Where a U.S. Buyer’s Assumptions Break on a Canadian Target

What this is

A U.S. buyer’s acquisition template embeds market assumptions: what a cap looks like, how long representations survive, what silence on sandbagging means, what terminating an employee costs. On a Canadian target several of those assumptions fail quietly, because the two systems share a vocabulary and the same words sit on different law. Our essay on what U.S./Canada deals require covers the traps outside the purchase agreement: permanent establishment, at-will termination, securities exemptions, privacy, Quebec. I cover the agreement itself here, and it sits inside my cross-border transactions practice.

The empirical anchor is the pair of studies published through the Mergers and Acquisitions Committee of the American Bar Association’s Business Law Section: the U.S. Private Target Mergers and Acquisitions Deal Points Study, 2025 edition (139 agreements signed 2024 through the first quarter of 2025), supplemented by the 2023 edition (108 agreements, US$30 million to US$750 million) where 2025 distributional data has not yet been republished; and the Canadian Private Target Mergers and Acquisitions Deal Points Study released in early 2025 (83 agreements of C$5 million and up, signed 2020 through 2022). The ABA’s file naming calls the Canadian study the 2025 edition, several firm summaries call it the 2024 edition; same dataset. The two current editions cover different deal vintages, so every comparison here is between each market’s most recent published data, not between contemporaneous deals.

The divergence table

Fourteen points in seven categories. Each cell states its base and its edition; the sections below expand each category and name the cases.

Sources: the ABA U.S. Private Target M&A Deal Points Studies (2023 and 2025 editions) and the ABA Canadian Private Target M&A Deal Points Study (released early 2025, deals signed 2020 through 2022), each read through the law-firm summaries described in the method note; the SRS Acquiom 2024 M&A Deal Terms Study; and the named cases and statutes. Compiled August 24, 2026.

Indemnity caps and survival: the same clause, different centers of gravity

Start with the cap. In the ABA 2023 U.S. study, the mean indemnification cap came in near 10.5 percent of transaction value, and roughly 86 percent of deals capped indemnification below the full purchase price. In the current Canadian study, 49 percent of the 83 sampled agreements cap indemnity at the full purchase price, double the 24 percent of the 2018 edition. A Canadian seller proposing a full-price cap is inside that market’s norms; a Texas buyer reading it as an aggressive opener is applying the wrong baseline. There is no shared market to cite; negotiate the number on its own terms. One vintage caveat: the 10.5 percent mean is single-sourced and one edition behind the 2025 headline data.

Survival is trending in opposite directions

In the ABA 2025 U.S. study, 41 percent of deals had no survival period for representations at all, up from 30 percent in 2023. The Canadian data moves the other way: 24 months is now the most common survival period for general representations and warranties, at 26 percent of deals, up from 18 months as the modal term in 2018. Survival is disappearing in the U.S. because the policy replaces it, and lengthening in Canada because, in the large majority of sampled Canadian deals, there is no policy.

Scrapes and baskets

The double materiality scrape, which removes materiality qualifiers for both breach determination and damages, appeared in 82 percent of ABA 2025 U.S. study deals, up from 69 percent in 2023. In the current Canadian study, scrapes of any kind appear in roughly 45 percent of deals, and the double variant in 56 percent of deals that have a scrape, up from 35 percent in 2018. On baskets, roughly 67 percent of U.S. baskets were deductibles in the 2023 U.S. study, against 46 percent deductible and 33 percent first-dollar in the Canadian study. For all three terms, propose the U.S. position expressly and expect to negotiate it rather than presenting it as boilerplate.

Representation and warranty insurance: a majority practice meets a minority one

Representation and warranty insurance was referenced in 63 to 64 percent of deals in the ABA 2025 U.S. study, up from 29 percent in the 2016-2017 edition. The current Canadian study puts it in a minority of deals, and summaries disagree on the size: two report usage of 13 to 15 percent, while two others report that 76 percent of deals made no express mention of a policy, implying roughly 24 percent. The direction is unambiguous either way, and the most likely explanation is deal size: 67 percent of the Canadian sample sits in the C$5 million to C$50 million band, below typical underwriting minimums. Fasken’s October 2025 bulletin describes the product as a fixture of Canadian private M&A, so adoption has probably grown since the 2020-2022 sample window. Probably grown is not bound coverage; confirm underwriter appetite before the structure depends on it.

