Can't Pay Your Mortgage in Canada: What Happens, and Where the Counting Stops

On 30 June 2026, 14,021 Canadian households were three or more months behind on their mortgage. Less than four years earlier, in August 2022, the number was 7,274 — the lowest figure in the thirty-one years of monthly record we rebuilt for this piece. If you are one of the 14,021, what follows next depends enormously on which province you borrowed in, and in Ontario a great deal of it can happen without a judge ever seeing your file.

This article does two things. It sets out what actually happens when the payments stop. And it shows you the four separate points at which Canada stops counting. The striking thing we found is not that someone is hiding the answer. It is that the systems built to measure this go blind in one specific configuration, and only that one, and no single one of them is doing anything wrong.

---

What actually happens when the payments stop

Three months is the number to know, but not for the reason most people assume. The Canadian Bankers Association counts a mortgage as being "in arrears" once it is three or more months past due, and that is the threshold at which a household appears in the only national statistic this country produces. It is a reporting threshold, not a legal one. When a lender may actually begin enforcing depends on the mortgage contract and on provincial law, and it can begin a great deal earlier than three months.

What happens after that depends enormously on where you borrowed, and that difference turns out to matter far more than most homeowners realise.

Canadian provinces run on two different models, and we read the statutes to confirm which is which.

Alberta is judicial. The Law of Property Act, RSA 2000, c. L-7, devotes Part 5 to the "Enforcement of Mortgages and Agreements for Sale of Land": section 39 governs foreclosure proceedings, section 41 the redemption time, section 48 the order of foreclosure. The phrase order nisi appears six times in the Act. The phrase "power of sale" appears zero times.

British Columbia is judicial. The Law and Equity Act, RSBC 1996, c. 253, runs from section 15 ("Order for sale") through section 22 ("Right of redemption controverted"), including section 18, "Power of court in foreclosure actions to make order without regular hearing," and section 21, "Venue in foreclosure proceedings." Procedure sits in Rule 21-7 of the Supreme Court Civil Rules. Some form of the word "foreclosure" appears thirty times. "Power of sale" appears zero times.

In both provinces a judge is involved, a file exists, someone signs an order, and that order is a public record. In Ontario, the lender's principal remedy is contractual and needs no court at all. The two models produce completely different public records: one leaves a countable trail, the other does not.

In Ontario, the dominant remedy is different. A lender holding a mortgage with a power-of-sale clause can serve notice and then sell the property, without commencing a court proceeding at all. There is no judge, no order, and consequently no court file recording who bought the house. The transfer is registered in the land registry, where inspecting the instrument costs money, one document at a time.

---

Less than four years separates the record low of 7,274 households from today’s 14,021.

How many Canadians are where you are: thirty-one years of record

As of 30 June 2026 there were 14,021 residential mortgages in Canada three or more months in arrears, out of 4,924,612 outstanding, a rate of 0.28%. The count has risen 92.8% since August 2022. It remains 25.0% below the all-time peak. Both halves of that sentence are true, and anyone who gives you only one of them is selling something.

The rate matters as much as the count. There are far more mortgages in Canada now than in 1997: 4.9 million against 2.47 million. A similar number of families in trouble therefore represents a much smaller share of borrowers. In January 1997, one mortgage in 155 was in arrears. In June 2026 it is one in 351.

What has changed is the direction and the speed. The 2022 low of 7,274 was not normal; it was the artificial floor of an era of emergency-low interest rates and pandemic payment deferrals. What we are watching now is that floor giving way. The honest description is not "a crisis worse than 2011." It is "the fastest deterioration in the record, from the lowest base in the record, and not yet finished."

---

Ontario carries 7,010 of the 14,021, and 73% of the entire national increase.

Ontario is not having the same crisis as Canada

National averages conceal the thing you would most want to know. When we broke the 31-year series out by province, one jurisdiction accounted for almost the entire national deterioration, and it is not the one with the highest arrears rate.

Three things in that table deserve to be read slowly.

Ontario has gone from 1,247 households in arrears to 7,010. That is a 5.6-fold increase in four years, and it puts the province within 6.6% of the worst month of the 2008–09 financial crisis. Of the 7,942 households added to arrears across all provinces since each one's own low point, 5,763 of them (73%) are in Ontario. Nationally the picture is a sharp deterioration from a record low. In Ontario it is a return to crisis conditions.

