Government-Tenanted Commercial Real Estate in Canada

The premise most investors get wrong

The instinctive case for a government-tenanted building is that the tenant cannot go bankrupt. It is true, and it is close to irrelevant. Insolvency was never the risk. The risk is the lease ending early, and on that question the tenant's name tells you almost nothing.

The United States demonstrated this at scale in 2025. Federal leases there commonly run a firm term followed by a soft term in which the government holds a termination option exercisable on a few months' notice — a clause long treated as dormant. When the U.S. General Services Administration began exercising it, notices went out on hundreds of leases covering roughly nine million square feet. Research published by academics at Yale and RIT measured the consequences on affected properties: net operating income fell by an average of 5.6 percent, and the first-loss tranches of the associated commercial mortgage-backed securities repriced by roughly 3.8 percent. The most instructive finding was indirect — private leases within five miles of a notified building fell 10.1 percent, a larger effect than the direct one.

The pricing evidence points the same direction. Among the small number of U.S. government sale comparables published in 2026, a federally-tenanted building with nineteen years remaining traded at a capitalization rate roughly half a percentage point inside the national single-tenant office average. A state-tenanted asset with about seven and a half years remaining priced several points wider. Same category of tenant. Entirely different pricing.

That is the frame this page applies to the Canadian market, and it is the reason the sections below lead with lease structure rather than covenant.

Where Canadian federal leasing demand actually sits

Public Services and Procurement Canada is custodian of roughly a quarter of the federal government's real property and houses more than 300,000 public servants from over 100 departments. It owns comparatively few buildings and leases a great many locations, and it publishes its requirements.

To establish where that demand actually falls, this firm analysed the federal government's own open tender-notice data covering fiscal years 2024‑25 through 2026‑27, filtered to Public Services and Procurement Canada solicitations for leased space. Thirty-six non-residential lease solicitations were identified. The pattern is consistent enough to be worth stating precisely:

Analysis by JD Capital of Government of Canada open tender-notice data published via CanadaBuys, covering fiscal years 2024‑25 to 2026‑27 (the latter partial). Term and size statistics are drawn from the eighteen notices carrying full requirement text; the remainder direct respondents to attachments. Figures describe published solicitations, not completed transactions.

The locations named across those notices include Labrador City and Grand Falls‑Windsor in Newfoundland and Labrador; Penticton and Maple Ridge in British Columbia; Collingwood, Petawawa, Smiths Falls, Carleton Place, Welland and Kingston in Ontario; Victoria County in Nova Scotia; Charlottetown; Rothesay and the Moncton region in New Brunswick; and Trois‑Rivières in Quebec. Seven of the thirty-five sat in Toronto, Montreal, Laval or Gatineau.

The significance is the intersection of size and geography. A requirement of roughly seven thousand square feet on a ten-year federal lease, in a community of modest population, is too small to interest an institutional buyer and too specialised to attract a passive one. It is also, in most of those communities, among the better credit covenants available in the local commercial market.

How a private landlord actually reaches the tenant

The process is public, and it is gated. It rewards preparation and penalises arriving late, which is precisely why it stays under-competed.

Two implications follow. First, monitoring is not optional — the relevant federal notices are published under the real estate services procurement category, and a landlord who checks quarterly will miss most of them. Second, the qualification burden is a barrier to entry, and barriers to entry are why a market stays priced for the participants willing to clear them.

There is a further structural feature worth understanding. Canada's internal trade agreement, which governs public procurement, expressly excludes from its tendering obligations the "acquisition or rental of land, existing buildings, or other immovable property." Governments here are therefore under no trade-law requirement to competitively tender a real property lease. Public postings reflect policy and practice rather than obligation — which means a meaningful share of government leasing is transacted through brokers, incumbent landlords and direct approach. This is a relationship market by legal design.

The shrinking-footprint question

Any honest treatment of this asset class has to address the argument against it: that the federal government is reducing its office footprint and will not need the space. The evidence is genuinely mixed, and both directions deserve to be stated.

The case for caution. Budget 2024 committed funding to cut the federal office portfolio roughly in half by 2034, explicitly including money to accelerate lease terminations. That is a stated policy intention to occupy less leased space.

The case that it is overstated. The Auditor General of Canada found the portfolio shrank by less than two percent between 2019 and 2024, attributing the shortfall principally to a lack of funding and departmental reluctance. The department's own published projection now anticipates reaching roughly a third rather than half, with the largest annual reductions deferred to the 2030s. And in 2026, as return-to-office attendance requirements increased, the department stated publicly that certain departments would need more workstations or more space in certain locations, with solutions including renewing existing leases and possibly acquiring additional space. The federal public service has grown by more than fifty thousand positions since 2020.

The countervailing force. Budget 2025 targets a reduction of roughly forty thousand public service positions by 2028‑29 as part of a broad expenditure review. Fewer people eventually means less space.

These pressures work against each other and the net effect is genuinely uncertain. The defensible conclusion is narrower than either camp would like: the leased portfolio contracts through non-renewal at expiry rather than through mid-term exits, gradually, and behind schedule. That is an argument for underwriting renewal probability building by building — not for assuming either a wave of vacancies or a guaranteed tenant.

