HCE Allocations in Failed ACP Test

Although an employer makes matching contributions during the plan year pursuant to the plan document's formula, they only contribute half of the prescribed amount to the HCEs because the plan usually fails the ACP test and it's how the employer attempts to prevent refunds of the HCEs' match.  The ACP test, though, still usually fails, even with such small allocations to the HCEs.  Historically, they have corrected the failed test by making refunds just based on the small HCE matches.  However, the new TPA is suggesting that the correction method must be based on the HCEs receiving enough additional matching allocations to satisfy the plan's formula before calculating the appropriate refunds.  I know it's always a good idea to follow the provisions of the doc, but because this seems somewhat counterintuitive and perhaps may yield different results, I just wanted to double check that what the TPA is saying is the correct way to handle the failed test.  What do you all think?      

You only mention the ACP test.  It's not often that a plan fails the ACP test but passes the ADP test.  Are there also ADP refunds, or is there something funky about the match formula?

You are correct to note that under-funding the match for a select group of individuals is not following the plan document.  It's important to keep in mind that, even though a plan seemingly indiscriminately discriminate against HCEs, HCEs are participants who rights must be protected.

Keep in mind that the amount of refunds to be made from the plan is calculated based on each individual's deferrals and match,  but the actual refunds are determined by starting with the highest percentages being refunded first.  This second step can shift the refund amount from one HCE to another.

This likely is the reason for the TPA's position.

Thank you for your response, Paul.  The ADP test passes because a good number of HCEs have catch-up contributions.