The 462% figure needs a guard rail, and we are putting one on it. Ontario's April 2022 low of 1,247 is extraordinarily small by any historical standard. A percentage change measured from a record-low denominator will always look enormous, and quoting it alone would be a cheap trick. The defensible sentence is the one with both anchors: Ontario's arrears have risen more than fivefold from a record low and are now back within touching distance of the financial-crisis peak. We would rather give you a smaller number you can trust than a larger one you cannot.

Saskatchewan and Manitoba are the quiet story. Saskatchewan has the highest arrears rate in the country at 0.46% and sits 43% above its own 2008–12 peak. Manitoba is 14.2% above its 2010 peak. Neither province is deteriorating quickly right now (Saskatchewan's count is actually at its lowest since 2021), but both have settled at a level their own history says is bad. They are invisible in national coverage because they are small. In a country that reports one number, small provinces with chronic problems disappear.

And Alberta runs the other way entirely: 6.9% above a low point only eight months old, and 63.3% below its 2011 peak. There is no single Canadian mortgage story. There are at least three.

---

The national arrears count covers nine banks. At least one in five outstanding mortgages sits with a lender it cannot include.

Blind spot one: the national count is a count of nine banks

Here is the sentence that reframes everything above it. The Canadian Bankers Association's arrears table names its own coverage on the page: BMO, CIBC, National Bank of Canada, RBC Royal Bank, Scotiabank, TD Canada Trust, plus Manulife Bank (from April 2004), Laurentian Bank (from October 2010) and Equitable Bank (from November 2020).

That is nine institutions. The series does not include credit unions. It does not include monoline mortgage lenders. It does not include mortgage investment corporations. It does not include private lenders. If your mortgage came from any of them, you do not appear in Canada's national count of mortgage arrears at all. Not as a statistic. Not as a data point. Not anywhere.

This is not a criticism of the CBA, and it is not a secret. It is a bank trade association publishing its own members' data and saying so clearly in a footnote. The organisation is doing precisely what it says it is doing. The problem is what happens when that number is then treated, by everyone downstream, as the number for Canada, because there is no other.

The obvious next question is how much lending actually sits outside those nine banks. That is answerable, and we answered it.

CMHC publishes a table of market share by lender type, drawn from its Survey of Non-Bank Mortgage Lenders and its own mortgage reporting. It is a data table rather than a judgement, so we used it and did the arithmetic ourselves. Four of the six lender categories in it — credit unions, other non-bank mortgage lenders, non-bank OSFI-regulated lenders, and mortgage investment entities — contain no member of the CBA's nine and therefore cannot appear in the national arrears count under any circumstances.

At least one in five outstanding Canadian mortgages, and at least two in five newly originated ones, sit with lenders that structurally cannot appear in Canada's national arrears count. Both figures are floors rather than estimates: any chartered bank outside the nine named adds to the uncovered share, and we have credited the entire "other chartered banks" category to the count even though only three of its members belong to it.

Now put the two columns beside each other. The count covers about 79% of the mortgages Canadians currently hold, but only about 59% of the mortgages Canadians are currently taking out. Today's originations become tomorrow's outstanding balances. The blind spot is not stable. It is growing, and it is growing without anyone deciding that it should.

---

Blind spot two: the epicentre is suppressed

If arrears tell you who is in trouble, property-transfer statistics should tell you what happened to the houses. Statistics Canada's housing programme publishes a variable for this: the number of buyers of properties sold in a non-market sale. Its own published definition of that term is worth quoting exactly:

"Non-market sale refers to non-arm's length transactions, which includes distressed sales, foreclosures, trade and forfeitures, redemptions, sales of part interest, and special interest sales."

That is as close as any free Canadian dataset comes to counting what happens when someone loses a home. We pulled the whole table from raw vectors and computed the share of those buyers that were businesses rather than people, against the same share in ordinary arm's-length sales.

Two findings come out of this, and the second is bigger than the first.

First: in three of the five provinces we could measure, the popular story runs backwards. In Nova Scotia, British Columbia and Manitoba, a business is less likely to be the buyer in a distressed transfer than in an ordinary sale. In Nova Scotia the gap is a factor of nearly three. Only New Brunswick shows corporate buyers meaningfully over-represented, and there the pattern is remarkably stable: between two and four times their market-sale share, in every single year from 2019 through 2024. New Brunswick is the one place in Canada where something resembling systematic corporate acquisition of distressed housing shows up in the data, and almost nobody is looking at it.

