Expanding to the US: Branch or Subsidiary? The Branch Profits Tax, the Liability Wall, and How the Choice Actually Gets Made

Short version: Branch Profits Tax Explained for Canadian Companies

The branch-versus-subsidiary question arrives dressed as a tax decision and mostly isn't one, because the US designed the branch profits tax to equalize the two. The structures first. A branch is the Canadian corporation itself operating in the US — registered in its states, filing Form 1120-F on the income effectively connected with its US business (or attributable to its permanent establishment under the treaty), with no separate legal person: US creditors' claims reach the whole corporation, Canadian assets included. A subsidiary is a US corporation (the C-corp from the entity-selection playbook) owned by the Canadian parent — its own taxpayer filing Form 1120, its own liability perimeter, dividends home at the treaty's 5% parent rate. The tax comparison: the subsidiary pays US corporate tax and then 5% withholding on dividends to the Canadian parent; the branch pays US corporate tax on its ECI and then the branch profits tax — a second-layer tax on the branch's earnings not reinvested in US assets (the dividend-equivalent amount), which the treaty caps at the same 5% and softens with a cumulative exemption on the first C$500,000-equivalent of branch earnings — the small-expansion sweetener that makes early-stage branches genuinely cheaper. Losses run opposite: a branch's early US losses consolidate into the Canadian corporation's results immediately (the startup-phase argument for branches), while a subsidiary's losses are trapped in the US entity awaiting its own profits; foreign tax credits and surplus accounts then govern how each structure's US tax integrates at home, with the subsidiary's dividends flowing through Canada's exempt-surplus system cleanly for active business income. So the tax rails are close to parallel — and the decision gets made on everything else: liability (the branch exposes the Canadian mothership to US litigation; the subsidiary is the wall, and US commercial reality prices it — landlords, lenders, and enterprise customers contract more easily with a domestic entity); state administration (a branch drags the Canadian corporation itself into state registrations, returns, and franchise taxes everywhere it operates); financing and credibility (US banks and partners know what a Delaware corporation is); employees and equity (US benefit plans and stock plans want a US employer); privacy (the branch's 1120-F exposes the Canadian corporation's treaty-based positions and, in states, sometimes more); and the exit (buyers acquire US subsidiaries every day; carving a branch out of a Canadian corporation for sale is a restructuring project). The pattern that follows is the standard arc: test the market entity-free behind the treaty (the nexus playbook's protective-filing stage); branch briefly where early losses and the C$500,000 exemption genuinely matter and liability exposure is modest; incorporate the subsidiary at the first employee, lease, or enterprise contract — and paper the conversion (branch assets into the new corporation) with the rollover and valuation care that turns a growth milestone into a filing rather than a taxable event.

Key takeaways

• The tax gap is engineered small: subsidiary = corporate tax + 5% dividend withholding; branch = corporate tax + branch profits tax at the treaty's 5% on unreinvested earnings, with the treaty's cumulative exemption on the first ~C$500,000 of branch earnings favoring small operations.
• Losses favor the branch early: immediate consolidation into the Canadian corporation versus losses trapped in a startup subsidiary — the honest tax argument for branch-first expansions with a planned conversion.
• Liability favors the subsidiary always: the branch puts the whole Canadian corporation in every US courtroom; the subsidiary is the wall, and most boards price that above the tax rounding.
• Commercial gravity favors the subsidiary: customers, landlords, lenders, benefit plans, and eventual acquirers all prefer contracting with a US corporation — the structure decision is a sales and HR decision wearing tax clothes.
• Administration differs in kind: the branch registers the Canadian corporation state by state and files 1120-F with its treaty positions; the subsidiary contains US administration inside a US entity — with Form 5472 and the transfer-pricing file arriving as the price of the intercompany relationship.
• The conversion is a project, not a formality: branch-to-subsidiary incorporation transfers assets, contracts, and people — executed with rollover planning, valuations, and both countries' filings coordinated, ideally at a moment chosen for it rather than forced by a lawsuit or a deal.

The staged-entry playbook

Stage one, sell in: no US entity; the treaty and the protective 1120-F carry the income tax answer while the sales-tax registrations run on their own thresholds. Stage two, plant lightly: where a physical toehold and early losses coincide — the demo facility, the first warehouse — a branch can serve for a defined window, the C$500,000 exemption absorbing the branch profits tax while losses consolidate home; the window is written down in advance with its conversion triggers. Stage three, incorporate: at the first US employee, lease, or enterprise MSA, the subsidiary forms, the branch (if any) rolls in, transfer pricing papers the new intercompany flows, and the structure the buyers, banks, and benefit plans expect is simply there. Companies that skip stage discipline pay in one of two currencies: the branch that lingered into litigation exposure, or the subsidiary formed at first curiosity that trapped three years of losses behind its own wall.

Worked example

A Burlington industrial-equipment maker expands south over four years. Year one: US$1.8M of sales shipped from Ontario, two trade shows, a rep network — no entity; the protective 1120-F discloses the no-PE position; four state sales-tax registrations run on thresholds. Year two: a Cleveland demo-and-service depot with two Canadian technicians rotating through — a branch, deliberately: the depot loses US$300,000 in its build-out year, consolidating immediately against Ontario profits; the branch profits tax computation shows the cumulative exemption covering the modest early earnings; the conversion triggers (first US hire, first US-signed MSA, or month thirty) are minuted. Year three: the first US employee and a national accounts contract trip two triggers — a Delaware subsidiary forms, the depot's assets roll in under the incorporation planning both firms paper, the intercompany distribution and services agreements start the transfer-pricing file, and Form 5472 joins the calendar. Year four: the subsidiary carries US$6M of revenue, a US benefits plan, and a term sheet from a strategic acquirer whose diligence request list assumes — correctly, now — that the US business lives in a US corporation. The tax paid along the arc barely differs from any alternative path; the liability, the losses, and the exit all landed where the staging put them.

Official sources

"A foreign corporation files this form to report their income, gains, losses, deductions, credits, and to figure their U.S. income tax liability." It also transmits treaty-based return positions (Form 8833) and calculates the branch profits tax. — Internal Revenue Service, About Form 1120-F, https://www.irs.gov/forms-pubs/about-form-1120-f

"The business profits of a resident of a Contracting State shall be taxable only in that State unless the resident carries on business in the other Contracting State through a permanent establishment situated therein." — Canada-United States Tax Convention, Article VII, https://www.canada.ca/en/department-finance/programs/tax-policy/tax-treaties/country/united-states-america-convention-consolidated-1980-1983-1984-1995-1997.html

Practitioner note

Branch versus subsidiary is the expansion question where the tax answer is a rounding error by design and the real drivers — liability, losses, customers, exit — point different directions at different stages. Our playbook is staged entry with written conversion triggers, because the two expensive versions are mirror images: the branch that overstayed into a lawsuit, and the day-one subsidiary that spent three years hoarding losses it couldn't use.

See also: Browse every cross-border tax topic guide, organized by situation.

Next step

Fairlight prepares the US expansion structure plan — staged entry design, branch profits tax and exemption modeling, the conversion project with rollover planning, and the subsidiary-era transfer pricing and 5472 setup. See cross-border pricing or book a call.