Two conventions travel under one label

A January 2026 DLA Piper comparison distinguishes U.S.-style RWI from the European W&I convention it reports is increasingly used in Canada. Under the W&I convention the data room and the buyer’s diligence reports are deemed disclosed to the insurer, coverage is negotiated warranty by warranty on a spreadsheet, underwriting runs through formal written questions, and in an auction the policy can be seller-initiated. Premiums run roughly 0.5 to 1.5 percent of the insured limit against roughly 2.5 to 3.5 percent for U.S.-style RWI. Ask which convention the underwriter is using before relying on a quote; the answer decides whether the data room narrows coverage and who leads the process.

Sandbagging, good faith, and the limits of the anti-reliance clause

Eagle Force Holdings, LLC v. Campbell (Del. 2018) posed the sandbagging question in dicta, Arwood v. AW Site Services, LLC (Del. Ch. 2022) reaffirmed the pro-sandbagging position, and In re Dura Medic Holdings, Inc. Consolidated Litigation (Del. Ch., February 2025) settled it: reliance is not an element of a contract claim, so a buyer’s pre-closing knowledge of a breach does not bar recovery where the agreement is silent.

Common-law Canada is roughly where Delaware sat before 2025. A 2001 Alberta Court of Appeal decision in the Eagle Resources line leaned pro-sandbagging (cited as Eagle Resources Ltd. v. MacDonald, 2001 ABCA 264 by one summary; the method note flags a naming conflict). Transamerica Life Canada Inc. v. ING Canada Inc., a 2003 Ontario Court of Appeal decision, cast doubt on the approach. And 82 percent of deals in the current Canadian study are silent on sandbagging, 10 percent expressly pro, 8 percent expressly anti. Silence over an unsettled default is the weakest position; the clause has to be express.

The good-faith overlay has no U.S. counterpart

Since Bhasin v. Hrynew, 2014 SCC 71, Canadian contract law recognizes an organizing principle of good faith and a duty of honest performance. C.M. Callow Inc. v. Zollinger, 2020 SCC 45 extended the duty to half-truths and sometimes silence; Wastech Services Ltd. v. Greater Vancouver Sewerage and Drainage District, 2021 SCC 7 pulled the doctrine back, holding in substance that good faith does not require a party to surrender a contractual advantage. No court has applied the trilogy to a buyer that closed silently on a known breach; commentary argues both directions. The drafting response: an express acknowledgment that the buyer’s knowledge, however acquired, is not unfair dealing and that closing without disclosure is not dishonest performance.

Fraud cannot be drafted away

In Delaware, the ABRY Partners line enforces a buyer-authored anti-reliance clause against extra-contractual fraud claims; Pearce v. NeueHealth (Del. Ch. 2024) let fraud claims proceed where the agreement lacked one. In Canada the ceiling is lower: 10443204 Canada Inc. v. 2701835 Ontario Inc., 2022 ONCA 745 holds that no entire-agreement or exclusion clause shields a party from its own fraudulent misrepresentation, whatever the parties’ sophistication, on an appellate line running back to 1964. For negligent misrepresentation, Canadian courts appear more receptive to specific, prominent non-reliance language than to generic integration boilerplate, a principle stated at the level the sourcing supports. Keep the buyer-authored clause, enumerate what it disclaims (the data room, projections, oral statements, everything outside the disclosure schedules), and handle fraud structurally: a closing bring-down certificate, an uncapped fraud carve-out, and a clear view of what the policy’s fraud exclusion leaves uninsured against the seller.