Second, and decisively: Ontario is not published. The buyer-by-owner-type figures exist only for resident owners, and Ontario's cells are suppressed. So the largest housing market in the country, the province carrying 73% of the national increase in arrears, the one jurisdiction where this question actually matters, cannot be answered from the national statistical record at all.

---

Where enforcement runs by contract rather than court order, no file is ever opened and no purchaser is ever recorded.

Blind spot three: no judge, no record

In a judicial-foreclosure province, the sale of a defaulting borrower's home passes through a court. That produces a file, and files can be counted. In Ontario, the power-of-sale route produces no such file, because no proceeding is commenced. The only trace is the transfer instrument in the land registry, and that registry is a pay-per-document system.

Follow what that means for a citizen, a journalist, or a member of Parliament who wants to know how many Ontario families lost homes last year and who bought them. There is no court list to count. There is no statistical table, because Ontario is suppressed. There is only a registry that charges by the instrument, meaning the price of answering the question scales linearly with the size of the question. Answering it for one house costs a few dollars. Answering it for the province is priced out of existence.

We should be precise about what this is and is not. It is not secrecy: the instruments are public, and anyone may buy one. It is not illegality: power of sale is a contractual remedy of long standing, and the registry's fees long predate the current situation. It is simply that the architecture, taken together, makes population-level scrutiny impossible without an institutional budget. The result is that a question a great many Canadians would like answered has no answer, and the absence is nobody's fault in particular.

Which brings us to the fourth gap, and the one that surprised us most.

---

11.00% of Canada’s federal property corporations have a public ownership record naming no human being.

Blind spot four: the company that bought it may name no human

Suppose you cleared the previous three obstacles and found that a numbered company bought the house. You would then want to know who is behind it. Since January 2024, Canada has had a public beneficial-ownership registry for federal corporations: they must file information about their individuals with significant control, and part of it is published free on each corporation's page.

Nobody had measured whether it works, because no bulk file of it exists — the registry is public one corporation at a time. So we asked it, one corporation at a time: a random sample of 6,000 of Canada's 644,202 active federal corporations, drawn with a fixed seed so anyone can reproduce the exact sample.

Projected across the register, that is between 51,223 and 60,386 active federal corporations whose public beneficial-ownership record names no person.

What we did not expect was that there are four entirely lawful routes to that outcome, and that only one of them is a failure to file:

1. Nothing filed. The register displays "No information has been filed."
2. Declared none. The corporation files, and states that no individual has significant control over it.
3. Unable to identify. The corporation files a declaration that it cannot identify who controls it — a statement of self-ignorance about its own ownership, accepted and published.
4. Excluded by statute. Listed public corporations, reporting issuers under provincial securities law, and their wholly-owned subsidiaries, are excluded from filing altogether.

Then we ran the same instrument as a 100% census — not a sample — over all 7,489 federal corporations whose registered name contains a property term: realty, real estate, properties, property, estates, apartments, rentals, housing, homes, residences.

Property companies are significantly less likely to have a public record naming a human being. But read the last row of the middle block before drawing the obvious conclusion: they are significantly more likely to file than corporations generally. They are not dodging the registry. They are using it precisely as written, and as written it produces no name.

More compliant, and less transparent. That is a different and more uncomfortable finding than a story about shells and evasion.

The statutory exclusions are the category that matters here. Among property corporations, 138 are excluded from beneficial-ownership filing because they are reporting issuers or the wholly-owned subsidiaries of one. Seventeen of them are "GP Inc." entities — general partners of limited partnerships, the standard vehicle for holding institutional real estate. A publicly traded real estate investment trust is a reporting issuer; the entities through which it holds property are its wholly-owned subsidiaries. Canada's beneficial-ownership registry does not apply to them, by design.

That is not evasion and we will not describe it as such. Those corporations are fully compliant; securities law imposes its own disclosure on reporting issuers, and the exclusion has a coherent rationale. It is simply, once again, a hole in the place a citizen would want to look.

---

What Parliament did: 458,340 words

Governments announce things. The only way to know whether an announcement became law is to read the law. So we enumerated every chapter of the Annual Statutes of Canada for 2025 and 2026 and searched the full text of each: 28 Acts, 458,340 words of enacted federal legislation.

