Phantom Stock for Key Employees: Keep Your Best People Without Giving Up Equity

The usual trigger is a conversation the owner didn't plan for. Your operations director, the one who actually runs delivery, asks for 10% of the company. You don't want a minority partner. You also don't want to lose the person most responsible for whether a buyer pays five times earnings or three. Phantom stock is often the right answer to that conversation. It's rarely the right first answer to the cash question underneath it.

A phantom stock plan is a written promise to pay a key employee cash equal to the value of a set number of notional shares, or just the growth in that value, when a defined event happens, usually a sale of the company. No real shares are issued, so the employee gets no vote, no distributions and no K-1. Because the payment is deferred compensation, the plan has to be built around IRC Section 409A from day one. Get that wrong and the employee, not you, pays the penalty.

What a phantom unit actually promises

There are two designs, and the cost difference between them is large.

Appreciation-only units behave like a stock appreciation right settled in cash. For a key employee joining or stepping up now, they're usually the better fit. The owner keeps the value that already exists, and the employee shares in what they help add.

Section 409A decides whether the plan works

Section 409A governs nonqualified deferred compensation, and phantom stock that pays in a later year is squarely in it. The IRS's final regulations, published in T.D. 9321, make the point that 409A's application isn't affected by the medium of payment. Paying in cash instead of stock doesn't get you out.

The statute, 26 U.S.C. § 409A, allows deferred amounts to be paid only on these events:

1. Separation from service
2. Disability
3. Death
4. A specified time or fixed schedule set when the compensation is deferred
5. A change in ownership or effective control of the company, or of a substantial portion of its assets, as defined by Treasury
6. An unforeseeable emergency

It also bars accelerating the payment schedule except as the regulations permit.

If a plan fails these rules in form or in operation, all compensation deferred under it becomes taxable to the employee once it's no longer subject to a substantial risk of forfeiture. The tax is then increased by interest at the underpayment rate plus 1 percentage point, and by an additional tax equal to 20% of the amount included. That's an expensive way to reward someone you were trying to keep.

The short-term deferral exception

Not every bonus is deferred compensation. Under the short-term deferral rule in Treas. Reg. § 1.409A-1(b)(4), a payment isn't a deferral if it's actually paid by the 15th day of the third month after the end of the employee's or the company's tax year (whichever is later) in which the employee's right to it stops being subject to a substantial risk of forfeiture.

Some phantom plans are designed to fit that exception: the unit vests only when a sale closes, and it's paid at closing. Others pay on a 409A-compliant event like a change in control. Either can work. A plan that pays "whenever the board decides" usually works for neither. Have employee-benefits counsel draft the document. This is a design article, not a legal opinion.

A phantom plan is only as fair as the value number underneath it. If your adjusted EBITDA hasn't been tested, run a Profit & Tax Leak Check on it before that number goes into a plan document.

Write the value formula into the plan

The fight you most want to avoid is over value. Remove the argument by defining it.

Equity value      = (Adjusted EBITDA × agreed multiple) − net debt
Unit value        = Equity value ÷ total notional units
Appreciation pay  = (Unit value at event − grant unit value) × units held
At a sale:        Equity value = actual net proceeds to equity, as defined in the plan

Three choices to make on paper:

• The multiple. Use a conservative, fixed multiple for interim valuations, and actual sale proceeds at a sale. Don't let the interim number become an implied promise of the exit price.
• The EBITDA definition. Spell out the owner-compensation normalization and one-time items. If the number would change after a buyer's quality-of-earnings review, it'll be contested.
• Net debt. Debt the owner takes on to fund distributions shouldn't cut the employee's payout, and debt used to buy a competitor arguably should. Decide now.

Worked example: a hypothetical $8M service firm

All figures are illustrative. Revenue is $8.0M, adjusted EBITDA is $1.6M, and the plan uses a 5.0x interim multiple with no net debt. The company defines 1,000,000 notional units.

The operations director receives 50,000 appreciation-only units, 5% of the notional total, at the $8.00 grant value.

• Appreciation payout at sale: ($13.20 − $8.00) × 50,000 = $260,000
• If they'd been full-value units: $13.20 × 50,000 = $660,000
• If you'd given 5% real equity: about $660,000 at sale, plus four years of pro rata distributions, voting rights, and a minority holder at the closing table.

The appreciation design costs the owner $400,000 less than either alternative at this exit, and it pays nothing if value doesn't grow. For the employee, $260,000 is meaningful only if they believe the EBITDA growth is achievable. Show them the plan to get there.

The payout has to be funded

At a sale, the payment usually comes out of the proceeds at closing. Buyers generally treat it as a transaction cost, so it reduces what the owner nets, and it belongs in the exit planning model years ahead of the sale.

A fixed-date plan is harder. A plan that pays out in year five whether or not there's a sale is a cash obligation the business has to fund from operations. Model it in the long-range forecast, and put a cap on any single year's payment, or allow installments set at the time of deferral, so the obligation can't collide with a weak year.

Separation payouts are the riskiest. If a vested employee can leave and trigger payment, the company may owe cash at exactly the moment it lost the person producing it. The safer default is to pay only at a change in control and forfeit unvested units on resignation.

Vesting, forfeiture, and the retention point

The plan exists to keep someone through a sale. Build it that way:

• Vest over three to five years, with the remaining units vesting at a change in control.
• Forfeit units on voluntary resignation or termination for cause before the event.
• Require the employee to stay through a transition period after closing, typically 6–12 months, for the final tranche.
• Pair with confidentiality and non-solicitation terms where your state allows them.

A buyer pays more for a business that isn't dependent on the owner. A phantom plan that keeps your second-in-command through the sale is one of the most direct ways to reduce the key-person discount on your exit price.

Phantom stock against the alternatives

For annual performance, use profit sharing. For the exit, use phantom stock. Plenty of companies need both.

Before you say yes

Calculate what the phantom grant costs at three exit values: today's formula value, a realistic exit, and a strong one. If the realistic payout doesn't change the employee's life, the grant is too small to retain them. If the strong payout would hurt your own retirement number, it's too large. Build those three cases into the long-range model with fractional CFO support before counsel drafts anything. If the EBITDA you'd anchor the formula to hasn't been pressure-tested, a Leak Check is the fastest way to find out whether it will hold up when a buyer looks.