Profit-Sharing Plan

A profit-sharing plan is a qualified defined contribution plan under which the employer may determine annually how much, if anything, to contribute, and the plan's own formula then allocates that amount among eligible participants. The IRS describes it as a plan under which "the plan may provide, or the employer may determine, annually, how much will be contributed to the plan (out of profits or otherwise)" — and that parenthetical is the whole story about the name. IRC §401(a)(27)(A) is literally headed "Contributions need not be based on profits." A business with no profits at all can make a profit-sharing contribution, and a profitable one can skip it. The label is a historical artifact.

One scoping point prevents most of the confusion in this area. A cash or deferred arrangement — the salary deferral feature everyone calls a 401(k) — cannot stand alone; it must be part of a profit-sharing, stock bonus, pre-ERISA money purchase, or rural cooperative plan. So virtually every 401(k) plan in the country is, legally, a profit-sharing plan with a 401(k) feature bolted on. This page is about the other half of that structure: the discretionary employer contribution that is not a match and not a deferral. Employee deferrals, matching, and the limits that apply to them belong to 401(k) and employer match. (Older material sometimes calls a self-employed person's version a Keogh plan. That label fell out of use once the tax law stopped distinguishing between corporate and unincorporated plan sponsors; the IRS still prints it in places but notes that, because the law no longer draws the distinction, the term is seldom used.)

The discretion is over the amount, not the allocation. This is the most frequently mangled point about these plans. An employer may decide in February that last year's contribution will be $0, or $200,000 — but it may not decide who gets what. Treasury Regulation §1.401-1(b)(1)(ii) requires the plan to contain a definite predetermined formula for allocating contributions among participants. "The owner decides who deserves a bonus" is not a profit-sharing plan; it is a compensation arrangement that would disqualify one.

Which formula you pick decides whether you test every year. A pro-rata or comp-to-comp allocation gives every participant the same percentage of pay, and a permitted disparity — or "integrated" — formula tilts modestly toward pay above the Social Security wage base to account for the fact that Social Security itself replaces less of a higher earner's income. Both are design-based safe harbors: they are deemed nondiscriminatory, so no annual §401(a)(4) test is required. A new comparability or cross-tested design, which places participants in allocation groups and can give an older owner a much larger percentage of pay than younger staff, is not a safe harbor. It converts each allocation into an equivalent benefit at the plan's retirement age and must pass general testing every single year — so a design that worked last year can fail this year purely because the workforce's ages or salaries shifted. That annual dependence is the real cost of the flexibility, and it is why cross-tested plans need an administrator running the numbers before the contribution is finalised, not after.

Two limits get conflated constantly. The employer's deduction limit under IRC §404(a)(3) is a percentage of the aggregate compensation of all plan participants — a company-level ceiling. The §415(c) annual additions limit is a per-participant ceiling on everything credited to one person's account for the year: deferrals, match, profit sharing, and after-tax contributions combined. Saying "you can contribute up to the annual additions figure as profit sharing" merges an individual cap with a company deduction cap, and the two bind in different situations. Compensation counted for either purpose is capped at $360,000 for 2026, which matters most in owner-heavy plans.

Discretion has an outer limit. Contributions must be "recurring and substantial," and the IRS flags that making no contributions in three of the last five consecutive years may amount to a complete discontinuance of contributions. That is not merely a paperwork issue: a discontinuance triggers 100% vesting for affected participants, which permanently removes the employer's forfeiture leverage. So "we can just skip years" is true in any given year and false as a long-run policy.

Finally, IRC §401(a)(27)(B) requires the plan document to designate the type of plan it is. Whether a plan is a profit-sharing plan or a money purchase pension plan is an election recorded in the document — never something inferred from how the employer has behaved.

The employer decides the total amount (subject to the deduction limit), the plan's formula divides it, and each participant's share is tested against their own annual additions limit. Nobody has to defer anything for a profit-sharing contribution to arrive.

A hypothetical example of a pro-rata allocation. A firm has three employees: the owner at $300,000, and two staff at $60,000 and $40,000, for total plan compensation of $400,000. The employer declares a $40,000 profit-sharing contribution. Pro-rata means everyone receives the same percentage of pay, and that percentage is 40,000 ÷ 400,000 = 10%. So the owner is allocated 10% × $300,000 = $30,000, the first employee 10% × $60,000 = $6,000, and the second 10% × $40,000 = $4,000 — which sums back to the $40,000 declared.

Two checks follow, and they are different checks. The firm's deduction is measured against total participant compensation, a company-level test. The owner's $30,000 is measured against his own annual additions limit for the year, which also has to accommodate whatever he deferred into the 401(k) feature and any match. A cross-tested formula could have shifted a larger share of the same $40,000 toward the owner — but only if the design passed general testing for that specific year. Figures are illustrative.

Pros

• The employer chooses the amount annually, so a bad year costs nothing — unlike a money purchase pension plan's fixed obligation.
• Contributions reach employees who do not defer anything themselves, which also helps with other plan testing.
• Allocation formulas allow real design flexibility, including tilting contributions toward owners nearing retirement.
• It pairs naturally with a 401(k) feature, since almost every 401(k) plan is already structured as one.

Cons

• Employees cannot count on it, which weakens it as a recruiting and retention tool compared with a stated match.
• Cross-tested designs must pass general testing every year, so the contribution cannot be finalised without an administrator's calculation.
• Repeatedly skipping contributions can be treated as discontinuing the plan and force 100% vesting.
• The employer's deduction limit and the per-person annual additions limit are different ceilings, and mixing them up produces excess contributions that have to be corrected.