The single occurrence is not a reform. It appears in the Budget 2025 Implementation Act, No. 1, and reads:

"Subsections (1) and (2) apply to taxation years that begin after 2024, except that if a corporation was controlled by or for the benefit of a real estate investment trust (as defined in subsection 122.1(1) of the Act) on April 16, 2024, subsections (1) and (2) apply to taxation years of the corporation that begin after 2025."

Parliament's only mention of real estate investment trusts across two years and 458,340 words gives corporations controlled by or for the benefit of a REIT an additional year before a measure applies to them.

And the government's flagship housing statute, the Build Canada Homes Act (5,191 words), contains none of the six terms. Not one.

---

What was asked, and what was answered

Parliament did examine this. In October 2023 the House of Commons Standing Committee on Human Resources tabled a report titled Financialization of Housing, containing eight recommendations. Recommendation 4 asked the government "to examine the social and economic costs and benefits of the current tax treatment of Real Estate Investment Trusts (REITs) and assess the option of taxing REITs like other corporations."

The Government Response was tabled on 26 February 2024. It runs to 3,122 words and answers each recommendation in turn. We read all of it and counted.

The word "REIT" appears exactly twice in the entire document — both times inside the committee's own recommendation. The government's answer never uses the word at all. Here is that answer, complete and verbatim. It is 107 words:

"While large corporate investors own a significant share of Canada's rental units and must play an important role in solving the housing crisis by building new homes, the Government recognizes that too many Canadians have experienced renovictions, above-guideline rent increases, and other actions that have made rent more expensive. More needs to be done to ensure these rental units are affordable for Canadians, which is why it is best to consider policy changes applicable to large corporate landlords to ensure best outcomes on affordability and fair treatment of tenants without undermining the important role the private sector must play in building more rental housing for Canadians."

It does not agree. It does not decline. It does not say the examination will happen, or that it will not. Of the eight recommendations, this is the only one whose response never returns to the subject it was asked about — recommendation 5 answers on tax measures, recommendation 6 declines explicitly on jurisdictional grounds, recommendation 8 says "supports this recommendation in principle and will examine it further." Recommendation 4 changes the subject to renovictions and closes on the importance of the private sector.

Then, on 27 November 2024, the House of Commons voted to adopt the report — recorded division No. 897, 184 to 145.

As of 28 August 2026, twenty-one months later, the Income Tax Act still defines a SIFT trust as one "other than an excluded subsidiary entity, or a real estate investment trust, for the taxation year." That parenthesis is what holds REITs outside the tax regime applied to other public flow-through entities. The most recent amendment recorded on the live section is 2017.

---

Why this is not a conspiracy, and why that is worse

Before making the argument, we should deal with the objection that would sink it. We went looking for measurement systems with gaps in them, and we found four. That is not the same as showing the gaps converge, because we chose the systems by looking for gaps. So the honest first question is: how many systems bear on this, and how many of them actually fail?

We can answer that, because two of the systems we checked work perfectly well.

Alberta and British Columbia both put mortgage enforcement in front of a judge. We read their statutes: a proceeding is commenced, an order is made, a file exists, and the purchaser is named in a public record. Those systems are not blind at all. Statistics Canada does publish buyer composition in non-market transfers — for New Brunswick, Nova Scotia, British Columbia, Manitoba and Prince Edward Island. That system works too. And the arrears count is complete and accurate for the nine banks it covers, which is most of the outstanding stock.

So the blindness is not general. It is conditional, and the condition is specific: it appears where a household borrowed from a non-bank lender, in a province that enforces by contract rather than by court order, and lost the home to a corporate purchaser. Remove any one of those three and some system sees you. Hold all three at once and none does.

That is a narrower claim than "four independent systems all fail," and we should also retire that phrasing, because it was overstated. The four gaps do not answer one question. One of them — the arrears count — answers who is in trouble. The other three all answer who ended up with the house, and two of those three are Ontario-specific. The accurate description is two questions and four institutional gaps, three of them stacked on the same question.

Narrower, and harder to dismiss.

Take each exclusion at its strongest

The Canadian Bankers Association is a trade association; it publishes its members' data because that is the only data it has, and it says so on the page. Statistics Canada suppresses small cells to protect the confidentiality of identifiable individuals, which is a legal obligation and a good one. Power of sale is a contractual remedy of long standing, and no court record exists because no court is involved — which is the entire point of the remedy, and on its own rationale it is faster and cheaper than litigating. The beneficial-ownership registry excludes reporting issuers because securities law already imposes continuous disclosure on them, and duplicating it would achieve nothing.