Two further facts belong in any assessment. The department's assistant deputy minister for real property services stated publicly in 2024 that roughly half of the federal government's leases expire within five years — a concentration of renewal decisions rather than a slow roll. And in 2026 the federal government agreed to purchase a fourteen-storey Ottawa office building it had occupied as a tenant, for a reported $148.2 million, describing the move as transitioning from tenant to owner. Where the government elects to buy rather than renew, the landlord loses the tenancy permanently. Both facts cut against a passive hold assumption.

What the covenant is worth

No Canadian brokerage publishes a capitalization rate series for government-tenanted property. CBRE, Colliers, Cushman & Wakefield, Avison Young and Altus all publish by asset class and market; none publish a government or credit-tenant category. Any figure presented as "the Canadian government cap rate" is therefore either derived or invented, and this page will not print one.

One public benchmark does exist. True North Commercial REIT holds a portfolio in which roughly seventy-three percent of revenue derives from government and credit-rated tenants. Its interim financial statements for the period ended 30 June 2026 disclose a weighted average terminal and direct capitalization rate of 6.83 percent, within a range of 5.50 to 9.50 percent, and a weighted average discount rate of 7.44 percent. Those are IFRS portfolio valuation inputs on a largely office portfolio — not transaction evidence — but they are the only published Canadian figures attached to a majority government-tenanted portfolio. For context, CBRE's national all-property average was 6.58 percent in the second quarter of 2026.

The same filing carries the more important disclosure. That portfolio's weighted average lease term is 4.2 years. Its federal government leases averaged 3.2 years remaining; its provincial leases ranged from 1.9 to 3.9 years. And in the fourth quarter of 2025 the REIT recorded an early lease termination at an Ottawa property, booking $13.9 million in termination income and subsequently classifying the asset as held for sale. Revenue for the following period would have risen without that property; instead it declined.

That disclosure requires no reference to events in the United States, and it makes the point directly: government tenants in this country do leave before the outside date, and when they do, the building becomes the landlord's problem. The section that follows sets out what that looks like on the ground.

Firm term, not headline term

Recent federal solicitations reviewed in the analysis above increasingly embed mid-term termination rights. A fifteen-year requirement in Labrador City carried a right to terminate after the twelfth anniversary. A fifteen-year requirement in New Brunswick carried the same structure. A ten-year requirement in British Columbia carried a termination right that reduces the reliable term to eight years. Notably, this language appears in solicitations published from late 2025 onward and did not appear in the earlier notices in the dataset — a pattern suggestive of a drafting trend rather than proof of one, given that half the notices do not publish full text.

The underwriting consequence is direct. A fifteen-year lease terminable at year twelve is a twelve-year asset with a three-year option attached, and it should be valued as one. Where a residual is being underwritten beyond the firm term, that residual belongs to the sensitivity analysis, not the base case.

When the tenant leaves

The strongest argument against this asset class is not theoretical, and the honest way to present the category is to state the failure mode precisely. Canadian examples are readily available, and several are current.

The purpose-built building in a small town is the hardest case. A sandstone armoury in Amherst, Nova Scotia, built in 1915, lost its regiment in 2015 when the unit relocated. The federal government declared the building surplus in 2016. In April 2024 Ottawa announced it would be offloaded for housing. In the summer of 2025 the Department of National Defence closed it outright over mould. As of early 2026 the redevelopment remained stalled, a proposed conversion was costed at roughly $20 million, and the town's own $50,000 feasibility study had been budgeted but not commissioned. The building has now been empty for a decade. The local member of Parliament observed that one could probably find at least one surplus federal building in every riding in the country.

That outcome was, in effect, predicted by the Auditor General, who warned that prioritising the disposal of properties suitable for housing "could delay the disposal of properties less suitable for housing and risks resulting in an increase in maintenance and operating costs." The least convertible buildings sit longest. The same office reported that disposing of a federal building takes, on the department's own account, an average of nine years.

Small-market federal exits are happening now. In January 2026 Agriculture and Agri-Food Canada announced the closure of seven research operations — research and development centres at Guelph, Quebec City and Lacombe, and satellite farms at Scott and Indian Head in Saskatchewan, Nappan in Nova Scotia and Portage la Prairie in Manitoba. Roughly 1,050 employees received notices. The Lacombe site had operated for 119 years; Indian Head's for 140. These are precisely the community types where a federal tenancy represents an outsized share of the local institutional market.

The effect is measurable even in the capital. Ottawa's office vacancy rate rose from 13.2 percent at the end of 2025 to 14.3 percent by the end of March 2026 on CBRE's measure, with roughly 413,000 square feet of negative absorption in a single quarter driven substantially by two government departures — one federal department vacating about 114,000 square feet on Carling Avenue, and a federal agency vacating about 112,000 square feet on Camelot Drive. One local brokerage principal estimated the federal government had given up approximately one million square feet of privately leased downtown Ottawa space over five years. Public sector employment represents close to a quarter of the Ottawa market, and the region recorded roughly 30,500 job losses in the twelve months to February 2026.

