The $12 Billion That Isn’t There
What the land value capture line in the McGill TRAM financial model actually rests on — and why a number doing the heaviest lifting in ALTO’s only public financial model is a planning placeholder, not a financing prospect.
The McGill TRAM financial model assumes that land value capture — the public capture of property-value uplift around new stations — will contribute $12 billion toward ALTO’s capital cost, reducing the amount that must be borrowed from roughly $53 billion to $41.23 billion.
This brief traces that figure to its origin, tests it against the international precedents the model invokes, against the realised Canadian record, against the legal authorities ALTO actually holds, and against the timing of when capture revenue could plausibly arrive. On every test, the $12 billion comes apart.
The $12 billion land value capture line is reverse-engineered from a 15-percent rule of thumb, not built from any property analysis. It contains no parcel-level valuation, no station-area market study, no comparable transactions, and no discounted cash flow.
A defensible figure for the present value of plausible station-area capture is in the low single billions — well under 5 percent of capital cost — and it accrues over decades rather than during the construction window when borrowing must actually be priced. The line is the difference between a model that reads as “tolerable on paper” and one that reads as “permanently subsidised.”
A percentage, not a forecast
The $12 billion originates in the McGill TRAM financial analysis, where it is described as land and real estate development gains “equivalent to roughly 15 percent of the total cost.” Fifteen percent of the assumed $79.8 billion capital cost is $12 billion. The ratio is asserted; the dollar figure follows arithmetically.
That is the whole of its derivation. The report contains no parcel-level valuation, no station-area market analysis, no comparable transaction work, no discounted cash flow of expected development revenues, and no sensitivity analysis. Change the cost assumption and the “capture” number moves with it — without any change to the underlying property economics, because there are no underlying property economics in the figure to begin with.
The line is also structurally load-bearing. Remove it and the borrowed principal rises from $41.23 billion to roughly $53 billion. At the model’s own 8 percent rate over 50 years, that adds about $1.05 billion a year in debt service. The companion brief concedes the consequence directly: its “No LVC” scenario requires average annual subsidies of $2.12 billion and never reaches self-sufficiency by Year 50.
The international examples do not transfer
The TRAM brief grounds its capture case on three precedents — Hong Kong’s West Kowloon, an Australian East Coast HSR pre-feasibility study, and California’s High-Speed Rail. None is institutionally analogous to the ALTO corridor.
Hong Kong West Kowloon
The only case with realised capture at scale: a single super-prime tower site sold for HK$42.2 billion. But Hong Kong’s land is overwhelmingly state-owned under a colonial leasehold system, and the government grants development rights as a primary fiscal instrument. It bears no resemblance to Peterborough, Trois-Rivières, Laval, or even Ottawa-Gatineau.
Australia East Coast HSR
The cited evidence is a 2022 preliminary investigation with a near three-fold range ($43–126 billion), for a project that remains unbuilt. Citing an aspirational range from an unconstructed project as proof that ALTO can capture $12 billion is circular reasoning.
California HSR
Cited for proposed tax-increment financing concepts. After fifteen-plus years and over $13 billion of spending, California HSR has captured essentially zero, while costs escalated from $33 billion to over $128 billion. It is a cautionary precedent, not a supporting one.
Two precedents the brief omits are more directly relevant. The UK’s HS2 explicitly considered capture and recovered a negligible fraction of capital cost — property values along the route fell on construction blight, and the government spent more on compensation than it recouped. Brightline in Florida, the closest North-American analogue with vertically integrated real-estate interests, is in distress on its Private Activity Bonds despite favourable conditions: no winter operations, sustained population growth, and no expropriation politics.
The most relevant evidence is Canadian — and it comes from a source the federal government itself supports. A 2023 study by the University of Toronto’s Infrastructure Institute, prepared for and supported by the Canada Infrastructure Bank, surveyed the realised Canadian record:
- Per-deal ceiling: realised Canadian capture deals — joint development and surplus land sales — have typically raised $30 million to $110 million, with only the largest sales in the most expensive markets exceeding that band.
- Corridor analogue: Montréal’s REM, the closest comparable, raised a $512 million station-area contribution — covering just 7.4 percent of the project’s $6.9 billion cost, itself well below a 2014 estimate of up to 35 percent.
- Single station: Vancouver’s Capstan Station, described as having near-ideal conditions for capture, raised only $32 million over nine years.
- The Hong Kong verdict: the same CIB-supported study attributes West Kowloon’s success to a combination of factors unique to Hong Kong, and concludes the model is fundamentally different from most capture models.
A CIB-supported source thus reaches the same conclusion this note does: the marquee precedent does not transfer, and realised Canadian capture operates two to three orders of magnitude below the $12 billion line.
The authorities required do not exist
Capture at the scale TRAM assumes requires legal authorities ALTO does not have and that no level of government has proposed. Property and land use are provincial jurisdiction. Municipal zoning, development charges, and the property tax base lie outside federal control. There is no Canadian equivalent of U.S. tax-increment financing as a station-area capture tool, and Ontario’s closest analogue — Section 37 / community benefits charges — generates modest, parcel-by-parcel sums and has been further constrained by recent provincial reform.