Every one of those justifications is sound. We accept all four. No individual, agency or company named in this article has done anything improper, and we found no evidence that any of them intended the aggregate result.

The mechanism is in that acceptance. These systems were built by different bodies, at different times, for different purposes, under different statutes. A trade association's disclosure practice, a statistical agency's confidentiality rule, a nineteenth-century contract remedy and an anti-money-laundering reform have no common designer and no shared subject. Ask what each exists to serve and the answer is never the same: the banks, the survey respondent, the efficient enforcement of a contract, the exposure of private companies to law enforcement.

Not one of them was built to see the household in default. The gap is not a place where a system looks away. It is the space between four systems, none of which was ever pointed at the person standing in it.

Which is why it survives. A deliberate cover-up has an author, and authors can be found and stopped. This has no author. It has no owner either — and that is the operative problem, because an unowned gap generates no one whose job it is to notice that it is there.

Picture the household this actually describes, because the argument has been abstract and the subject is not. They borrowed from a lender outside the big nine, probably because no bank would take them — which is to say they were already the most exposed borrower in the market. They fell behind. They did not appear in the national count, because their lender does not report into it. Their home was sold without a judge, so no file records the sale. A company bought it, and that company may be one of the 11% whose public ownership record names no human being. At the end of it there is a family who lost a house, and a country in which no institution holds a single line of data about any part of what happened to them.

That changes what the remedy has to be. If the invisibility were deliberate, the fix would be exposure. Because it is structural, exposure accomplishes nothing — there is nothing hidden to reveal. The fix is duller and more achievable: give some existing body the explicit job of counting this population. No one holds that mandate today. That is a description of the present arrangement rather than a law of nature, and it is why the last section of this article is a list of six things a person can actually file.

What would prove us wrong

A thesis that cannot be falsified is not a thesis, so here are the three findings that would defeat this one.

One. If any Canadian body already publishes a count of forced residential sales and their purchasers — a provincial regulator, a court administration, a land-titles office, a lenders' association — then the central gap does not exist and this article is wrong about the thing it is most confident about. We searched for such a count and did not find one. We would rather be shown it than be right.

Two. If the arrears coverage gap is materially smaller than we computed — if credit unions or non-bank lenders in fact report into some national series we did not locate — then the first exclusion shrinks and the argument weakens at its base. Note that this no longer turns on the unverified two-thirds private-lender figure: the coverage gap is now established directly from CMHC's own market-share table, and it holds whether that claim is true or false.

Three. If the conditional pattern is wrong — if households in the judicial provinces turn out to be equally uncountable in practice, whatever their statutes say — then the "conditional blindness" framing collapses into something vaguer and less useful, and the specificity we have claimed is false precision.

What the official record says, and why this piece does not rest on it

We were asked to build the argument in this article on evidence we gathered and computed ourselves, and not on any official body's characterisation of its own data. We have done that. But setting official statements aside quietly would be its own kind of dishonesty, so here they are, with our reasons.

The distinction we have worked to hold throughout is this: an agency's data is a record, and an agency's opinion about its own data is a claim. Records we used and recomputed. Claims we quoted and set down.

---

Six petitions you can file

This is the third arm of Petitions to Fix Canada, a series about the instruments an ordinary person can actually use. Every gap in this article is a measurement gap, and measurement gaps are unusually good petition subjects: they cost little, they do not require anyone to concede a political argument, and refusing them is hard to justify in public.

Three of these are federal. Three are provincial, because property, land registration and mortgage enforcement are provincial matters — a distinction that trips up most petition drafts and gets them ruled out of order. Take each to the right legislature.

If you have not filed one before, the mechanics (the sponsorship requirement, the signature thresholds, the 45-day response rule and its limits) are set out in the series hub. The two earlier arms cover federal programs you can claim now and the transparency instruments that actually bind.

---

Questions people actually ask

The questions below were harvested from Canadian search suggestions on 28 August 2026 rather than invented. They are answered in the words people used to ask them.

How many Canadians are behind on mortgage payments?

As of 30 June 2026, 14,021 residential mortgages were three or more months in arrears, out of 4,924,612 outstanding, a rate of 0.28%. Ontario accounts for 7,010 of them. Important caveat: that count covers nine banks only, and excludes credit unions, monoline lenders, mortgage investment corporations and private lenders entirely.

