401(k) compliance testing, explained
Annual testing exists to stop a plan from working mainly for the people who run the business. Most small employers meet it as a surprise refund in March. It is more manageable than that, and some of it is avoidable by design.
A 401(k) faces three annual tests. ADP and ACP check that deferrals and matching contributions do not favour highly compensated employees. The top-heavy test checks whether key employees hold more than 60% of plan assets.
Testing is the part of a 401(k) that employers hear about only when it fails, usually as a refund cheque to an owner in the first quarter. Understanding which test does what turns that from an annual surprise into a design decision made once.
ADP and ACP: are deferrals lopsided?
The Actual Deferral Percentage and Actual Contribution Percentage tests exist to check that deferred wages and employer matching contributions do not discriminate in favour of highly compensated employees. ADP looks at employee salary deferrals; ACP looks at matching and after-tax contributions.
The mechanic is a comparison of averages. If the highly compensated group's average deferral rate runs too far ahead of everyone else's, the plan fails, and the usual correction is refunding excess deferrals to the highly compensated employees — which is why the failure lands on the owner rather than on the workforce.
The uncomfortable implication is that these tests can only be passed by rank-and-file participation. An owner cannot fix an ADP test failure by contributing less next year and hoping; the plan's result depends on how much everyone else saves. That is the whole reason automatic enrolment and safe harbor designs exist — both raise or bypass the dependency.
Top-heavy: who holds the balances?
A plan is top-heavy if the account balances of key employees exceed 60% of the account balances of all employees. The ratio is tested every year, based on account balances on the last day of the prior plan year.
If a plan is top-heavy, the employer must generally make a minimum contribution of 3% of total compensation for the year to non-key employees. Two details matter. The minimum is based on total compensation for the year, not only on compensation during the period of plan participation. And if the highest contribution percentage for any key employee is less than 3%, non-key employees receive that lower percentage instead.
Top-heavy status is common in small businesses and is not a sign of anything wrong. A five-person company where the two owners have been saving for a decade will very likely be top-heavy on arithmetic alone. It is a cost to plan for, not a failure to avoid.
The three tests side by side
| Test | What it checks | What removes or reduces it |
|---|---|---|
| ADP test | Whether the average salary deferral rate of highly compensated employees runs too far ahead of everyone else's. | A safe harbor design removes it. Automatic enrolment raises rank-and-file participation, which makes passing easier without removing the test. |
| ACP test | The same comparison, applied to employer matching and after-tax contributions. | A safe harbor matching design is deemed to satisfy it, where the safe harbor match is the only match made. |
| Top-heavy test | Whether key employees hold more than 60% of plan assets, measured on the prior plan year's final balances. | Not removable by design in the same way. Some safe harbor designs satisfy the top-heavy minimum through the contribution they already require. Worth confirming for your specific design. |
Percentages here are statutory plan-design figures — the top-heavy threshold and the minimum contribution rate. They are not fees.
Why safe harbor is the usual answer
A safe harbor 401(k) removes the ADP test, and a safe harbor matching design is deemed to satisfy ACP where no other match is made. It does that by committing the employer to a contribution that is fully vested when made — the trade is a guaranteed cost in exchange for a removed uncertainty.
For an employer whose testing is marginal, that trade is usually worth taking, because a failed test is not free either: refunds create taxable income for the people receiving them, corrections take professional time, and the outcome is not known until after the year has closed. A safe harbor commitment is known in advance and can be budgeted.
The detail of the formulas, the notice window and what safe harbor actually buys is covered in the safe harbor guide below.
What to do about it before the year ends
Testing is retrospective, but the levers are not. Three things are worth checking during the year rather than after it.
- Participation among non-highly compensated employees. This is the variable that decides the ADP test, and it is the one most improved by automatic enrolment.
- Whether the plan is heading for top-heavy status. The ratio uses the prior year's closing balances, so you can see it coming a full year ahead.
- Whether the plan's compensation definition is being applied correctly. Testing runs on data from payroll, and a wrong compensation figure produces a wrong test result that looks perfectly plausible.
In a pooled employer plan the testing work sits with the pooled plan provider as plan administrator, but the inputs still come from the employer's payroll — which is the same boundary that applies everywhere else in plan administration.
Common questions
What is the ADP test in a 401(k)?
The Actual Deferral Percentage test checks that employee salary deferrals do not discriminate in favour of highly compensated employees, by comparing the average deferral rate of that group against everyone else's. If the gap is too wide the plan fails, and the usual correction is refunding excess deferrals to the highly compensated employees.
What is the difference between the ADP and ACP tests?
ADP looks at employee salary deferrals. ACP applies the same kind of comparison to employer matching contributions and after-tax contributions. Both exist to confirm that deferred wages and matching contributions do not favour highly compensated employees.
What makes a 401(k) plan top-heavy?
A plan is top-heavy if the account balances of key employees exceed 60% of the account balances of all employees. The ratio is tested annually using balances on the last day of the prior plan year, which means an employer can see it coming a year ahead.
What does a top-heavy plan have to contribute?
Generally a minimum of 3% of total compensation for the year to non-key employees — based on total compensation for the year rather than only compensation while participating. If the highest contribution percentage for any key employee is below 3%, non-key employees receive that lower percentage instead.
Can a small business avoid 401(k) testing?
A safe harbor design removes the ADP test, and a safe harbor match is deemed to satisfy ACP where it is the only match made. Top-heavy is different — some safe harbor designs satisfy its minimum through the contribution they already require, but that should be confirmed for your specific design rather than assumed.
Sources
- Internal Revenue Service, 401(k) Plan Overview — that the ADP and ACP tests verify deferred wages and employer matching contributions do not discriminate in favour of highly compensated employees, the 60% top-heavy definition, and that safe harbor contributions are fully vested when made
- Internal Revenue Service, Is my 401(k) top-heavy? — the 60% key-employee threshold, annual testing on the prior plan year's final balances, the 3% minimum contribution to non-key employees on total compensation for the year, and the lower-percentage exception
- Internal Revenue Service, 401(k) Plan Fix-It Guide — The plan was top-heavy and required minimum contributions were not made — the consequences of missing a required top-heavy minimum contribution
- Internal Revenue Service, 401(k) Plan Qualification Requirements — that contributions or benefits must not discriminate in favour of highly compensated employees, and that plans may satisfy this with safe harbor contributions or the ADP and ACP tests
aipep401k.com is a pooled employer plan for small business 401(k) plans, operated by ThinkNirvana Foundation — plan administration handled by automation rather than assembled by hand.
Related: what a safe harbor design commits you to, the payroll data testing actually runs on. To see what a plan design would look like for your own headcount, the plan design tool is free and needs no sign-up.