Anonymised, illustrative composite. Buying a farm holding company is not buying the farm directly, and the federal loan programme that finances one does not treat them as the same purchase.
At a glance
A Saskatchewan grain operation had been run through a two-tier structure since the founder's estate freeze fifteen years earlier: a holding company owning 100% of the shares of the operating farm corporation, with the founder's growth shares later sold down to his daughter. She was now selling to a neighbouring arm's-length buyer — not the farmland directly, and not the operating company's shares directly, but the holding company that sat above both. Price for 100% of the holding company's shares was $2,550,000. The buyer had $380,000 of his own capital and had financed his equipment purchases for years through the programme that had worked for every deal before this one: the Canada Small Business Financing Programme. It excludes farming by definition, which sent the file to the Canadian Agricultural Loans Act instead — the same statute that financed the farm deal the CSBFP could not touch on the last file to test it.
That earlier file ran the numbers on a direct purchase: buy the operating company's shares outright, test the operating company's own property against the 90% floor in para 4(1)(g.1), done. This file was not that purchase. The buyer was not acquiring the operating company. He was acquiring the holding company two tiers up, whose only real asset was the operating company's shares. The lender's first-pass credit memo treated the two as interchangeable — “the holding company owns nothing but a 100%-farming operating company, so it is obviously all-farm too” — and ran a single composition test on the combined, look-through numbers. The statute does not test combined numbers. It tests each entity in the chain against its own property, one link at a time, and a look-through calculation is not the calculation para 4(1)(g.2) actually performs.
The operating company's own property, at fair market value: land under crop $1,900,000, grain storage and buildings $230,000, machinery and equipment $310,000 — $2,440,000 in the categories the Act lists. Against that, the operating company had also built up a $290,000 GIC ladder from several strong crop years, parked at the corporate level rather than distributed — non-qualifying property. Total operating-company property: $2,730,000. Qualifying share: $2,440,000 / $2,730,000 = 89.4%. Below the 90% floor, at the lower link.
The holding company's own property was simpler on its face: nothing but the operating company's shares, appraised at $2,520,000, plus a $30,000 cash buffer left over from a prior dividend — $2,550,000 total, matching the sale price. Naively, that reads as 100% farm-related. It is not what the statute tests. What it tests is whether those operating-company shares are themselves property “referred to in paragraph (g.1)” — which turns on the operating company's own 89.4%, not on how the holding company's balance sheet looks from the top.
Paragraph 4(1)(g.1) of the Canadian Agricultural Loans Act funds “the purchase of shares in a corporation… if 90 per cent or more of the fair market value of the property of the corporation… was attributable to property referred to in this subsection that was used principally in the course of carrying on the business of farming in Canada.” That is the single-tier test the prior file ran: the target corporation's own property, measured against the Act's own list of farm-use categories.
Paragraph 4(1)(g.2) funds the next tier up, and it is written as a chain, not a look-through: “the purchase of shares in a corporation… if 90 per cent or more of the fair market value of the property of the corporation… was attributable to shares, an interest in a partnership or memberships in a corporation, partnership or cooperative referred to in paragraph (g.1) or to property referred to in this subsection that was used principally in the course of carrying on the business of farming in Canada.” Read the phrase that does the damage: “shares… referred to in paragraph (g.1).” Not shares in a company that happens to farm — shares that themselves satisfy paragraph (g.1)'s own 90% test, measured against that company's own property. The holding company's shares in the operating company only count toward the holding company's numerator if the operating company clears 90% on its own books first.
At 89.4%, the operating company did not clear it. That means its shares were not “referred to in paragraph (g.1)” at the moment of the test, which means the holding company's own composition test had nothing qualifying to count: not 100%, not 92%, effectively zero — the $2,520,000 of operating-company shares contributed nothing to the numerator, leaving only the $30,000 cash buffer, itself non-qualifying. A structure that looked, from the top, like it was obviously all-farm failed the statute's actual test completely, because the statute does not ask what the top of the chain looks like. It asks whether the link immediately below cleared its own floor.
The financing ceiling would have followed the same path as a direct purchase in any event: s.4(3)(c)(ii) caps “paragraphs (1)(a) to (e) and (g.1) to (i)” — a range that runs through (g.2) — at $350,000, not the $500,000 aggregate in subparagraph (i). Buying the second tier does not buy a bigger ceiling; it only changes which entity's books the eligibility test reads.
The fix ran at the lower link, not the upper one. The operating company paid a $30,000 dividend to the holding company out of the GIC ladder before closing, trimming its non-qualifying property to $260,000 against the same $2,440,000 of qualifying property. New ratio: $2,440,000 / $2,700,000 = 90.4%. The operating company now cleared paragraph (g.1) on its own books, which made its shares property “referred to in paragraph (g.1)” for the first time — and the holding company's shares, now backed by a qualifying lower-tier holding worth $2,520,000 against total holding-company property of $2,550,000, cleared paragraph (g.2) at 98.8%. The $30,000 that moved up a tier did not, on its own, fix anything; it was the composition of the entity that received it below that mattered. The loan funded at $350,000, the sub-limit for a (g.2) purchase, against a $2,550,000 equity price, with the balance in a vendor take-back and the buyer's own equity.
Financing the deal on the credit memo's original look-through math would have produced an approval the Act does not actually authorise — an eligibility finding resting on a calculation paragraph (g.2) does not perform. Caught after disbursement rather than before, the exposure sits with the lender under the loss-sharing guarantee in s.4(1), not with the borrower: a loan advanced on an eligibility test that was never really met is a loan the Minister's 95% coverage was never actually behind, discovered only if the loan later defaults and the guarantee is tested. The $30,000 dividend that fixed it cost a single pre-closing resolution. Finding it after a default would have cost the guarantee itself.
The tell is the word “referred to” sitting inside paragraph (g.2) — two words that turn a plain-English “is this basically all farm” question into a cross-reference back to a different paragraph's own test, applied to a different entity's own books. Whenever a target sits more than one tier above the operating assets, run the 90% test at every tier separately, starting at the bottom, before pricing the deal on how the group looks from the top.
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.
