16.6 Permissive Disaggregation & Restructuring to Pass the ADP and ACP Tests

Fix the Population Before You Fix the Numbers

Candidates lose points on ADP/ACP questions not because the percentage arithmetic is hard but because they run it on the wrong population. Before a single ratio is calculated, three questions must be answered:

1. What must be disaggregated? (mandatory)
2. What may be disaggregated? (permissive)
3. What must be aggregated? (mandatory, and constrained by the coverage election)

Mandatory Disaggregation

These separations are not optional. Under Treas. Reg. §1.410(b)-7(c), a single plan document is treated as several separate plans for testing:

Also structurally separate, though for a different reason: the §401(k) deferral portion, the §401(m) match and after-tax portion, and the §401(a) profit-sharing portion of a plan are each tested under their own statute. That is why a single plan generates an ADP test, an ACP test, and a coverage test on the nonelective source.

Permissive Disaggregation: Otherwise Excludable Employees

The election under IRC §410(b)(4)(B), applied to ADP/ACP by Treas. Reg. §1.401(k)-1(b)(4)(iv), splits the population into:

• Statutory group — employees who have attained age 21 and completed one year of service;
• Otherwise excludable group — everyone else who is eligible under the plan's more generous terms.

Each group runs its own ADP and ACP test.

Why it usually helps: newly eligible short-service employees defer at very low rates, and every eligible non-deferrer enters the calculation at 0.00%, dragging the NHCE average down. Removing them from the statutory group raises the NHCE ADP and therefore raises the ceiling on the HCE ADP.

Worked example. Alder Creek Bakery has three-month eligibility. Its 2026 raw results:

Population	Count	ADP
HCEs (all long-service)	4	7.80%
NHCEs, combined	90	4.10%

Combined test: maximum HCE ADP = 4.10% + 2.00% = 6.10%. HCE ADP of 7.80% fails by 1.70 points.

Splitting the NHCE population: 58 have a year of service and average 5.65%; 32 have under a year and average 1.29%.

Statutory group: HCE ADP 7.80%, NHCE ADP 5.65%. Maximum permitted HCE = 5.65% + 2.00% = 7.65%. Still fails — but by only 0.15 points instead of 1.70, cutting the refund by roughly 90%.

Otherwise excludable group: no HCEs are eligible in that group, so no ADP test is required for it.

Had the HCE ADP been 7.40% rather than 7.80%, disaggregation alone would have converted the failure into a clean pass.

Where it backfires: if an HCE falls into the otherwise excludable group — a newly hired executive, or an owner's child who is an HCE by attribution — that group must run its own ADP test with an HCE against a tiny, low-deferring NHCE population. That test almost always fails. Run it both ways before electing.

Mandatory Aggregation

Two aggregation rules constrain the disaggregation choices above.

1. Plans Aggregated for Coverage Must Be Aggregated for Testing

If the employer permissively aggregates two plans under Treas. Reg. §1.410(b)-7(d) to pass §410(b), those plans must be treated as a single plan for ADP and ACP purposes as well. An employer cannot aggregate to solve coverage and then disaggregate to solve testing. Aggregated plans must also share the same plan year and the same testing method (both prior-year or both current-year).

2. An HCE Eligible Under Two Plans

Under Treas. Reg. §1.401(k)-2(a)(3)(ii), if an HCE is eligible to defer under more than one cash or deferred arrangement of the same employer, all of that HCE's deferrals under all such arrangements are aggregated into a single actual deferral ratio, determined using the HCE's compensation from the employer as a whole.

Worked example. Delia is an HCE eligible under both the Parent Co. 401(k) and the Subsidiary Co. 401(k) — the same controlled group. Compensation from Parent is $180,000 with $14,000 deferred; compensation from Subsidiary is $120,000 with $4,000 deferred.

Delia's single ADR = ($14,000 + $4,000) ÷ ($180,000 + $120,000) = $18,000 ÷ $300,000 = 6.00%.

That 6.00% is the figure used for Delia in both plans' ADP tests. It is not $14,000 ÷ $180,000 = 7.78% in one and $4,000 ÷ $120,000 = 3.33% in the other. The exception: plans that may not be permissively aggregated — because they have different plan years, for instance — do not aggregate the HCE's ratio.

Restructuring Into Component Plans

Treas. Reg. §1.401(a)(4)-9(c) permits a plan to be restructured into component plans, each tested separately for §401(a)(4) and §410(b). It is a real technique for general nondiscrimination testing — but note the limit that the exam tests: restructuring is not available for the ADP and ACP tests. Those tests are run on the plan as determined under the §410(b) disaggregation and aggregation rules described above, not on self-defined components.

---

Common ASPPA QKA Exam Traps

• Trap 1 — Aggregating for coverage but disaggregating for ADP/ACP. The two must be consistent.
• Trap 2 — Computing a separate deferral ratio for each plan for an HCE in two plans. One combined ratio, one combined compensation figure.
• Trap 3 — Using restructuring to pass ADP/ACP. Restructuring is a §401(a)(4) technique, not an ADP/ACP one.
• Trap 4 — Treating union disaggregation as elective. It is mandatory.
• Trap 5 — Assuming otherwise-excludable disaggregation always helps. An HCE in the excludable group usually creates a failure.
• Trap 6 — Aggregating plans with different plan years or different testing methods. Both must match.