IstariISTARIStrategyB2BGet in touch

Istari · the housing freeze · live

Almost nobody is building the home you need.

Four Canadian housing markets have nearly stopped building — not because nobody wants homes, but because the math stopped working. Pick your city and your home. We’ll show you the arithmetic that froze it, put it next to a real paycheque, and let you find the plan that thaws it — then show you what each fix strains in return.

City

You're a couple starting out in Toronto — median family income: $102,570/yr (StatCan 2023). Last month, 2,558 apartment homes were started across the Toronto area.

No team jersey. Live data from Statistics Canada, CMHC and the Bank of Canada; every figure tagged Verified Reported Illustrative. Method →

01 · the diagnosis

What it costs to build vs. what you can pay.

One new 2-bed condo in Toronto, in three honest parts. If the bar towers over the sage line, this home doesn't get built — however many people need it.

Three honest parts: building it · government’s share (fees) · the bank’s interest plus the ~15% margin a lender requires before financing. Verified fee & cost lines · Illustrative land/soft split. Levers below move these bars live.
$131,000
the government's share of this one home — fees a council can change with one vote.
$48,362
bank interest carried through construction, at the rate you set below.
$165,954
the gap. Until it closes, this home stays on paper.

That's the freeze: not a conspiracy, arithmetic. $165,954 per home stands between you and the cranes. The levers below close it from both ends.

A home that costs more to build than the people it’s for can finance is a home that doesn’t get built. That gap — not laziness, not land, not labour — is the freeze.

The proof, checked live

Apartments — started vs finished, Toronto area

connecting…
StatCan / CMHC table 34-10-0154. started (tomorrow’s supply) · finished (started 3–4 yrs ago). Verified · live
2,558
apartment homes started last month (April 2026)
378
townhouses started last month
140
detached houses started last month

The trunk line (finished) still rides the 2021–22 boom. The sage line (starting) is the future — and it has fallen off a cliff. Today's glut of finished condos and tomorrow's shortage are the same event, five years apart.

02 · the reality check

Now put a paycheque next to it.

CMHC’s standard: housing is affordable under 30% of pre-tax income. Here’s your home against the actual median incomes, live from Statistics Canada’s tax data — normalized to what a regular household earns.

Renting your 2-bed condo ($2,939/mo)34% of income
30% = affordable
Share of family median ($102,570/yr, 2023) going to rent. CMHC's line is 30%.
Buying at today's price ($660,000)needs $107,171/yr
median $102,570
Income to qualify (stress-tested, 20% down) vs. the actual median.
Buying at what it costs to build ($891,954)needs $144,836/yr
median $102,570
The freeze in one bar: the income a household needs for a builder to break even.

The people the homes are for can’t finance the homes’ construction — that gap is the freeze, restated in paycheques.

03 · find the balance

Two futures. You author one.

The clay line is the do-nothing path. Every lever re-runs the building math and rolls it to 2032 — but no fix is free. Pull one and watch the friction readout show what it strains. There’s a balanced configuration in here; when you find it, we hand you the chart of the plan you built.

Who pulls it
do nothing   your plan
Top: monthly rent for your home. Bottom: homes finished per year, citywide. Illustrative model — method in the deep dive.
$4,269
your rent 2030 (do nothing $4,269)
building restarts
+0
extra homes by 2032
Still frozen. The math sits at 81% of working. Little gets built this decade; your rent runs $1,330 above today by 2030 as the shortage lands.
No levers pulled — no strain, and no thaw. Pull one and watch what it costs.

Demand levers — help buyers afford

fed+prov

Ottawa + Ontario refund the full 13% HST on new homes under $1.5M for every buyer — roughly $80–120k off. Expires March 2027. On = make it permanent.

Done before: Ontario+federal, live — expires Mar 2027

Bank of Canada rate central bank2.25%

Hits twice: cheaper construction loans and bigger mortgages. Shown for honesty — no mayor or premier controls it.

live: —

Supply levers — make homes buildable

Cut the fees on building city council−0% · $0 off

Toronto charges $52,676 in development fees on a one-bedroom and $137,846 on a detached house — before parkland charges. Up ~370% in a decade. A council vote can cut it.