Escrow and holdback: smaller, longer, differently motivated

U.S. practice, per the SRS Acquiom 2024 M&A Deal Terms Study (2,100-plus private-target deals closed 2018 through 2023), makes escrow near universal: more than 90 percent of deals carry a purchase-price-adjustment mechanism, typically a working capital adjustment, with escrow, and indemnity escrows run near 10 percent of deal value on uninsured deals, roughly half a percent where RWI is bound, with a 12-month median survival. One honesty note: the current ABA U.S. editions cleanly report only a separate adjustment-escrow figure (53 percent in 2023, 58 percent in 2025 per a single summary); no current general indemnity-escrow rate from the ABA U.S. study exists in the secondary literature, hence SRS carries the U.S. side here.

In the current Canadian study, 38 percent of deals used an escrow or holdback for indemnification, down from 60 percent in the 2018 edition, and average size fell from 10.85 percent to 4.56 percent of deal value. Those figures trace to a single authored summary, and the decline may reflect the sample’s smaller deals rather than a real loosening of seller protection, so I present them as the best available reading, not a settled trend. Directionally well supported: Canadian escrows are rarer and relatively smaller, and they sit against longer survival, commonly 24 months against the U.S. 12-month median. And one Canadian holdback has no U.S. cousin: the section 116 withholding holdback, covered below, which layers on top of the indemnity escrow.

Employment: the notice liability the form file does not price

Texas counsel prices an individual termination at close to zero because at-will employment makes it so. Canada has no at-will doctrine: every non-unionized employee, absent a valid limiting contract, is entitled to reasonable notice or pay in lieu under the factors from Bardal v Globe and Mail Ltd (1960): position, tenure, age, and availability of similar employment. The one-month-per-year heuristic with a ceiling near 24 months is a heuristic only. McCarthy Tétrault’s count of awards in Westlaw’s Canadian wrongful-dismissal quantum database found averages rising from 10.4 months (1970 to 2010) to 15.7 months (trailing three years), and Lischuk v K-Jay Electric Ltd, 2025 ABKB 460 awarded 26 months, the first Alberta decision past the informal ceiling. Alberta’s Employment Standards Code minimums top out at eight weeks at ten years of service, and they are a floor under the common-law entitlement, not the measure of it.

The termination clause is guilty until proven innocent

Waksdale v Swegon North America Inc., 2020 ONCA 391 voids the entire termination scheme if any part of it falls below the statutory floor; the Supreme Court of Canada denied leave in January 2021, and Dufault v Corporation of the Township of Ignace reaffirmed the rule in December 2024. Alberta’s Court of Appeal has not adopted that rule. The leading Alberta authority, Singh v Clark Builders, 2025 ABKB 3, is a trial-level decision requiring clear and unambiguous language before a clause displaces common-law notice. The approaches converge in effect without being identical. Treat every inherited clause as presumptively failing to cap liability until counsel in the target’s province confirms otherwise.

Deal structure does not reset the clock

In a share deal the employer never changes, so accrued seniority and the wrongful-dismissal back book come with the company. An asset deal does not help: Alberta’s Employment Standards Code deems employment continuous when a business is sold, a rule applied against an asset purchaser in Certified Redi-Mix Concrete Group v L’Hirondelle, 2022 CanLII 79247 (AB ESA). Diligence needs every employee’s original hire date, age, and title, and the price or a specific indemnity must absorb the inherited exposure.

Tax structuring: section 116, exchangeable shares, and the treaty

Section 116 of the Income Tax Act (Canada) has no U.S. analogue. Where a non-resident sells taxable Canadian property, generally shares deriving more than half their value from Canadian real or resource property at any time in the preceding 60 months, and no CRA clearance certificate is delivered, the purchaser is personally liable to remit the withholding, generally 25 percent, with no limitation period. The standard protection is a certificate closing condition plus a separate holdback of the statutory percentage. Two qualifiers: many operating-company share sales fall outside taxable Canadian property entirely, so the threshold test comes before the mechanism; and the 25 percent figure traces to commentary written while the 2024 budget’s proposed increase to 35 percent was still pending, before its cancellation on March 21, 2025, so verify the operative rate at signing. On the seller’s side, the capital-gains inclusion rate remains one-half, a cancellation formalized in Budget 2025. Any model built between June 2024 and March 2025 on the two-thirds rate needs rebuilding.

