Two Plans with different Plan Year Ends

[Dougsbpc]

Suppose you have a professional employer who has sponsored a Defined Benefit Plan and a Profit Sharing Plan, both with a December 31 plan year. There are eight participants in each plan. The arrangement has worked well for over 10 years.

The defined benefit plan is becoming over funded so they should terminate that plan as soon as possible. If the Defined Benefit Plan is terminated effective 7/15/2026 all participants would have accrued a benefit in 2026 because they all would have worked more than 1,000 hours by that time. Not a problem. However, I believe the Defined Benefit Plan would then be deemed to have a short year for 2026 (1/1/2026 - 7/15/2026). Can the 7/15/2026 Defined Benefit Plan be cross-tested with the 12/31/2026 Profit Sharing Plan?

We can wait to terminate the Defined Benefit Plan until year end but then we may be waiting around just watching that plan becoming more and more over-funded.

Unless there is some prohibition in doing so, it should be easy enough to test the 7/15/2026 DB with the 12/31/2026 PSP. After all, participants have already accrued 2026 benefits in the DB plan and most will receive at least 7.5% of salary contributions in the PSP.

Thanks.

 

[Bill Presson]

If it’s overfunded, why is the money still invested?

[bp parv]

Depending on the termination process, the excess assets can fluctuate during the wind-up period. So, before accelerating the termination date solely because of overfunding concerns, I'd want the actuary to quantify how much additional funding surplus is actually expected to arise between July and December 2026. 

[Dougsbpc]

Thank you for your answers.

Does anyone think a 7/15/2026 DB could be tested with a 12/31/2026 PSP?

[Nate S]

On 6/12/2026 at 8:43 PM, Dougsbpc said:

Thank you for your answers.

Does anyone think a 7/15/2026 DB could be tested with a 12/31/2026 PSP?

Yes, when you terminate the DB, do not create a SPY as of 7/15/2026.  The termination itself will create a short limitation year for 401a17 and 415 purposes, but the creation of SPY is a discrete act.  Unless you have a similar fiscal change at the sponsor level, a short plan year should always be avoided.

Regardless, you CAN test two plans with different year end dates, especially if the cross-testing has elements with reliance upon the other plan, such as the gateway; or you can't pass one without the other.  I'm paraphrasing but the guidance is effectively, "you can't test them, unless you have to."

[johncerten]

This is an interesting plan design question. My understanding is that the short plan year created by terminating the Defined Benefit Plan on 7/15/2026 does not automatically prevent it from being tested with the Profit Sharing Plan for the same limitation year, but there are several technical considerations that need to be reviewed.

The key issue is that the plans must be tested under the applicable coverage, nondiscrimination, and cross-testing rules using the correct plan year and benefit allocation data. A DB plan that terminates mid-year will generally have a short plan year, and the testing implications should be carefully coordinated with the PSP’s 12/31/2026 plan year.

I would pay close attention to:

• Whether the DB and PSP are part of a controlled group or otherwise required to be aggregated.

• Whether the plans have historically been tested together and whether the testing method remains consistent.

• How the short-year DB accruals and the PSP allocations are treated for the 2026 nondiscrimination testing.

• Whether the DB termination creates any timing or operational issues under the applicable IRS regulations.

From a practical standpoint, terminating earlier to avoid additional overfunding may make sense, but I would have the TPA/actuary run the 2026 testing scenario before finalizing the termination date. A projection showing the 7/15 DB termination combined with the year-end PSP should confirm whether the intended cross-testing approach produces an acceptable result.

The bigger concern may not be whether the two plans can be tested together, but whether the short-year DB termination affects the assumptions and testing methodology that have been used in prior years.