Retail trade

Tire retail is a service business with an inventory problem: the product is bulky, seasonal and price-shopped online before the customer arrives. National chains and the dealership service lanes hold the convenience positions, and the parts counter behind it competes with same-day delivery from a distributor the independent also buys from. The repair-shop software at 8111 and the dealer systems at 441110 are screened separately.

The pre-screen called this a candidate, and the screen does not overturn it. The instructive figure is Lowe's round trip. It bought RONA in 2016 in a transaction valued at C$3.2B (US$2.4B), creating a 539-store Canadian business with about C$6B of pro forma revenue [A]; on 3 February 2023 it sold the whole Canadian retail business — RONA, Lowe's Canada, Réno-Dépôt and Dick's Lumber, corporate and dealer-owned — to Sycamore Partners for US$491M in cash plus contingent consideration first valued at $21M, having already written US$2,061M off the long-lived assets and booked a further US$421M loss on sale [A]. The business was less than 6% of Lowe's fiscal 2022 sales and cost it about 60 basis points of operating margin [A]. What that establishes is that national scale in Canadian home improvement was not worth what the best-capitalised entrant paid for it. What it does not establish is that the independent yard is a good business. Independents live on contractor trade — delivered lumber, the estimating desk, the house account — and rent their purchasing scale from dealer-owned buying groups, so the big box's cost advantage is narrower here than in most retail. But this is not a small shop: a quarter of Canadian establishments in the group employ twenty or more people. Entry is by buying an existing yard, and the price includes land, a delivery fleet and a full inventory; after closing, the owner finances his contractors' receivables through a housing cycle he does not control. That is the nearest thing to a cut, and the screen does not find it decisive. A full study would have to test what banner dealers actually change hands for, contractor-account losses through a rate cycle, and whether the succession wave shows up in buying-group membership. The ERP sold to these dealers is screened separately.

Two unlike shops share this code. The garden centre is the one people want to open, and what is attractive about it is real: plant knowledge and locally grown stock are things a big box does badly, and the customer will pay for both. The cut is the calendar. The only Canadian operating data found — a 2010 trade-magazine reader survey, small and self-selected — has 45% of a garden centre's sales landing between March and May and 7% in winter, with 27% of respondents reporting margins of 1–4% and another fifth answering 'Profit? Are you kidding me?' [C]. A year of fixed cost — land near enough to households to draw them, greenhouses, heat, a core staff — is carried on roughly ten spring weekends, in perishable stock, against weather; and in those same weeks every grocer and home centre sets up a parking-lot tent selling bedding plants at or near cost. The survivors are overwhelmingly families that already own the land and often grow their own stock — which is to say the business works once the capital is sunk and written off, and not before. The outdoor power equipment dealer runs on different rails: the right to sell a major line is granted dealer by dealer by the manufacturer, and the service bay rather than the showroom earns the living. That half was not examined beyond noting it, and it sits closer to the landscaping-contractor trade already screened at 561730 than to retail. The dealer ERP on this branch is screened separately.

Net margins of one to two percent leave no room for a purchasing disadvantage, and an independent buys through a wholesaler at prices the national chains beat directly with the manufacturer. The defensible versions are geographic — a town the chains will not enter — or specialised ethnic and natural grocery, both of which are real and neither of which scales. The grocery technology stack is screened separately at 4451.

A butcher, cheesemonger, bakery or ethnic grocer is small-format and genuinely reachable — and that is most of what is attractive about it. The cut is the shape of the money: a lease, a fit-out, refrigeration and opening inventory are all spent before the first sale, the inventory spoils, and the recovery comes back one basket at a time from a catchment measured in minutes of walking or driving. Gross margin on food retail does not stretch far enough to service that fit-out quickly, and the landlord — not the operator — captures the value of a location that works, at renewal. This is the same capital-shape cut applied across the retail records in this research, and nothing about specialty food exempts it.

In most provinces the licence is the asset and it is capped, transferable and expensive — which makes entry an acquisition priced on the licence rather than the store. Wholesale price is set by a provincial board that is also, in several provinces, the competing retailer. Margin is therefore administered rather than earned, and the operator's only levers are location, hours and selection.

The showroom has become a fitting room for a purchase completed online, and the national chains and manufacturers price directly against the independent while carrying the delivery and warranty infrastructure a single store cannot. Inventory is bulky, financed and slow-turning. The defensible version is installation and service on appliances someone else sold — a trade business, not a retail one.

