank J. Bitzer and Bernadine Topazio for not properly recognizing their contributions to an amazing resource.  You deserve similar recognition, and my failure to properly do so reflects my familiarity with Mr. Poje's writing/posts.)  Anyway, I also recommend that you do a search of the message boards on this testing, and pay special attention to comments by Mr. Poje and several others who have an obvious expertise with this form of testing (BG5150, Bird, John Feldt, CB Zeller, Lou S., jpod, Kevin C. and Mike Preston are a few off the top of my head.)  Good luck!

2 hours ago, Below Ground said:

A plan level (all sources) average benefit test is needed if any of the rate groups fail the special coverage testing; which assume that no group is classified as below the unsafe harbor percentage. 

This is indeed complicated stuff and your post does justice to the complexity. The only part I would tweak is above. The ABT is company wide not plan level. And the threshold is the midpoint not the unsafe harbor. Good job. 

Thanks for the clarification Mike.  As always, your comments are most helpful.

thanks for the compliments, I am a bit humbled by the compliments, Amy Cavanaugh wrote the first edition to the book, and shortly thereafter had me helping on the book for several editions. after she 'retired' from the book Frank and Bernadine joined in, though we each primarily handle certain chapters. Yes, I do the cross testing portions. wonder how you could tell. ha. My hope has always been not just providing the yes or no answer but when possible the 'why'.

but even my time is drawing rapidly to a close. I've resigned from the book early this year. mom is 94 and I'd rather spend the time at home taking care of her. Need to tie a few more things up at work on some plans and then be out the door.

3 minutes ago, Tom Poje said:

Need to tie a few more things up at work on some plans and then be out the door.

You will be missed! 

Yes, you will indeed be missed.  Sorry if I made you uncomfortable, but I think you'd have to look pretty far to find someone in disagreement.

Yes, Tom, you will be missed. Best of luck with your Mom. I hope you may find some time to post some of your humorous anecdotes at times.

Tom,  you will be missed. I hope you can still hang out here some.  Thank you!!

Tom, you do also need to find time to publish a pension song book.

Best wishes to you!