An add-back is a cost you take out of reported earnings to show what the business will earn for its next owner. Buyers price off the result, which makes the arithmetic simple and unforgiving: at a 7.3x multiple, every dollar of add-back you defend is worth $7.30 of enterprise value, and every dollar a diligence team strikes costs you the same $7.30. A $400,000 schedule that survives intact is $2,920,000 of price. The same schedule with a quarter of it struck is $2,190,000.
Which of those you get is decided by documentation, not by argument. The add-backs that survive are the ones a stranger can verify without taking your word for anything. The ones that die are the ones that need a story.
The multiple that turns add-backs into money
GF Data’s contributed set put average valuations at 7.3x trailing twelve-month adjusted EBITDA across 80 completed transactions in the first quarter of 2026. Note which EBITDA that is. The reported number is not what gets multiplied; the number that survives diligence is. Sector matters too - the ranges in EBITDA multiples by industry and the business valuation calculator will give you a starting figure to run your own schedule against.
The SEC drew the line first, and it is the clearest one available
Public companies have been fighting about this in writing for twenty years, and the staff guidance that came out of it is the best free definition a private seller can get.
Start with the name. Under the SEC’s non-GAAP interpretations, measures “calculated differently than those described as EBIT and EBITDA in Exchange Act Release No. 47226 ... should not be characterized as ‘EBIT’ or ‘EBITDA,’ and their titles should be distinguished from ‘EBIT’ or ‘EBITDA,’ such as ‘Adjusted EBITDA’”. The moment you add anything back, you are no longer talking about EBITDA. That is not pedantry; it is the first thing a buyer’s accountant re-labels in the model.
Then the substance. The staff treats a measure that “excludes normal, recurring, cash operating expenses necessary to operate a registrant’s business” as one that could be misleading - and it defines recurring broadly: an operating expense “that occurs repeatedly or occasionally, including at irregular intervals” is recurring.
Read that last clause again, because it is the one that kills schedules. “Non-recurring” does not mean “unpredictable” or “not every year.” A legal settlement in 2023 and a different legal settlement in 2026 is a company that settles litigation. An equipment failure every third year is a maintenance capital expenditure with bad timing. Both get added back constantly in sale memoranda, and both come straight back out.
Three add-backs the SEC staff has challenged, on the record
Comment letters between the SEC staff and public filers are published on EDGAR, and they are the closest thing this market has to case law on add-backs. Three are worth knowing, because the same categories turn up in lower-middle-market schedules constantly. Watch what actually happened in each one: the staff’s objection is the useful part, and it is not the same thing as the filer conceding.
Sales commissions. The staff asked Bill.com how it had considered C&DI 100.01 given that “sales commissions and service costs appear to be normal, recurring cash operating expenses necessary to operate your business”. The company’s February 2021 reply agreed it would “revise its calculation method to remove such adjustments from its calculation of non-GAAP metrics in its future periodic reports and earnings releases.” The read-across: the cost of winning revenue is a cost of revenue, whatever line it sits on.
Acquired in-process R&D. In February 2022 the staff questioned Cardiovascular Systems’ adjustment for an IPR&D charge arising from an asset acquisition, on the grounds that the expenditure “appears to be similar to research and development expenses and normal, recurring, cash operating expenses necessary to operate your business”. The company did not concede. It replied that acquired IPR&D charges “are not normal, recurring, cash operating expenses in the Company’s business,” citing the nature and infrequency of a $3.3 million purchase price charged entirely to expense. Buying capability rather than building it is arguable - which is the point. It gets argued.
Recruiting, relocation and sign-on bonuses. In August 2024 the staff told BRC Inc. that its Adjusted EBITDA “includes several adjustments that appear to relate to normal operating expenses necessary to operate your business”, listing executive recruitment, relocation and sign-on bonuses, strategic-initiative costs, ordinary legal costs, and start-up and production costs on its ready-to-drink line. The company defended all of them as appropriate under C&DI 100.01 and 100.04.