Provincial assets strand the same way. New Brunswick closed its provincial visitor information centres between 2018 and 2020. One of them — a thirty-five acre highway site near the Maine border — was sold as surplus for $323,000 in 2022, was never developed, and was subsequently relisted at $599,900, below its 2024 assessed value of $742,200.

None of this makes the category uninvestable. It makes single-asset, single-purpose exposure in a thin market the wrong way to hold it. The buildings that survive a tenant departure are the ones with a plausible second tenant, a conventional floorplate, and a location that would have supported commercial use regardless of who occupied it. That test — would this building let without the government — is the one worth applying before the covenant is discussed at all.

Adjacent tenant classes worth understanding

"Government tenant" covers several structures whose credit differs materially, and confusing them is a common error.

• Federal departments under a head lease. The covenant is the Crown, but day-to-day lease administration is largely outsourced — a landlord's operational counterparty is generally the government's real property services contractor rather than the department itself.
• Canada Post. Its disclosed lease obligations are substantial and long-dated, with the large majority of building right-of-use assets held on net leases. The credit picture is more complicated: the corporation recorded its largest loss on record in 2025, and the long-standing moratorium on rural post office closures was lifted in September 2025, unfreezing several thousand rural corporate outlets. The franchise-operated channel — which is where growth is directed — places the covenant with the retail operator, not the postal service.
• RCMP detachments. Under municipal police service agreements, the municipality generally provides and pays for the detachment building. The covenant is therefore usually municipal, not federal. Contract policing agreements run to 2032.
• Privately operated service counters. Several provinces deliver front-counter services through private operators rather than public employees. In Ontario, 195 of 276 in-person service centres were privately operated as of the province's most recent published analysis; Alberta delivers registry services through more than two hundred independent agents; Manitoba through several hundred insurance agents. In each case the tenant of any retail premises is a private business holding a provincial contract — not the Crown.

The last category deserves particular care, and a caution runs against it. Ontario has placed a number of service centres inside an existing national retailer's stores, and the responsible minister described the resulting savings as arising largely from not paying leasing costs. That is a provincial government stating plainly that the object is to exit leased premises by folding counters into space a large retailer already occupies. The direction of travel in that channel is consolidation into existing big-box retail, not new demand for small-bay space. The operator's revenue may be underwritten by a government contract, which is a genuine strength, but the lease covenant is the operator's own — and the contract can move to a different host. Those are not the same credit, and diligence should treat them separately.

Why Canadian sale-leasebacks are rare

Investors familiar with the U.S. market often ask why Canadian governments do not sell and lease back their buildings. The answer is structural, and it is worth knowing before pursuing the idea.

Canada's Income Tax Regulations exclude from "exempt property" a building leased primarily to a Canadian government, municipality or other public authority that owned the building before the lease commenced. The building becomes specified leasing property, and the buyer's capital cost allowance is restricted accordingly — removing much of the depreciation shelter that ordinarily supports sale-leaseback economics. Separately, public sector accounting guidance treats a sale-leaseback that results in a leased tangible capital asset as, in substance, a financing arrangement: the selling government generally re-recognises the asset at its prior carrying amount and does not report the holding gain. In Ontario, municipal finance rules compound the problem by excluding real property sale proceeds from the revenue base used to calculate borrowing capacity, while the resulting lease payments consume it.

The historical record matches the incentives. The federal government completed one significant office sale-leaseback, in 2007. One province has completed a comparable transaction, in 2008; its auditor general later concluded that elementary rules of sound management and prudence had not been followed, finding that concessions granted exceeded the price improvement obtained and that the transaction had been driven by a fiscal year-end. An Ontario programme announced in 2012 was never completed. No completed Canadian municipal sale-leaseback of a civic building could be identified. This is not a market awaiting discovery; it is one the tax and accounting rules have largely foreclosed.

How this firm approaches the category

Government and government-adjacent tenancies fall within the firm's mandate where the lease structure supports the price — not because the tenant is a government. In practice that means underwriting to the firm term rather than the headline term; treating renewal beyond the outside date as sensitivity rather than base case; assessing whether the building has an alternative use and an alternative tenant in its local market if the lease is not renewed; and treating a single-tenant government asset in a small market as what it is, which is concentrated exposure requiring a debt structure that can absorb a vacancy.

The transaction size that recurs throughout the federal data above — several thousand square feet, ten-year term, secondary and tertiary markets — sits within the firm's stated acquisition parameters, and the absence of published cap rate coverage in most of those markets is a reason for care in underwriting rather than confidence in pricing. The standards governing every acquisition are set out in the firm's published mandate, and the structures through which accredited investors participate are described in the companion guides to private commercial real estate in Canada and preferred equity.

Readers evaluating any government-tenanted asset — offered by this firm or any other — should ask four questions before the cap rate: what is the firm term, what termination rights exist and on what notice, what happens to this building if the tenant does not renew, and who exactly is the counterparty on the lease. A sponsor who leads with the tenant's name rather than the term sheet has answered the last question by avoiding it.