The pre-screen called furniture retail 'structurally squeezed'. The filings say otherwise, and that is the problem. Leon's Furniture — Leon's, The Brick and 300 stores — reported 2025 revenue of $2,573.7M, up 3.0%, same-store sales up 3.0%, furniture up 6.3%, and a gross margin of 45.04%, 65 basis points better than the year before, with net income of $157.0M [A]. In 2024 Fairfax took Sleep Country private at Cdn$35.00 a share, US$881M for the equity of a chain of more than 300 locations [A], a business the buyer describes as holding about 40% of the Canadian market [B]. Neither incumbent is in retreat. The existing 449 record argues the showroom has become a fitting room for an online sale; this record adds that the Canadian chains are profitable, growing in furniture and freshly capitalised — they are not the soft target that story implies. An entrant reaches a 45% gross margin only if it lands product at comparable cost, which means joining a buying group, and then carries the same fixed stack — showroom, warehouse, two-man delivery — across one store instead of three hundred. The incumbents also sell credit, warranties and insurance on top of the sofa, which a single store can only broker. Where Leon's itself is shrinking — mattresses and electronics, down mid-single digits [A] — is no opening either: that is the ground the specialist mattress chain and the online bedding brands already hold. Floor covering and window treatment retail (449121, 449122) is installation-led, closer to a trade than a shop, and was not examined here.

The pre-screen says electronics retail 'has largely moved online'. Best Buy Canada says something more exact. The International segment of Best Buy — which is Canada and nothing else, 142 stores — took US$3,413M of revenue in fiscal 2026, up 3.7%, with comparable sales up 2.3% [A]. So the store-based format has not disappeared. What it earns is the finding: a gross margin of 21.6% and an adjusted operating margin of 3.4% [A], at national scale, with vendor funding, its own distribution centres and a services arm. Computing and mobile phones were 49% of the mix; services were 6% [A]. An independent selling the same televisions, laptops and phones buys them for more than Best Buy does, sells them at the price on the customer's screen, and starts from a gross margin below 21.6% with nothing beneath it. Nothing on the shelf is unavailable elsewhere at the same price the same day — that is the cut, and it holds even if rent and capital were free. Appliances, where independents have traditionally held on through delivery and installation, fell 5.1% on a comparable basis at Best Buy Canada [A], and the existing 449 record already makes the point that the defensible version of that business is installation and service, which is a trade, not retail. Where the 78% of Canadian establishments with fewer than ten staff actually earn a living — phone repair kiosks, carrier dealerships, custom audio-video installation — was not examined; those are different propositions wearing this code.

This is not an entry market, and the record exists to say why in one event rather than in adjectives. Hudson's Bay filed under the CCAA on 7 March 2025 with more than $1.1B of secured debt and about $3M of cash; liquidation sales began within weeks, every store was closed by 1 June 2025 and 8,347 employees were terminated [B]. No going-concern buyer emerged for the chain. What sold was the name: Canadian Tire paid about $30M for the stripes, the coat of arms and the private labels [B]. An attempt to take up to 28 of the leases and relaunch a department store was blocked by the court for 25 of them [B]. The mechanism is that the format's two jobs — assembling brands under one roof and anchoring a mall — have both gone elsewhere: brands sell direct, and landlords now prefer to re-let an anchor box in pieces than to a department store. The pre-screen cut this on capital, and the StatCan shape agrees that it is a scale format — 86% of establishments employ a hundred or more. But capital is not what failed here: Hudson's Bay had 350 years of brand, the locations and a billion dollars of credit, and the category still shrank out from under it. The definition's other door — operating a department on concession — makes the entrant an unsecured creditor of the host, which in this proceeding was the class standing behind more than $1.1B of secured debt. What is reachable is selling to the survivors: the merchandise planning software on this branch is screened separately.

The title says warehouse clubs, and nobody needs a screen to rule out building a Costco. The reachable proposition sits in the residual: the dollar store and the small general merchandise shop, which is what most of the 6,987 Canadian establishments here are — a third of them fall in the 10–19 employee band, the footprint of one chain variety store. The incumbent in that niche is about as far from vulnerable as a retailer gets. Dollarama reported fiscal 2026 sales of $7,255.8M, up 13.1%, Canadian comparable sales up 4.2%, a gross margin of 45.0% and an operating margin of 26.7%, with 1,691 Canadian stores, 75 of them net new in the year [A]. A 26.7% operating margin in discount retail is not a pricing accident. It comes from buying direct from factories in container volumes, engineering products to fixed price points, and running a store design with almost no labour in it — none of which a single store can imitate. An independent buys the same category of goods through importers and wholesalers at a multiple of Dollarama's landed cost and then has to sell them beside a chain that is still opening a new store every five days. The other big name in the group, Canadian Tire, is dealer-operated, but dealerships are awarded by the corporation rather than bought on an open market and were not examined. The one exception the screen leaves standing is the rural general store with a catchment too small for any chain: a real business, a geographic monopoly, and one that has to be found town by town rather than sized.

Fuel is the loss leader, the store is the margin, and scale decides both. Couche-Tard runs close to 17,300 stores on more than $76B of fiscal 2026 revenue and opened 130 in the year — a buying power in fuel supply, private label and card processing that a single site cannot approach. The consolidation is also removing the independents' comparables: Sunoco completed its acquisition of Parkland and its roughly 4,000 locations on 31 October 2025. Add environmental liability on underground tanks, which is the specific risk that makes a cheap site expensive. The forecourt software at this code is screened separately.