The personal dollar limit is not the only constraint
A traditional 401(k) plan can fail its actual deferral percentage test even when every participant stays within the personal elective-deferral limit. The test compares relevant deferral rates for highly compensated and non-highly compensated employees under applicable rules. An owner should therefore not treat the maximum payroll election as a guaranteed amount that will remain in the plan. Depending on the permitted correction method, the employer may make additional contributions for non-highly compensated employees or return excess contributions and allocable earnings to highly compensated participants. An ACP testing failure involving matching or employee after-tax contributions is related but distinct. A valid safe-harbor design may change the testing requirement, but the company's actual plan terms and operation determine that result. The administrator should identify the precise failed test before the owner decides how to respond.
Show the owner refund and employee-funding alternatives
Assume an administrator's completed ADP test determines that a founder must receive a $6,000 corrective distribution plus $240 of allocable earnings under the selected refund method. The founder had contributed below the annual personal limit, so the correction is not proof that payroll exceeded that limit. The $6,240 leaves the plan under the corrective rules and affects personal tax reporting. Suppose an alternative permitted correction would require $18,000 of employer contributions for eligible non-highly compensated employees. The company must compare that real employee expense with the owner's loss of intended retirement funding. These amounts are hypothetical outputs of a test, not a method for calculating ADP results. Ownership attribution, compensation definitions, testing method, participant groups, and earnings rules control the actual figures. The owner cannot choose an arbitrary refund merely to make the test appear to pass.
Deliver a complete census and compensation definition
Provide the administrator all employees, related employers, ownership and family relationships, compensation by source, eligibility dates, hours, deferrals, matches, and after-tax contributions. Highly compensated status is not established solely by a job title or a single current salary amount. Confirm the plan's compensation definition and testing method rather than using financial statement payroll totals. Missing a non-highly compensated employee can materially distort the average. Bonuses and midyear hires should be reflected consistently with plan rules. Review payroll-to-recordkeeper reconciliations so contribution amounts are not missing or duplicated. Preserve the final test report and the input census with the annual plan file. A refund calculated from incomplete data can be wrong even if the arithmetic is correct. The owner should understand what facts changed from the preliminary illustration to the final result.
Meet the applicable correction deadlines
Ask the administrator for the deadline and tax consequences of each correction route. Timing can affect employer excise tax and the available correction process, with different rules for eligible automatic contribution arrangements and other plans. Do not assume the employer's income-tax extension postpones the plan testing correction deadline. A refund of excess contributions under nondiscrimination testing has its own Form 1099-R reporting and generally is not an eligible rollover distribution. It should not be casually rolled into an IRA to preserve tax deferral. Determine the inclusion year and applicable treatment of Roth sources and earnings. If deadlines were missed, evaluate correction under the proper plan compliance program and address associated excise taxes rather than issuing an unlabeled ordinary distribution. Keep authorization, calculation, participant payment, and reporting confirmations together.
Use testing history to choose next year’s design
Compare a traditional tested plan with a supported safe-harbor design, improved employee participation, changed contribution formulas, and total company funding capacity. Employee education may improve participation, but no participation forecast guarantees a testing result. Model the owner contribution after expected corrections rather than showing only the maximum statutory election. Review testing trends before the next plan year's adoption and notice deadlines. The practical result is a deliberate funding choice: accept a supported owner refund, make a permitted employee correction, or adopt a different prospective design. Do not describe a refund as an unexplained provider error when the plan validly failed a required test. Accurate reporting and a clear employee-cost comparison let the owner evaluate the actual retirement benefit retained instead of assuming every dollar deducted from salary is permanently sheltered.
Sources & further reading
Consult the original guidance and its effective dates. These references support the educational discussion; their application depends on the specific facts.