Anonymised, illustrative composite. A specialty contractor's seller add-back schedule showed a different one-time cost every year for seven straight years. Each was genuine. None of them ever repeated by name. All seven together told a different story than any one of them alone.
At a glance
The target was a specialty contractor being sold by its founder after eighteen years in business. The buyer's diligence team requested seven years of financial statements, rather than the customary three, because the seller's confidential information memorandum leaned heavily on a normalized EBITDA figure built from add-backs, and the buyer wanted to see the pattern behind it rather than just the most recent year.
What the seven-year file showed was a different unusual item every single year: a customer lawsuit settlement in year one, uninsured storm damage to a job site in year two, an ERP system implementation in year three, a facility relocation in year four, one large customer's bankruptcy write-off in year five, a second lawsuit settlement in year six, and a system upgrade in year seven. The current year carried an eighth item, a $210,000 payment to settle a dispute with a subcontractor, which the seller wanted treated the same way as the previous seven.
Tested one at a time, every item held up. Each had a real invoice, a real cause, and no connection to the owner's personal affairs or to how the buyer would run the business differently. A conventional add-back review, which tests each claimed item against the question “did this specific expense recur?”, would have cleared all eight, because none of the eight expense TYPES repeated — there was one lawsuit settlement in year one and a different lawsuit settlement in year six, not the same lawsuit twice.
But run across all seven years at once, a second pattern appeared that a single-year or even three-year review would not catch: this business incurred at least one material, unbudgeted disruption cost, averaging roughly $185,700, in every single year for seven consecutive years. The specific cause changed. The fact that something happened, at roughly that scale, did not.
Seven years of add-backs: approximately $210,000, $165,000, $195,000, $140,000, $230,000, $175,000, $185,000 — a total of roughly $1,300,000 and an average of $185,700 a year. The current year's proposed add-back was $210,000, close to the seven-year average.
The buyer's normalization treated the average annual amount, not zero and not the full current-year figure, as an ongoing cost of running the business. At the deal's 6.5x multiple, the difference between adding back the full $210,000 and normalizing to the $185,700 average was modest in a single year — but stripping out all seven years' worth of “non-recurring” items from the trailing multi-year average the seller's own banker had used to establish the EBITDA trend added roughly $1,300,000 to reported cumulative earnings, worth six turns of the deal's multiple if left uncorrected across the valuation base.
No statute or accounting standard defines “non-recurring” for the purpose of a private-company sale; it is a quality-of-earnings judgment, and the judgment turns on a distinction that a single-year review cannot make: non-recurring in KIND, meaning the specific expense type does not repeat, versus non-recurring in AMOUNT, meaning the CATEGORY of expense — irregular, unbudgeted, but real — does not recur.
Treadstone Law's guidance on questioning add-backs states the ordinary, single-item version of this test plainly: “If it recurred, it isn't one-time, and it belongs in the ongoing cost picture of the business rather than out of it.” That is the right test applied to one repeating expense type. This file is the harder case the same principle implies but does not spell out: apply it to the CATEGORY across a multi-year window, not to any single line, and a business that has produced a genuine, individually-defensible “one-time” item every year for seven years is not experiencing seven unrelated events. It is a business whose normal operations include roughly $185,700 a year of irregular cost, and a buyer who acquires the business should expect to keep encountering it.
The parties agreed to normalize EBITDA by adding back the seven-year average of $185,700 rather than the full current-year figure of $210,000, and to disclose the methodology in the purchase agreement's EBITDA definition so a future dispute over the closing adjustment could not simply re-litigate which year's number applied.
The seller's banker initially resisted, arguing each item's individual documentation made it fully add-backable. The buyer's position held once the seven-year table was put side by side: no reasonable buyer models a business that has never, in seven years, gone a single year without a material unbudgeted cost, as if next year will be the first exception.
Accepting each year's add-back on its own individual merits, without ever testing the pattern across years, would have let a genuinely recurring cost of the business — averaging $185,700 a year — disappear from the earnings base entirely. Capitalised at the deal's multiple, that is real purchase price paid for earnings the business has never actually produced in any of the seven years reviewed, and is unlikely to produce in the first year after closing either.
The tell was not in any single year's file — it was in having seven years of files side by side instead of three. A three-year lookback would have shown two or three unrelated one-time items and cleared every one of them without complaint. Only pulling the full seven years turned eight individually true stories into one pattern: this business has never had a clean year, and the buyer should not model one either.
A 30-minute call is enough to tell you whether AI pays for itself here.
Canadian small-business transaction data is not published anywhere, so most valuations in this country quote an American benchmark. The Deavo–Treadstone Acquisition Index is a daily record of Canadian listings built to replace that: asking-price distributions by province and city are published now, and days on market, departure rates and asking-to-sale spreads follow as the series lengthens. Leave an email and we will tell you as each measure lands.
No pitch, no listings. One email when the first report lands.