A structural obstacle compounds the jurisdictional one. The same CIB-supported study identifies fragmented land ownership as a core constraint: unlike Hong Kong’s state leasehold system, prime station-adjacent land in Canada is held by many separate owners. ALTO’s catchments — especially built-out central areas like Toronto Union and Montréal Central — are precisely this kind of fragmented holding, where capturing uplift at scale would first require slow, costly, politically fraught land assembly.
The brief’s recommendation that government “empower Alto to lead development and value capture within 2 km around the stations” implies development authority over roughly 88 km² of station catchment — about 12.6 km² around each of seven stations. No mechanism in Bill C-15, the Cadence consortium structure, or any published ALTO document contemplates this. The Bill C-15 expropriation provisions are scoped to the right-of-way, not to station catchments; acquiring 88 km² would be a separate expropriation programme of significant scale, with compensation costs the model never nets against the $12 billion gross.
Housing and TOD intent does exist in the procurement. A federal housing and TOD presentation to bidders — released under access to information — sets out a four-pillar housing strategy and contemplates that Canada would acquire project lands and explore station-hub development with the developer partner. That intent carried forward into the ALTO procurement, which required a high-speed rail proposal from all bidders.
But the presentation is explicitly provisional throughout: “provisional guidelines,” requirements “to be refined,” an affordable-housing threshold “to be determined.” It attaches no budget, no land-assembly cost, no carrying-cost provision, and no capture-revenue target — and it describes a federal-acquisition-then-explore model that is the opposite of ALTO-led capture across catchments. The procurement confirms an intention to pursue TOD; it does not supply the costed mechanism on which the $12 billion depends.
Even a generous bottom-up envelope falls short
The TRAM model is corridor-wide and does not allocate the $12 billion to specific stations. Spread across the seven announced stations, it implies an average of roughly $1.7 billion per station. A station-by-station review of catchment characteristics shows how implausible that is — most of the corridor’s stations serve small markets or are already built out, so most uplift would accrue to existing landowners rather than to a public capture programme.
Toronto:$1.0–2.0B — incremental only
Montréal:$1.0–2.0B — incremental only
Note:Most uplift to existing owners
Ptbrgh:$0.1–0.3B — CMA ~90k
T-Rivières:$0.1–0.3B — CMA ~85k
Québec:$0.3–0.8B — heritage limits
Ottawa:$0.5–1.5B — core receding
Laval:$0.3–0.8B — greenfield TOD
Total:$3.3–7.7B gross envelope
Summed, a generous corridor-wide envelope — gross, undiscounted, spread over 20–30 years — reaches $3.3 to $7.7 billion. Even its upper bound falls short of the $12 billion the model requires. And that envelope still assumes full institutional empowerment of ALTO as a development corporation, which is not on the table, while ignoring both the carrying cost of land assembly and the compensation cost of catchment-area expropriation.
Most of the value, in present terms, is fictional
The model treats $12 billion as available during construction, to reduce the principal borrowed. In practice, capture accrues over decades. Land sales and development gains around new stations typically materialise five to fifteen years after a station opens, and construction on the full corridor is projected to take well over a decade. A realistic capture stream would produce most of its value between roughly 2040 and 2060 — long after the borrowing is priced.
Discounted at the model’s own 8 percent rate, $12 billion realised over Years 15–35 has a present value of only about $3 to $4 billion at financial close. That is the figure that can actually reduce the borrowing requirement. The remaining $8 to $9 billion in the arithmetic is, in present-value terms, fictional — and the construction debt still has to be priced against the full undiscounted principal.
One line, three improvements, all of them collapse
The $12 billion capture line is the single most important — and least scrutinised — financing assumption in the only publicly available financial model for ALTO. It does three things at once, and all three depend on the same unsupported number.
It cuts the borrowed capital
From roughly $53 billion to $41 billion — the difference being the $12 billion the model assumes capture will supply.
It pulls self-sufficiency forward
From “never” to Year 48. Without the capture line, the companion brief’s own “No LVC” scenario never reaches self-sufficiency by Year 50.
It lowers the annual subsidy
From $2.12 billion to $1.23 billion a year on average — the gap between “tolerable on paper” and “permanently subsidised.”
Professor El-Geneidy has said publicly that the model uses “very generous” assumptions, particularly on demand, and that breakeven “can happen … but it requires a lot of work from the government to make it happen.” The capture assumption falls into the same category. Even on its own optimistic terms, the model shows cumulative subsidies of $61.6 billion through Year 50, on top of the initial $26.6 billion federal investment — a combined taxpayer exposure of $88.2 billion before any recovery from project revenues.
A placeholder, not a pillar
The $12 billion figure should be treated as a planning placeholder rather than a financing prospect. Any business case, public communication, or appraisal that relies on it as a stable revenue pillar is overstating ALTO’s financial position by an order of magnitude — at the present-value point that matters most, the moment construction debt is priced. The defensible number is in the low single billions, it arrives over decades, and it cannot be borrowed against today.