A Charlotte HVAC contractor, about $3.9 million in revenue, is for sale at 3.5x adjusted EBITDA. The seller's package shows reported EBITDA of $1,184,272 for the last fiscal year, then a schedule of add-backs totaling $165,500, for an adjusted figure of $1,349,772. That is the number the price is built on.
We ran the same business through the full workup: three years of P&L, two years of bank statements, the tax returns, the lease, the payroll register. The number that survived was $1,118,468.
The gap is $231,304, which is 17% of the seller's figure. At the 3.5x the seller is asking, that gap is $809,564 of purchase price. Nothing in the seller's package is false. Every line ties to a document. The gap comes from three habits of add-back schedules that are so common they have stopped looking like choices, and each one is testable with material the buyer already has. That is this issue: the three findings, and the test behind each, in the order a lender's credit desk will apply them.
Finding 1: the one-time expense that recurs
The schedule adds back a $14,500 legal settlement as a one-time expense. The label is doing all the work. The P&L legal line shows spend of the same order in all three years, and the operating account shows the same payee, a Charlotte law firm, paid in August in both years of bank data we were given.
An expense that recurs every year in the bank statements is an operating cost, whatever the schedule calls it. It may have been a settlement this year and something else last year; the business still pays a lawyer every August. Rejected, credited at $0.
The habit: for every one-time claim, find the payee in the bank data and check whether the payee appears in the other years. If it does, the add-back fails before you read the explanation. This takes ten minutes per line and it is the single highest-yield test in a workup, because sellers almost never check it themselves.
Finding 2: the owner who works for free
Officer compensation is booked at $39,996, and the schedule adds all of it back. Read literally, the next owner runs a $3.9 million contractor for nothing.
The right treatment adds back what the owner actually took and charges what it costs to replace the owner. For a business this size in Charlotte, a first-line supervisor or general and operations manager in specialty trade contracting, priced off BLS occupational data at the midpoint and adjusted for the revenue base, is about $110,000. Net effect against the seller's treatment: ($70,004).
The bank data confirms the business already pays its owner far more than the payroll line suggests. Distributions over the two years total $255,357. That money is real, it left the company, and the schedule treats it as if it did not exist.
The habit: never accept an owner comp add-back until you have priced the replacement. The replacement cost is a number a lender will insist on, so put it in the model before the lender does. If the buyer intends to run the business personally, that is a financing decision, not an earnings adjustment; the credit desk still charges a manager's salary because the loan has to survive the buyer getting hit by a bus.
The premises are leased from a landlord affiliated with the owner at $6,500 a month. Market for comparable space is about $9,500. The seller's schedule shows the rent adjustment at zero, which is the same as saying the lease is at market.
It is not. Normalizing to market reduces earnings by ($36,000) a year. This one is not an accounting argument; it is what will happen. After close, the owner-affiliated landlord either grants the buyer the same below-market rent, which no landlord does for a stranger, or the rent goes to market. The lender underwrites whichever number survives, and the lender will assume market. We entered this as a reviewer override pending a formal market-rent opinion, which is the one document in this workup still to be obtained.
The habit: put a market number on every related-party line, rent first. Related-party rent, family payroll, management fees to an affiliate, and equipment leased from the owner are the four places a seller's earnings quietly borrow from a relationship the buyer does not inherit.
What the rest of the schedule looked like
The remaining add-backs were a mix: some accepted in full with bank support, some accepted in part, a few rejected for the same reason as the legal line. None of them individually moved the number the way the three above did, and that is typical. Add-back schedules are usually right about the small things and wrong about the large ones, because the large ones are the ones with a story attached.
Why this decides deals nearer the line
This business still clears. Debt service coverage on the normalized $1,118,468 is 2.1x against the 1.25x standard, with $445,290 of annual earnings to spare. On the seller's number it would have read 2.5x. The buyer is safe either way, and the price negotiation is the only place the $809,564 shows up.
Move the same arithmetic to a deal at 1.4x on the seller's number and the 17% gap takes coverage below the standard. The loan gets smaller, the equity injection gets larger, or the price comes down; those are the only three exits. Starting October 1, on SBA 7(a) acquisitions of $3 million or more, the lender commissions the Quality of Earnings that produces the second number, engaged when the loan number issues and delivered before closing, and the borrower pays for it. The seller's schedule is no longer the last word on what the business earns. It is the first draft. The same test is what a buyer runs before the LOI, and what a seller's advisor runs eighteen months out; the rule only made it mandatory for the lender.
If you want to see what a haircut does to a loan, the 1.25x calculator at getworkup.ai/tools has a slider for exactly this: drag it to 17% and watch the maximum loan move. That slider is the whole letter in one control.
Next issue: how a Main Street deal actually dies between the LOI and the credit desk, and who is standing at each step.
Gray
Workup is the analyst's read on small business acquisition: pre-LOI screens and full diligence for buyers, and the lender-commissioned Quality of Earnings now required on SBA acquisitions over $3 million. Findings and implications, in five business days. getworkup.ai