A third of Ottawa's 2026 sales shortfall landed in August

OREB's August release said prices held steady and conditions stayed broadly balanced. Both of those are true. Neither is the most interesting thing in the file.

The most interesting thing is that OREB published the year-to-date shortfall as a percentage in two consecutive months and never subtracted one from the other.

End of July, Ottawa's 2026 sales were running 5.2% behind 2025. End of August, 6.9% behind. One month moved the year-to-date gap by 1.7 points. I wanted to know what that was worth in actual houses.

The arithmetic

OREB gives you everything you need to do this yourself.

Year to date through August, 9,283 homes sold in Ottawa, down 6.9% from the same period in 2025. Divide 9,283 by 0.931 and 2025's eight-month total works out to 9,971. So Ottawa is 688 sales behind where it was a year ago.

Now August on its own. 1,002 homes sold, down 18.6% year over year. Divide 1,002 by 0.814 and August 2025 was 1,231 sales. August 2026 came in 229 sales short.

229 divided by 688 is 33.3%.

August is one month out of eight. It produced a third of the entire year-to-date shortfall.

Cross-check it the other way if you don't trust me. Year to date through July was 9,283 minus 1,002, so 8,281, and that was down 5.2%, which puts 2025's seven-month figure at 8,735. That is a 454-sale gap over seven months. Add August's 229 and you get 683, which lands on 688 once you allow for OREB rounding its percentages to one decimal.

Seven months built 454 sales of shortfall. The eighth month added 229 more. One month did half of what the previous seven managed between them.

Why the price numbers didn't flinch

Here is the part that trips people up.

The August average sale price was $688,253, up 0.3% year over year. The MLS benchmark was $637,700, up 1.0%. Homes sold at 97.9% of asking, unchanged from last August. Median days on market went from 28 to 29.

Every one of those four numbers describes a house that sold. They are computed on the 1,002 transactions that closed. The 229 sales that did not happen contribute nothing to any of them, because a sale-to-list ratio requires a sale.

That is not a knock on OREB. It is how the statistics are defined. But if you are a seller reading "97.9% of asking" and taking comfort from it, understand you are reading the scoreboard for the people who got a deal done, not the full field.

OREB says as much in the same release, quietly. Its own review of listing records found that terminations, cancellations and expirations became more prominent relative to completed transactions through the summer. I went through that in detail when the August numbers first landed. More homes are leaving this market without selling than the headline implies.

The eight-month price story is not the one-month price story

One more derived number, and this one surprised me.

Year-to-date dollar volume was roughly $6.5 billion, down 7.2%. Year-to-date sales were down 6.9%. Dollar volume fell faster than unit count.

Divide 0.928 by 0.931 and you get 0.9968. So the average price per sale across all of 2026 so far is running about 0.3% below last year's pace, even though August's monthly average came in 0.3% above.

Small number. I am not going to dress it up as a correction. It is worth naming because "prices are up slightly" and "prices are down slightly" are both defensible readings of the same release, depending on whether you look at one month or eight. If someone quotes you a price trend this fall, ask which window they used. The average versus benchmark gap I wrote about earlier this year is the same trap wearing a different hat.

Put 4.5 months of inventory on a scale

Months of inventory went from 3.5 in July to 4.5 in August. OREB notes that over the previous ten years the median July-to-August change was zero, and no increase ever exceeded 0.4 months. So a full one-month jump is about two and a half times the largest normal seasonal move in a decade, and it is the highest August reading since 2016.

That is the alarming framing. Here is the other half of it.

Four and a half months is still inside the band that gets called balanced. Single-family sat at 4.0 months. Townhouses at 4.1. Active listings actually fell 3.9% from July, and new listings were flat year over year. This inventory build is not a flood of sellers arriving. It is buyers stepping back, which is a different problem with a different fix.

The softness also has a postal code. Ottawa Centre came in at 7.0 months and Rural East at 6.5, while Suburb West ran 3.6 and Rural West posted the only year-over-year sales increase in the city. I pulled that apart last Sunday. A citywide number is a weighted average of six markets that are not doing the same thing.

What I would actually do with this

Selling this fall, the number that matters is not 97.9%. It is that a third of this year's demand gap opened up inside the last four weeks of data anybody has. Price to today, not to July.

Buying, you have more leverage than the balanced-market label suggests, and you have the most of it in the central core and the east rural belt.

Waiting, the September OREB release lands in early October and the Bank of Canada decides on October 28. One soft month is a data point. Two is a direction. I would want the second release before calling it either way.

On the record so you can score me later: if September's year-to-date shortfall widens past 7.5%, August was not an interruption.

Want this arithmetic run on your street or your building instead of the whole city? Fifteen minutes on your actual situation beats another market summary in your inbox. Reach me at 613-262-6545 or fil@613realtor.ca.