What happens if you can't pay your mortgage in Canada?

Not what most people expect. Three months past due is the threshold at which you enter the national arrears statistic, but it is not the point at which a lender may act — enforcement rights come from your mortgage contract and provincial law, and can arise considerably earlier. Where enforcement runs through the courts a judge supervises it; in Ontario, a lender holding a power-of-sale clause can serve notice and then sell the property without commencing a court proceeding. You generally retain a right to bring the mortgage back into good standing by paying the arrears and the lender's costs within the notice period — the specific timelines are set by provincial statute, so confirm yours with a lawyer in your province.

How many power of sale in Ontario?

Nobody knows, and that is not a figure of speech. Ontario power-of-sale transactions do not generate a court file, Statistics Canada does not publish Ontario's non-market-sale buyer composition, and the land registry charges per document. We searched for a free public count and could not find one. If you have seen a number quoted, ask what record it came from.

Is a mortgage arrears crisis happening in Canada right now?

Depends where you live, and the honest answer has two parts. Nationally, arrears have nearly doubled since August 2022 but remain 25% below the January 2011 peak and well under the 1997 rate. In Ontario, arrears have risen more than fivefold and are within 6.6% of the 2009 financial-crisis peak. Saskatchewan has the country's highest rate at 0.46% and is 43% above its own historic peak. Alberta has barely moved.

Who buys houses in a power of sale?

In Ontario, there is no public record that would let anyone answer this at a population level. Where it can be measured, the picture is mixed and mostly contradicts the popular story: in Nova Scotia, British Columbia and Manitoba, businesses are less likely to buy in a distressed transfer than in an ordinary sale. New Brunswick is the exception, with corporate buyers running two to four times their market share every year from 2019 to 2024.

Do corporations buy up foreclosed homes in Canada?

Not at the scale the phrase implies, on the evidence that exists — and in most measurable provinces the direction is the opposite. What is true is narrower and better documented: 11.00% of Canada's federal property corporations have a public ownership record naming no human being, and the corporate vehicles that hold institutional real estate are exempt from the beneficial-ownership registry altogether.

Can I stop a power of sale once it has started?

Usually yes, within the notice period, by paying the arrears plus the lender's enforcement costs — that is what the notice period is for. Options narrow sharply once the property is sold. Because the timelines and rights are provincial and unforgiving, this is a situation to take to a lawyer immediately rather than after the notice period runs. Do not rely on an article, including this one, for your own deadline.

What happens to the money left over after the house is sold?

Sale proceeds are applied in an order set by provincial statute: broadly the costs of the sale, then the mortgage debt, then any subsequent encumbrances, with any surplus belonging to the borrower. We could not verify that order against the statute itself from any free source, so treat the sequence as indicative and confirm it with a lawyer. What we can say without qualification is that whether borrowers actually receive their surplus is unknown. There is no national record of power-of-sale surpluses anywhere. We looked.

---

What this article cannot tell you

A piece that spends this long on other people's blind spots owes you its own.

• We cannot tell you who bought the houses in Ontario. That was the question we set out to answer. Every free route dead-ends, and we have documented each one rather than substituting a proxy and hoping you would not notice.
• Arrears are not foreclosures. Three months past due describes a household in trouble rather than one that has already lost its home. Many recover. Canada does not publish a national count of completed forced sales, so the step from 14,021 to "families who lost the house" cannot be made with public data.
• The arrears series is not a constant-coverage panel. Manulife Bank joined the reporting group in 2004, Laurentian in 2010 and Equitable in 2020, so level comparisons across those boundaries are affected. The source also footnotes reporting adjustments to Manitoba and Saskatchewan figures as at November 2006. Yukon is reported inside British Columbia; the Northwest Territories and Nunavut inside Alberta.
• The corporate-ownership work covers federal corporations only. Most Canadian companies are incorporated provincially and sit outside that frame entirely. Our property-corporation census identifies companies that name themselves after property; a company holding houses under an unrelated name is invisible to it.
• The statute scan proves the words are absent. It does not prove that nothing was done. A measure achieving the same end without using any of those six phrases would not be caught, and regulations made under existing authority would not appear in the Annual Statutes at all. Both are separate checks we have not yet run.
• "No information has been filed" is what the register displays, not proof of wrongdoing. Filing obligations key to annual-return timing, and a corporation may be well within its window.
• We verified Alberta's and British Columbia's mortgage-enforcement statutes directly and quoted them. We could not retrieve Ontario's, after exhausting every free automated route in a single sitting. The Ontario process description therefore rests on secondary legal sources, clearly marked where it appears.
• The lender-share figures are CMHC's data, and our arithmetic. We computed the uncovered share from its published market-share table; the categories, definitions and survey behind that table belong to CMHC, and we did not audit them.