What matters to a private seller is not who won. It is which categories a trained reader flags on sight - staffing, growth spending, ordinary legal costs, and anything bought rather than built - and that two of these three filers, with counsel and an audit committee behind them, still had to spend a letter defending them. Your buyer’s diligence team reads the same way. You will not get a letter; you will get a lower number.
What lenders allow, in writing: 15 to 30 percent
Credit agreements are public record, and they are the only place you can read exactly how much adjustment a professional lender will tolerate before it stops believing the number.
In MasterBrand’s March 2026 amendment to its credit agreement, run-rate cost savings and synergies can be added to Consolidated EBITDA only if they are “reasonably expected and factually supportable”, only for actions “taken or committed to be taken within eighteen (18) months,” and only up to “15% of Consolidated EBITDA for such period (calculated before giving effect to such add-backs)”. Earn-out and contingent consideration add-backs in the same definition are capped at 10 percent.
Viasat’s January 2026 credit agreement sets a combined ceiling: unusual, extraordinary or non-recurring losses, together with pro forma run-rate cost savings and synergies, “shall be limited to 20.0% of EBITDA for such period”, again measured before the add-backs themselves.
Gen Digital’s March 2026 amendment runs looser, and the way it runs looser is instructive. Its run-rate clause requires the savings to be “reasonably identifiable and factually supportable” and projected from actions “taken, committed to be taken or with respect to which substantial steps have been taken,” within 24 months rather than eighteen. Standalone cost-saving add-backs then “shall not exceed 30% of Consolidated EBITDA for such period (calculated after giving effect to any such add backs and all other add backs for such period)”. Read the parenthesis: measuring a 30 percent cap after the add-backs is a materially bigger allowance than measuring a 20 percent cap before them. The headline percentage is never the whole term.
The clause immediately after it is the one that should change how you prepare. It admits adjustments “evidenced by or contained in a due diligence quality of earnings report” prepared by a “big four” firm or another firm the agent finds acceptable, and “consistent with Regulation S-X.” The lender does not ask the borrower to argue for its add-backs. It names the document that counts as evidence - which is the same thing the table below says in longhand.
Three agreements, three ceilings, one shape: a percentage cap, a factual-support test and a clock of eighteen to twenty-four months. That is the market’s own working view of how much of an earnings figure can be adjustment before the figure stops being useful - and the looser the cap, the more the lender leans on the evidence requirement instead. It matters to a private seller for a direct reason: the bank financing your buyer’s purchase price is applying the same discipline to your schedule. Push past what it will underwrite and the debt cannot be sized on the number you are asking the buyer to pay. The buyer does not argue about it. The buyer lowers the price, or moves the contested amount into an earnout and makes you prove it.
There is evidence about what happens when those ceilings are set too high. A Federal Reserve Bank of St. Louis working paper, “EBITDA Add-backs in Debt Contracting: A Step Too Far?” (Faria-e-Castro, Gopalan, Pal, Sánchez and Yerramilli, September 2022), used the 2013 Interagency Guidance on Leveraged Lending as a shock that pushed non-GAAP add-backs up, and found that the add-backs raise the likelihood of loan delinquency and default and of the borrower being downgraded - and that lead arrangers become more likely to sell their share down in the secondary market. That is the syndicated loan market rather than a $5M-EBITDA sale, so read it as direction and not as your number. The direction is that the people who lend against adjusted EBITDA have learned what a padded schedule is worth.
The add-backs that survive
| Add-back | What has to exist before you claim it |
|---|---|
| Owner compensation above market | A compensation study or a real quote for a replacement general manager. Claim both a $600,000 salary add-back and a twenty-hour week and you have told the buyer the business needs a manager it does not have. |
| Related-party rent above market | The lease, and a broker opinion or appraisal of market rent for the same building. The add-back is the spread, not the rent. |
| Family members on payroll who do not work in the business | Payroll records and a job description that nobody fills. The role has to disappear on closing day without anything else changing. |
| One-time professional fees | Engagement letters and invoices tied to a specific matter that has ended - a financing, an acquisition that did or did not happen, a single piece of litigation. |
| Discontinued product line or closed location | A segment-level P&L showing the costs leave with the line. If overhead was allocated to it, that overhead stays. |
| Personal expenses run through the company | Invoice-level detail, coded. Vehicles, travel, club memberships. Expect every item to be read, and expect the ones without receipts to come out. |
One line on that list runs the other way. If the related-party rent is below market - the common case where an owner charges their own company a token rent for a building they also own - the adjustment is negative: occupancy gets restated upward to market before any multiple is applied. Sellers who own their buildings are often surprised by this, and it is not a buyer being difficult. The next owner will pay market rent or buy the building, and the earnings have to reflect whichever it is.
