Adjusting earnings is standard practice; each adjustment is still a claim to be proven
Start with the reassurance, because the suspicion behind the question is half right. A private company's reported profit is shaped by its owner's personal and tax decisions: what they pay themselves, what they run through the company, what rent their holding company charges the business. Restating profit without those choices is not manipulation; it is the only way to compare the business to any other or to price it for a stranger. Done honestly, normalization can even cut against the seller, as when an owner underpaying themselves must have a market manager's salary inserted, reducing the earnings.
Where your instinct is right: a normalization schedule is an advocacy document. The seller or broker prepared it to support the asking price, every judgment call in it leans upward, and the aggressive versions bundle genuinely personal expenses together with costs any owner would incur, hoping the pile gets accepted whole. So the answer to is that legitimate is: the practice, yes; the schedule in front of you, only line by line, with evidence. Your job is not to reject normalization but to re-perform it.
Normalized EBITDA is the deal's common currency, used three ways
EBITDA is earnings before interest, taxes, depreciation and amortization, and the deal world uses it because it strips out how the current owner financed the business and how they depreciated its assets, leaving a rough proxy for operating cash generation that can be compared across companies. Normalized EBITDA then applies the owner-specific adjustments to that base. The result is used three ways at once, which is why one number gets so much attention.
First, pricing: private businesses are typically priced as a multiple of normalized EBITDA, with the multiple reflecting size, industry, growth and risk. Second, financing: the lender underwriting your acquisition sizes the loan against its own normalized EBITDA and the debt service it can cover, and later writes covenants against a defined EBITDA in the credit agreement. Third, comparison: normalized EBITDA is how you weigh this business against the other one you looked at, and against simply keeping your money. Because all three uses hang off the same number, an inflated add-back does not just flatter the seller; it distorts your price, your borrowing case and your judgment simultaneously.
One sizing caveat before the taxonomy: on smaller deals, brokers often quote seller's discretionary earnings instead, which adds back the owner's entire compensation on the assumption the buyer works full time in the business. That is a different base from normalized EBITDA after a market salary, the two can sit far apart, and multiples quoted for one applied to the other misprice the deal from the first conversation. Always ask which base a quoted multiple refers to.
Sort every add-back into one of three buckets, and demand evidence accordingly
Every line on a normalization schedule fits one of three buckets: routine adjustments any professional would make, claims that are legitimate only if proven, and items that should be refused on principle. The table is the sorting guide we apply in practice.
| Add-back category | Legitimate when | Evidence to demand |
|---|---|---|
| Owner compensation to market | Always adjusted, in whichever direction the facts point | Payroll records, and a defensible market salary for the role |
| Personal expenses through the company | The expense truly disappears with the owner | General ledger detail, invoices, and who actually used it |
| Rent to a related landlord | Restated to the market rent you will actually pay | The lease you are offered, and comparable local rents |
| Family members on payroll | Only the portion beyond real work performed | Roles, hours and who will do that work after closing |
| One-time costs (lawsuit, flood, move) | Genuinely non-recurring, and matched by removing one-time gains | The documents behind the event, and a multi-year expense history |
| Run-rate and pro-forma savings | Rarely; changes contracted and priced before closing at most | Signed agreements, not intentions or projections |
Two patterns in that table deserve emphasis. One-time items must be tested in both directions: a seller who removes the flood repair but keeps the insurance recovery, or removes the bad year's write-off but keeps the good year's subsidy, is normalizing asymmetrically, and a three-to-five-year expense history exposes it, since a one-time cost that appears in some form every year is simply a cost. And the run-rate bucket is where legitimacy usually ends: adjustments for the price increase just announced, the customer about to sign, or the savings you could achieve after closing are projections dressed as history. The improvements you will create belong to you; paying the seller a multiple of them is buying your own future work.
The multiple arithmetic: every accepted dollar is several dollars of price
The reason to fight over small add-backs is that none of them are small once multiplied. A deal priced at a multiple of normalized EBITDA converts every accepted add-back dollar into several dollars of purchase price: at four times, a modest-looking pile of soft adjustments totalling one hundred thousand dollars is four hundred thousand dollars of price, funded with your equity and your borrowed money. Sellers understand this arithmetic precisely, which is why the schedule exists at all, and why the negotiation over which lines survive is frequently worth more than the negotiation over the multiple itself.
The arithmetic also gives you the honest frame for the conversation. You are not accusing anyone of dishonesty by rejecting a line; you are declining to pay a multiple of a cost that will exist for you, or of an improvement you will have to earn. In practice we build a buyer's version of the schedule beside the seller's, agree the routine items quickly, and concentrate the negotiation on the contested middle, where the evidence requested in the table settles most arguments faster than positioning does. A seller whose add-backs collapse under ordinary documentation has told you something useful about the rest of the file; that finding belongs in your wider financial due diligence, where the earnings work connects to revenue quality, the balance sheet and the tax history.
EBITDA is not cash flow, and your lender will prove it to you
Even a perfectly normalized EBITDA overstates the cash the business will hand you, because it ignores four real claims on it: tax, equipment replacement, the working capital that growth consumes, and the debt service on the loan that bought the business. Walk the bridge honestly, EBITDA less cash taxes, less the capital spending that keeps the equipment running whether or not it is called maintenance, less the cash absorbed by growing receivables and inventory, less principal and interest, and what remains is what pays you. On capital-intensive businesses, and most trades, logistics and manufacturing businesses qualify, the gap between EBITDA and free cash is the difference between a deal that works on paper and one that works.
Your lender will run exactly this bridge, which is worth knowing before you present a number. The bank underwriting the acquisition builds its own normalized EBITDA, usually more conservative than the seller's, tests debt service coverage against it with a margin, and expects financial projections that carry the normalized earnings forward under your ownership with assumptions a credit analyst can verify. After funding, lender reporting continues the discipline: annual statements, covenant certificates calculated against a defined EBITDA, sometimes quarterly interims. Preparing that package, business financing and projections work by a CPA firm for Ontario buyers, is standing scope in our Business Financing Advisory practice, and it is far easier to build when the EBITDA underneath it was tested rather than accepted.
The same number also interacts with the deal's legal shape. The transaction structure decides what the earnings are worth to you after tax, since an asset purchase gives fresh deductible cost against future profits while a share purchase inherits the corporation's old tax positions; the trade-offs sit in asset purchase versus share purchase. And where the parties cannot agree on a contested slice of EBITDA, structure can bridge it: an earn-out pays for the disputed earnings only if they materialize, provided the earnings definition and the accounting policies behind it are drafted tightly enough to survive the first measurement date.
Five facts decide how far to trust the number, and how we test it
How much weight a normalized EBITDA deserves comes down to five facts worth naming before you negotiate. The size and direction of the total adjustment, because a schedule that doubles reported profit demands proportionally more proof than one that trims it. The documentation quality behind the contested lines, since legitimate add-backs survive a general ledger pull and aggressive ones do not. The trend of the normalized number across three to five years, because a figure that only looks good in the latest year is a projection, not a history. The capital intensity of the business, which sets the gap between EBITDA and cash. And the earnings base being quoted, discretionary earnings or EBITDA after a market salary, because the multiple only means something attached to the right base.
We test seller schedules for buyers across Mississauga and the GTA as defined-scope work under Strategic Projects: a re-performed normalization with each line evidenced, accepted or rejected, the cash flow bridge and debt service case built on the result, and the findings carried into the price negotiation and the financing package. The wider purchase, structure, tax, diligence and closing, is mapped in buying a business in Canada, and the schedule is worth testing before the letter of intent anchors the price to it, not after.