Stock consideration needs a structure, not a swap

A U.S. buyer expects a tax-deferred stock-for-stock exchange under Internal Revenue Code section 368; for Canadian holders, a direct swap into U.S. shares is a taxable disposition. The standard fix is the exchangeable-share structure: a Canadian Acquisitionco issues exchangeable shares on a section 85 rollover, engineered for economic equivalence to the buyer’s stock, plus a Callco call right and a parent support agreement. The decision belongs at the letter of intent stage: the structure adds entities, governance mechanics, coordinated tax review on both sides, and confirmation that the shares avoid derivative-forward-agreement characterization under Canadian rules.

The treaty does not follow the flag

Canada’s statutory withholding on payments to non-residents is 25 percent; the Canada-US Tax Convention reduces dividends to 5 or 15 percent for qualifying recipients. Article XXIX-A denies those rates to a U.S. resident that is not a qualifying person; CRA’s guidelines set out the objective routes: public trading, the ownership and base-erosion tests, or an active U.S. trade or business. A closely held acquisition NewCo can fail all of them, and discretionary relief runs roughly six months. Separately, Article IV(7)(b) can deny treaty rates where the payee is a fiscally transparent LLC; a C corporation or an interposed U.S. partnership is the usual work-around. The LLC point rests on thinner sourcing than everything else here and is flagged as a mandatory question for tax counsel. Run the treaty analysis, then name the buying entity.

Investment Canada Act and Competition Act: two gates, neither of them HSR

A U.S. buyer’s regulatory reflex is HSR premerger notification; CFIUS, its closest screening analogue, reviews investment coming into the United States. The Investment Canada Act runs two independent tracks: net-benefit review, pre-closing ministerial approval required only for direct acquisitions of control above an enterprise-value threshold that reindexes every January, with approval typically riding on multi-year written undertakings; and national-security review, which has no dollar floor, reaches any non-Canadian investor, and since Bill C-34 extends to minority interests and certain asset transactions, on a 45-day initial clock, extendable. Bill C-34 received royal assent March 22, 2024; its first tranche came into force September 3, 2024, including a ministerial power to impose binding interim conditions during a pending review. The mandatory pre-closing filing regime for prescribed sensitive sectors was still awaiting regulations as of the newest official update I have verified, the ISED modernization page’s August 21, 2025 revision. State-owned buyers face a lower threshold and heavier scrutiny.

The Competition Act after the amendment wave

Canadian merger notification turns on a two-part test: a transaction-size threshold that is GDP-indexed but has been held unchanged for five consecutive years per the Bureau’s 2026 announcement, and a party-size threshold that is not indexed and has not moved since it was set, which is why mid-market deals trip Canadian notification more readily than HSR. A complete filing starts a 30-day waiting period, restarted by a supplementary information request. The efficiencies defence, former section 96, was repealed effective December 15, 2023; since June 20, 2024 the Act carries rebuttable structural presumptions where combined share exceeds 30 percent or the post-merger concentration index exceeds 1,800 with an increase of more than 100; and the window to challenge an un-notified merger runs three years after closing. An advance ruling certificate is a safe harbor; a no-action letter is not. Osler’s review of the first 18 months reports fewer certificates, longer reviews, and more parties closing at expiry. Final merger enforcement guidelines remained in draft after a consultation closing February 11, 2026.

The drafting consequences compress to four moves: separate conditions for Competition Act clearance and Investment Canada Act approval; an efforts covenant that anticipates interim conditions and closing at expiry; an outside date sized to two clocks that run independently; and survival for regulatory matters sized to the three-year window for un-notified deals. Where regulatory risk is material, Blakes’ Canadian public M&A guidance describes reverse break fees at or above the target’s own break fee, a public-company practice cited here as such.

