Market Insights

A Montreal Housing Crash: The Data

By Georges Matar · June 17, 2026 · 6 min read

The question comes back the moment a headline mentions a slowdown: is the market going to crash? I am not going to answer that, because nobody can. What I can do, and what is more useful, is show which numbers exist, what they actually measure, and where they stop letting you conclude anything.

Everything below comes from public sources, dated and named. You will find no price prediction here, no advice about timing, and no scenario.

A Word With No Statistical Meaning

Crash has no official definition. APCIQ publishes no crash threshold. Neither does Canada Mortgage and Housing Corporation, and neither does Statistics Canada. The word circulates in the media, not in the tables.

What does exist in the tables is four distinct measures: number of sales, number of active listings, median price by property category, and average days on market. None of them is the market on its own, and they do not move together.

A Sales Slowdown and a Price Decline Are Two Different Things

This is the most common confusion, and the August 2026 data takes it apart in one stroke.

In the Montreal census metropolitan area, APCIQ recorded 2,853 sales in August 2026, which is 13% fewer than in August 2025. In that same month, active listings reached 20,128, up 18% year over year.

And median prices, in that same month, were all higher than a year earlier: $650,000 for single family homes, up 3%, $437,250 for condominiums, up 4%, and $856,000 for plexes, up 2%. Those figures come from the monthly statistics APCIQ publishes through the Centris system, released on 4 September 2026.

Fewer transactions, considerably more choice, and median prices that are not falling. Sales and prices are two different series. A decline in one does not mechanically show up in the other, not in the same month and not in the same direction.

The Indicators You Can Check Yourself

You do not need me to follow a market. You need the right numbers at the right geographic level.

Average days on market. In August 2026, in the Montreal CMA, it stood at 42 days for single family homes, 62 days for condominiums and 52 days for plexes. The year over year change says more than the level: up 1 day for single family homes, up 12 days for condominiums, down 4 days for plexes. Three categories, three directions. I break that measure down in how long it actually takes to sell in Montreal.

Active listings. This is available supply. The 20,128 listings of August 2026 broke down into increases of 17% for single family homes, 19% for condominiums and 15% for plexes. Read it alongside new listings for the month, 5,874, up 7%.

Median price by category and by sector. A median is the midpoint of closed transactions, not the value of your property. It also moves when the mix of what sold changes.

The ratio of asking price to selling price. This is the indicator buyers ask for most, and it is not in the APCIQ monthly release. It is calculated sector by sector from closed Centris transactions. Ask your broker for it for your sector and your category, and ask how many transactions it is based on. A ratio with no sample size is worth nothing, including when it suits my argument. Same logic as with bidding wars.

Why a National Headline Describes Your Market Badly

Take that same month of August 2026 and compare two geographic levels, both published by APCIQ.

For the province of Quebec as a whole: 6,502 sales, down 7%, and a median price of $497,501 for single family homes. For the Montreal CMA alone: 2,853 sales, down 13%, and a median price of $650,000 for single family homes.

Same month, same source, same definitions. Six points of difference on the sales change, and more than $150,000 on the median price. A headline saying the market slowed by 7% is not wrong, it is simply talking about something other than your neighbourhood.

Splitting by category widens the gap further inside the CMA: single family sales fell 9% in August 2026, condominium sales fell 18%. That is the same market inside the same newspaper sentence.

Two Common Claims the Data Contradicts

“Population is exploding, so demand is guaranteed.” The Institut de la statistique du Québec published its 2025 review on 13 May 2026. Quebec’s population stood at 9.03 million on 1 January 2026, down 9,600 people over the year, or 0.1%. Natural increase was negative, with 78,200 births against 80,450 deaths. Total net migration was minus 7,150, made up of net international migration of 450 people and net interprovincial migration of minus 7,600. The number of non permanent residents in Quebec was estimated at 514,050 on 1 January 2026, down 51,400 from a year earlier.

That data covers all of Quebec, not the Montreal CMA, and says nothing about prices. But it removes the foundation from the demographic argument as it is usually stated.

“Employment is collapsing.” According to the Statistics Canada Labour Force Survey released on 4 September 2026, the unemployment rate in the Montreal CMA was 6.5% in August 2026, almost unchanged both from the previous month and from a year earlier. The Quebec rate was 5.6% and the Canadian rate 6.4%. That same release notes, however, that the month’s employment losses were concentrated in the Montreal CMA, down 21,000 or 0.9%. Both facts sit in the same document and deserve to be read together.

What the Data Does Not Let You Say

None of the series quoted here projects prices for the next twelve months. None tells you whether a given month marks a turn or a pause. A year over year change depends on the reference month as much as on the observed one. A category median does not describe one specific property. And a single month is a small sample as soon as you drop down to one neighbourhood.

The same caution applies in the other direction. A headline announcing a record is describing one month of one series in one territory, not a market, and a strong month is no more predictive than a weak one.

Anyone announcing a specific percentage decline or increase for next year is giving you an opinion, not data. Brokers included.

What a Cycle Actually Changes Over a 10 Year Horizon

If you are buying to live somewhere for ten years, the cycle moves three concrete things, and none of them is the future price.

Time and negotiating room. More listings and longer days on market in a category means more showings you can attend, less pressure to sign the same evening, and an inspection condition that is easier to keep in the offer.

Entry costs. Welcome tax, notary, inspection, moving: they are paid once and spread across the holding period. Over ten years they weigh far less than over two. That point, not a forecast, is what justifies a long horizon. The rent versus buy calculator puts those costs in the balance.

Renewals. Over ten years you will renew several times. The gap between your approved maximum and your real budget matters more than the price you paid at the start.

And there remain situations where buying is the wrong decision regardless of the cycle. I describe five of them here. For a wider reading of the same data, see also what the numbers say about the Montreal market.

Want the numbers for your sector and your property category rather than for Canada? Write to me and I will pull the recent data for your neighbourhood so we can read it together.