Anonymised, illustrative composite. A buyer who discovered the seller had engineered the revenue numbers wanted the deal undone, not a cheque. The purchase agreement said money was all it was ever going to get — and for most breaches, that clause would have held.
At a glance
A Nova Scotia-based wholesale bakery, supplying independent grocers and a handful of food-service accounts across the Maritimes, changed hands on a share purchase at $2,300,000. The buyer, a first-time operator with a food-manufacturing background, financed the purchase with a term loan and a modest vendor take-back. The purchase agreement followed a structure common across mid-market Canadian deals of this size: general representations surviving twelve months, an indemnity capped at 10% of purchase price, and a $23,000 basket.
Buried in Article 8, alongside the cap, sat a clause the buyer's lawyer had seen in nearly every deal she had closed and had never had reason to test: an “Exclusive Remedies” provision stating that, other than for fraud, the indemnity was the buyer's sole and exclusive remedy for any breach of any representation or warranty, and that the buyer waived rescission, specific performance, or any other remedy at law or in equity for such a breach.
Eight months after closing, the buyer's new controller noticed a pattern working through the general ledger: a disproportionate share of the largest customer's orders in the final quarter before the letter of intent had been shipped, invoiced, and then quietly credited back or returned within thirty days — timed to book revenue inside the measurement period the seller's broker had used to market the business, and reversed just after the deal closed. Internal emails, produced later by a former warehouse manager, showed the timing was deliberate: shipments were held back until the last week of the quarter specifically to hit a revenue number the seller had been circulating to buyers.
The buyer did not just want its money back for the overstated revenue. It no longer trusted the rest of the business it had bought against numbers it now knew had been engineered, and on the corrected figures the business was worth far less than a capped indemnity claim could ever recover. It wanted the deal undone — the shares returned, the price back — not a cheque bounded by a clause the seller's own counsel had drafted before anyone knew there would be a dispute.
Two principles were on a collision course, and the clause's plain wording only resolved one of them. Canadian courts generally hold sophisticated commercial parties to an exclusive-remedies bargain, including a bargain that excludes rescission for an inaccurate representation — that is precisely what the clause was drafted to do, and on its face it purported to reach “any breach of any representation.”
But the clause carved out fraud by its own terms, and Canadian law would have reached the same result even without that carve-out. Treadstone Law's own comparison of the three categories of business-sale misrepresentation states the effect of a contractual exclusion clause plainly, category by category: it “generally cannot shield a seller who committed fraud” — the opposite of its effect on a negligent or innocent misstatement, where the same clause “may limit or exclude the claim, depending on how the clause is worded” or stand as “the seller's main practical protection.” The same principle sits behind the modern Canadian law of exclusion clauses generally: in Tercon Contractors Ltd. v. British Columbia (Transportation and Highways), 2010 SCC 4, the Supreme Court of Canada held that even a clause whose wording plainly covers the loss claimed can still be refused effect where enforcing it would be unconscionable or contrary to an overriding public policy — and a party's own fraud is the paradigm case, not a marginal one.
So the fight was never really about what the clause said. It was about what the seller knew. Treadstone Law draws the line exactly where this file turned on it: fraudulent misrepresentation is where “the seller knew the statement was false, or was reckless as to whether it was true” — as opposed to negligent misrepresentation, where the seller simply failed to take reasonable care. A clause that validly excludes rescission for the second cannot exclude it for the first, and proving which one this was became the entire case.
The buyer's counsel spent four months assembling the evidence: shipment logs cross-referenced against credit-memo dates, and the internal emails instructing warehouse staff to hold shipments until the last days of the quarter specifically to hit a revenue figure. That was enough to move the claim from careless bookkeeping into deliberate concealment. Facing a credible fraud claim that the exclusive-remedies clause could not touch, the seller's counsel negotiated rather than litigated: a settlement of $1,100,000, well above the $230,000 cap though below the buyer's full valuation claim, paid in exchange for a release and no admission of liability.
Had the same revenue pattern turned out to be careless rather than deliberate — a seller who simply hadn't reconciled credit memos against the sale numbers, rather than one who had timed them — the clause would have held completely. The buyer's recovery would have been capped at $230,000 against $1,730,000 of overstated value, and the difference between negligence and fraud, on facts otherwise identical, would have been the difference between a full recovery and a ten-cents-on-the-dollar one.
Credits that cluster in the days after a measurement period closes, concentrated on the single largest customer, in a pattern that repeats every quarter and reverses immediately after the deal completes, is a revenue-recognition-timing test a buyer's diligence accountant can run before signing — comparing shipment dates against credit-memo dates for the top five customers across the trailing eight quarters — rather than a controller stumbling onto it eight months into ownership.
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.