Where the dollar thresholds live

Method, and where I am not certain

Doctrine, the named cases and statutes, can be checked directly. Market data cannot: the ABA’s primary study PDFs sit behind member access, so every deal-points figure here comes through named law-firm summaries, and where only one summary reports a figure, the dataset’s confidence field says so. Three U.S. distributions (cap size, survival breakdown, basket structure) are still 2023-vintage because the 2025 edition’s detail has not been republished. The documented uncertainties that matter most:

• The Canadian RWI rate is unresolved. Summaries of the same study support 13 to roughly 24 percent, depending on whether they measure usage or express mention; I print the range here, not a pick.
• The escrow category is the weakest point I have. The Canadian prevalence and size figures are effectively single-authored, and the U.S. side lacks a current ABA indemnity-escrow rate entirely.
• The Alberta sandbagging citation is disputed. One summary cites Eagle Resources Ltd. v. MacDonald, 2001 ABCA 264; another renders the respondent differently. The reporter could not be checked directly, so the discrepancy stands.
• The good-faith trilogy is untested on sandbagging facts. No Canadian court has applied Bhasin, Callow, and Wastech to a buyer that closed silently on a known breach. That is a gap in the law, not in the research.
• The Investment Canada Act sensitive-sector filing regime is a moving target. Its not-yet-in-force status is verified to the ISED page’s August 21, 2025 revision and re-checked at each reissue.
• Several granular Canadian figures are single-sourced. The survival-period distribution, the escrow decline, and the overall scrape rate each trace to one authored summary.

Each row carries the category, the point, the U.S. default assumption, the Canadian reality, the drafting consequence, the anchoring study edition or authority, and a stated confidence level keyed to the method note above.

Common questions

Can I rely on my pro-sandbagging clause on a Canadian deal?

An express clause is the right start, and far better than silence: 82 percent of deals in the current ABA Canadian study are silent on sandbagging, and Canadian law supplies no settled default to fill that silence. But no Canadian court has tested a pro-sandbagging clause against the good-faith duties recognized in Bhasin v. Hrynew, 2014 SCC 71 and C.M. Callow Inc. v. Zollinger, 2020 SCC 45, so pair the clause with an express acknowledgment that closing with knowledge of a breach is not dishonest performance.

Do I still need indemnity survival if I buy RWI?

On a US deal the market increasingly says no: 41 percent of deals in the ABA 2025 US study had no survival period at all, with the policy as primary recourse. On a Canadian deal, confirm the policy is actually bound before copying that structure. RWI appeared in a minority of deals in the current Canadian study, roughly 13 to 24 percent depending on the summary, and where there is no policy, the survival period and the indemnity are the protection.

What happens if I terminate the target's employees on US assumptions?

Canada has no at-will employment. Every non-unionized employee is entitled to reasonable notice or pay in lieu, assessed on the Bardal factors: position, length of service, age, and availability of similar work. One published count of recent awards averages 15.7 months over the trailing three years, and an Alberta court awarded 26 months in Lischuk v K-Jay Electric Ltd, 2025 ABKB 460. Statutory minimums, which top out at eight weeks in Alberta, are a floor, not the liability.

What is a section 116 certificate and who bears the risk?

It is a clearance certificate the Canada Revenue Agency issues when a non-resident sells taxable Canadian property. Without it, the buyer bears the risk: the purchaser is personally liable to remit the withholding, generally 25 percent, with no limitation period on that exposure. Buyers protect themselves with a closing condition and a separate holdback of the statutory percentage until the certificate arrives. Many operating-company share sales fall outside taxable Canadian property entirely, so test that first.

When does the Investment Canada Act touch my deal?

On two independent tracks. A direct acquisition of control needs pre-closing net-benefit approval only above an enterprise-value threshold that reindexes every January, published on the official ISED thresholds page. The national-security track has no dollar floor and can reach minority stakes and some asset purchases, so even a small deal needs condition language covering it. Clearing one track is not clearing the other.

Why not print the dollar thresholds?

Because they change on a schedule and a reference page should not quietly go stale. The Investment Canada Act net-benefit thresholds reindex to GDP every January, and the Competition Act transaction-size threshold is republished annually. The official pages linked in the threshold note above always carry the current figures, and I carry the analysis instead.

The table says where the assumptions break. It does not paper your deal.