The add-backs that get struck
- Savings from actions you have not taken. The route consolidation you are planning, the two positions you intend not to refill. Lenders will fund these only when they are factually supportable and scheduled within eighteen months, and a buyer applies a harder version of the same test, because the actions are now its risk and not yours.
- “Non-recurring” items that recur. See the SEC’s own standard above. If the same category appears in two of the last five years, it is a cost of doing business.
- Growth spending relabelled as one-time. The three salespeople hired last year are not a one-time expense; they are how the revenue you are also claiming gets produced.
- Lost-customer normalization. “EBITDA would have been $800,000 higher if we had kept the contract” is a statement about a company that no longer exists. Buyers price the one that does.
- Unsupported pandemic- or tariff-era adjustments. Normalizing a disrupted year is legitimate and routine; doing it from memory, four years later, without contemporaneous records, is not.
- Deferred maintenance and skipped capital expenditure. Adding back the repairs you did not make is an argument that the business earned more than it did, and it tells the buyer their own capex forecast is too low. It is the one add-back that costs you twice.
- Anything that double-counts - the owner’s salary and the management fee paid to the owner’s other entity for the same work. That is the fastest way to have every other line re-read.
A worked schedule, on an invented company
Take a company with $5,000,000 of reported EBITDA and no real-world counterpart, and a schedule its owner considers conservative.
| Line | Claimed | Survives diligence |
|---|---|---|
| Reported EBITDA | $5,000,000 | $5,000,000 |
| Owner compensation above a replacement GM’s market rate | $450,000 | $450,000 |
| Related-party rent above a broker’s market opinion | $120,000 | $120,000 |
| ERP implementation | $180,000 | $90,000 |
| Settlement of a single employment claim | $95,000 | $95,000 |
| Owner’s vehicles, travel and club dues | $85,000 | $85,000 |
| Run-rate savings from a planned route consolidation | $300,000 | $0 |
| Adjusted EBITDA | $6,230,000 | $5,840,000 |
| At 7.3x | $45,479,000 | $42,632,000 |
Two lines moved and the price moved $2,847,000. The route consolidation went to zero because nothing had happened yet - no notices, no contracts cancelled, no board minute. Half the ERP cost came back in because the licence and support fees continue after the implementation ends, which makes that half a normal, recurring operating expense in the SEC’s sense and in a buyer’s.
Notice the other number in that table. The claimed schedule totals $1,230,000, or 24.6 percent of reported EBITDA - above two of the three lender caps quoted above. The schedule that survived is 16.8 percent, which is inside all of them. A seller who had checked that ratio first would have known which line was going to be the fight.
The number you win on is the number you then have to live with
An aggressive add-back that somehow survives does not stop working once the price is agreed. Adjusted EBITDA is usually the definition underneath everything that comes after it.
It sets the covenant. Leverage tests in the acquisition debt are measured against the same adjusted figure, and as the arithmetic of a dividend recapitalization shows, a covenant computed off an inflated number leaves a company with less headroom than its owner thinks. It sets the earnout threshold, if there is one - winning an argument about a $300,000 add-back and then being handed an earnout target that assumes the $300,000 is real is a poor trade. And it sets the entry price on rollover equity, which means an inflated number today is equity you bought expensively in the company you still own a piece of.