If you can close any of these gaps — particularly if you have a primary source for the private-lender share of Ontario power-of-sale filings — we would like to hear from you at milad@zeusebikes.ca. Corrections will be published.

---

The bottom line

A household that no institution counts.

Fourteen thousand and twenty-one Canadian households were three or more months behind on a mortgage at the end of June. That number has nearly doubled in four years, three-quarters of the increase is in one province, and it counts only the people who borrowed from nine banks.

What happens to them afterwards is, in the largest housing market in the country, genuinely unknown — not classified, not concealed, just uncounted. Four institutions each doing their own job correctly, under four different statutes written for four different purposes, have produced a population that none of them is responsible for seeing.

The households are not invisible because anyone decided they should be. They are invisible because visibility was never anybody's assignment. That is a fixable problem, and it is fixable with the least controversial instrument in public policy: counting.

---

Sources

Every figure in this article was computed by the author from the records below. Retrieval date for all: 28 August 2026.

1. Canadian Bankers Association, Number of Residential Mortgages in Arrears, month ended 30 June 2026 (19-page PDF; monthly series from January 1995, Canada and by province). Extracted to a 3,402-observation monthly dataset by the author; continuity verified.
2. Statistics Canada, table 46-10-0030-01, Residential property owners by ownership type and residency status, with the number of buyers and sale price. Full-table download via the Web Data Service; buyer shares computed from raw vectors.
3. Statistics Canada, table 46-10-0023-01, Business and government ownership of residential properties, by legal type and industry. Full-table download, 619,542 rows.
4. Annik Gougeon and Oualid Moussouni, Residential real estate sales in 2018: Characteristics of properties sold, Statistics Canada catalogue 46-28-0001, 11 August 2021. Source of the verbatim definitions of "market sale" and "non-market sale."
5. Corporations Canada, Active business corporations bulk register (CBCA), 103,501,074 bytes, last modified 28 August 2026. Population frame of 644,202 corporations.
6. Corporations Canada, public corporation detail pages (individuals with significant control). 13,481 corporations queried individually by the author: a 6,000 random sample and a 7,489 census of property-named corporations.
7. Justice Laws Website, Annual Statutes of Canada 2025 and 2026. All 28 chapters, 458,340 words, enumerated and full-text searched by the author.
8. Justice Laws Website, Income Tax Act, R.S.C. 1985, c. 1 (5th Supp.), s. 122.1. Definition of "SIFT trust" and its amendment history.
9. House of Commons, Standing Committee on Human Resources, Skills and Social Development and the Status of Persons with Disabilities, Report 12, Financialization of Housing, 44th Parliament, 1st Session. Adopted 4 October 2023, presented 26 October 2023.
10. Government of Canada, Government Response to Report 12, tabled by Infrastructure Canada, presented 26 February 2024 (3,122 words).
11. House of Commons, recorded division No. 897, Sitting 377, 27 November 2024. Concurrence in the 12th report, 184–145.
12. Canada Mortgage and Housing Corporation, Renewal wave peaks but still dominates mortgage market, 12 May 2026, quoted in the disregard table above and not relied upon.
13. Canada Mortgage and Housing Corporation, Residential Mortgage Industry Report, Figure 8a "Market Share of Outstanding Mortgages (%)" and Figure 8b "Market Share of originated Mortgages (%)", Q3 2025 (source line: Survey of Non-Bank Mortgage Lenders, CMHC NHA MBS mortgage reporting, CMHC calculations). Uncovered-share arithmetic is the author's.
14. Law of Property Act, RSA 2000, c. L-7, Part 5 (Alberta) — read in full from the King's Printer.
15. Law and Equity Act, RSBC 1996, c. 253, and Supreme Court Civil Rules Rule 21-7 (British Columbia) — read in full from BC Laws.
16. Google autocomplete, Canada geography, 28 August 2026. 126 observed search suggestions used to select the questions answered in this article.

Visuals created by Playcut.ai