Done before: Vaughan −88% · Mississauga −50% · Metro Van rollback ’26

city+devco

Copenhagen’s trick: a public company builds transit and services first, repays by selling the now-pricier land. Counts as a ~60% fee cut without the budget hole.

Done before: Copenhagen’s metro, funded by land

Approve homes faster city hall0 mo

Toronto-area approval delays add $43,000–$90,000 per home. This slider cuts the wait.

Each month of delay ≈ $2,900 per home (BILD/Altus)

CMHC

A public guarantee on construction loans so projects break ground at half sold, not 70%. Moves borderline projects — and some risk onto the public.

Mechanism exists: CMHC already insures rental construction

Legalize more homes per lot province+city0%

Let townhouses and small apartments be built on regular lots, by right. Auckland did this to ¾ of its land in 2016 — rents landed an estimated 14–35% below the no-reform path.

Done before: Auckland ’16 · Tokyo · BC Bill 44

The plan you built

Pull some levers — then share the future you chose.

Pull some levers — then share the future you chose.

$4,269
your rent, 2030 (2-bed condo)
frozen
building restarts
+0
extra homes by 2032
0/100
system strain

Where this could be wrong.

Demand might stay weak. Immigration was cut and B.C.’s population fell in 2025 — fewer arrivals means the squeeze comes later and softer. The model assumes demand recovers slowly, not instantly.

The empty homes are the wrong homes. Most unsold units are small investor condos; the missing ones are family-sized. That means the squeeze hits families sooner — which is why we model four home types, not one.

Full method, the published algorithms, and the G7/OECD scoreboard live in the Deep Dive.

04 · the G7 / OECD scoreboard

Is this just us? No — and that’s the point.

Healthy is visible only against peers. Canada has the fewest homes per 1,000 people in the G7 — and the gap is a build target, not a mood. Verified

Japan
495
France
509
Germany
512
UK
433
US
427
Italy
587
G7 average
471
Canada
424

Reaching the G7 average means roughly 1.8 million more homes. Every lever in the cockpit is one country’s answer to closing a gap like this — none of them invented new physics.

Want the full picture?

The map, the math, and every honest trade-off.

You’ve pulled the levers. The Deep Dive has everything underneath it — the neighbourhood GIS explorer, the published algorithms, and the G7/OECD scoreboard, sourced and argued through.

Sources & verification
  1. verifiedStatCan table 34-10-0154 (CMHC Starts & Completions) — live starts/finishes by home type & CMA. www150.statcan.gc.ca
  2. verifiedStatCan T1FF 11-10-0008 / 11-10-0009 — median individual & family incomes by CMA, live. www150.statcan.gc.ca
  3. verifiedCity of Toronto DC schedule (June 2025); Metro Vancouver DCC schedules; Halton/Oakville & Burnaby fee by-laws. toronto.ca
  4. verifiedAltus 2026 Canadian Cost Guide — hard construction $/sq ft by type & city. altusgroup.com
  5. reportedUrbanation Q1-2026; TRREB / Greater Vancouver REALTORS®; Zolo / bccondosandhomes — May 2026 prices & rents incl. Oakville & Metrotown. urbanation.ca
  6. verifiedScotiabank Economics — Canada lowest in G7 at 424 homes / 1,000 (G7 avg 471). scotiabank.com
  7. verifiedBank of Canada Valet API — policy rate, fetched live. bankofcanada.ca
  8. illustrativeIstari model — Oakville & Burnaby pro-formas, lever effect sizes, friction weights, demand path. Argued in the Deep Dive. method

Starts/completions: StatCan 34-10-0154 (CMHC). Incomes: StatCan T1FF 11-10-0008/0009. Rates: Bank of Canada Valet. Oakville & Burnaby pro-formas and all model effect sizes are Istari assumptions, argued in the Deep Dive.