Build the schedule before someone builds it for you
A sell-side quality of earnings report is the mechanism for doing this on your own clock. GF Data looked at 360 transactions completed since the third quarter of 2024 and found that sellers who used a sell-side QoE saw average TEV/EBITDA multiples of 7.4x against 7.0x for those who did not, with the benefit most notable on deals above $50 million of enterprise value and little valuation boost on smaller ones. In the same reporting, a capital-advisory practitioner estimates sell-side QoE use at at least 90 percent of the private-equity-backed deals he sees, against about 50 percent among lower-middle-market founder-led businesses — his estimate from practice rather than a measured figure.
That gap is the opportunity. The institutional seller arrives with its adjustments already tested and negotiates; the founder arrives with a spreadsheet built by the person with the most to gain from it, and defends. How PE firms value a company shows where the schedule sits in the wider pricing process, and exit planning for business owners puts it where it belongs on the calendar - a year or two before the data room opens, not during it.
Frequently asked questions
What is a normal level of add-backs? There is no public dataset of private-company add-back schedules, so anyone quoting a precise average is guessing. The closest honest benchmark is what lenders write into their own documents: caps of 15 to 30 percent of EBITDA, and the tighter ones are measured before the add-backs rather than after. A schedule above the band is not automatically wrong, but it will be read line by line, and the looser the cap the more the lender wants a quality of earnings report behind it.
Are adjusted EBITDA and seller’s discretionary earnings the same thing? No. Seller’s discretionary earnings adds the owner’s entire compensation back, on the theory that a buyer-operator will take the owner’s place. Adjusted EBITDA adds back only the amount above what a hired manager would cost, because the buyer is paying for a business that runs without its owner. At $3M+ of EBITDA you are in the second world, and the two numbers are not comparable - neither are the multiples applied to them.
Do add-backs affect the multiple as well as the number? Yes, and this is the part sellers underestimate. A schedule that reads as carefully built invites a negotiation about price. A schedule with three obviously soft lines invites a reappraisal of everything else in the data room, and the discount shows up in the multiple and in the structure - more escrow, more earnout, more holdback - rather than in a line anyone names out loud.
Sources
- U.S. Securities and Exchange Commission, Compliance and Disclosure Interpretations: Non-GAAP Financial Measures (Questions 100.01 and 103.01)
- MasterBrand, Inc., Second Amendment to Amended and Restated Credit Agreement dated March 26, 2026 (Exhibit 10.1), definition of “Consolidated EBITDA”
- Viasat, Inc., Credit Agreement dated January 21, 2026 (Exhibit 10.1), definition of “EBITDA”
- Gen Digital Inc., Third Amendment to Amended and Restated Credit Agreement dated March 27, 2026 (Exhibit 10.36), definition of “Consolidated EBITDA”
- Faria-e-Castro, Gopalan, Pal, Sánchez and Yerramilli, “EBITDA Add-backs in Debt Contracting: A Step Too Far?”, Federal Reserve Bank of St. Louis Working Paper 2022-029
- Bill.com Holdings, Inc., response to SEC staff comments, 25 February 2021 (non-GAAP adjustments for sales commissions and service costs)
- Cardiovascular Systems, Inc., response to SEC staff comments, 17 February 2022 (acquired IPR&D adjustment)
- BRC Inc., response to SEC staff comments, 16 August 2024 (recruiting, relocation, legal and start-up adjustments)
- Association for Corporate Growth, “GF Data Reports Show Stronger Financing Markets Fueled M&A Gains in Q1”
- Middle Market Growth (ACG), GF Data on quality of earnings reports
The bottom line
Add-backs are the cheapest value in a sale process and the easiest to lose. Nobody raises your multiple for being conservative, but the seller who can hand over an engagement letter, a lease, a comp study and a segment P&L for every line keeps the number, and keeps it without spending the negotiation defending the parts that were never going to hold.
The same schedule does the same work in an Independent Buyout: the financing is sized off the company’s own adjusted cash flow, so the add-backs that survive underwriting set what the structure can pay the owner - with the difference that the equity moves to a trust owned by the employees, existing leadership keeps running the business, and there is no sponsor and no exit clock attached to the money.
If your company does $3M+ in EBITDA and you want to know what your earnings look like after someone else has read them, talk to IBO Advisors.
Curious whether an Independent Buyout fits your business?
Learn More