Sold prices, sold to list against both the last ask and the first, days on market and months of inventory, for one kind of home of one size in one place. Every figure prints the number of sales it was computed from, and anything resting on fewer than 12 sales is shown but refused as a publishable statistic rather than smoothed into something that looks like a market. Where nothing sold, this page says so. Then take a figure from here to the cost of ownership calculator and find out what carrying it would cost.
Months of inventory is the active listing count divided by the monthly sales pace, and 4 to 6 months is the band usually called balanced. Not included anywhere on this page: what a specific house is worth, what condition any of these homes were in, and what was agreed off the record. A sold price is the outcome of one negotiation between two parties, and a median of them is not a valuation.
46 sales in the 12 months to Aug 14, 2026, and 17 listed today.
The middle sale of 46 in the 12 months to 2026-08-14. Median rather than average, because one estate sale or one builder buying for the land moves an average of this many sales by more than a year of price drift and moves the median by nothing.
Against the ask the sale actually closed at, from 46 of 46 sales. A listing that cut its price twice and then sold at the third ask reads as 100% here, which is why the figure below exists.
Median of the 11 sales in 2026-06 to 2026-08 against the 13 in 2025-09 to 2025-11: up 11.7%. Two medians of this size are not a trend line, and a change under 2% is reported as flat because it is inside what one unusual sale does.
Seasonality. GTA sales are strongly seasonal: the same segment's median moves several percent between a February close and a May one with no change in the market. Separating that from a real move needs at least 24 months of the same segment and this one has 18, so no seasonal factor is applied and none is borrowed from a national series. Part of the direction of travel may be the calendar.
Every figure is shown with the number of sales behind it. A figure computed from fewer than 12 sales is marked as too thin to publish and is not the market.
The middle sale of 46 in the 12 months to 2026-08-14. Median rather than average, because one estate sale or one builder buying for the land moves an average of this many sales by more than a year of price drift and moves the median by nothing.
Shown next to the median rather than instead of it. It sits 21.3% above the median, and the gap is the skew: a few large sales pull an average and leave a median where it was.
Median of the per-sale figures from the 37 of 46 sales that state a floor area, not the median price divided by the median area. Listed areas are the seller's own measurement and are frequently a range, so this figure carries more error than the price above even where the sample is large.
Against the ask the sale actually closed at, from 46 of 46 sales. A listing that cut its price twice and then sold at the third ask reads as 100% here, which is why the figure below exists.
Against the price the listing first asked, from 46 of 46 sales. The gap between this and the figure above is how much of the segment's apparent discipline came from cutting the ask rather than from what buyers paid. No portal shows both.
Days from the current listing to the firm sale, for 46 of 46 sales. It counts the listing that sold, not the two that were terminated and relisted before it, so a house that has been available for six months can appear here as eleven days. That is a property of every days on market figure published anywhere, including this one.
17 listings on the market against 3.8 sales a month over the 12 months to 2026-08-14. Under 4 months is conventionally read as a seller's market and over 6 as a buyer's market, which is a convention of the trade rather than a law. It assumes the next few months sell at the pace of the last 12, and in a segment this size they may not.
The share of the 17 active listings that sells in a month at the pace of the 12 months to 2026-08-14. It is the reciprocal of the months of inventory and is here because it is the figure a listing agent quotes.
43 listings came to market in the 12 months to 2026-08-14, against 46 sales closing. Reconstructed from the listings this dataset holds rather than counted from a feed of new listings, so it is a floor and not a count. Listings that came and went before we saw the corpus are missing from it. Because new supply is the denominator of the ratio below, an undercount reads as a tighter market than the real one, which is the one direction this error runs. Derived from the listed date on each record, or the sold date minus days on market where the listed date is absent, both of which only see listings this dataset still holds.
46 sales against 43 new listings in the 12 months to 2026-08-14: 107% of new supply sold. Under 40% is conventionally read as a buyer's market and over 60% as a seller's, which is a convention of the trade rather than a law. It is not a share of these listings that sold: some of these sales were listed before the window opened. Reconstructed from the listings this dataset holds rather than counted from a feed of new listings, so it is a floor and not a count. Listings that came and went before we saw the corpus are missing from it. Because new supply is the denominator of the ratio below, an undercount reads as a tighter market than the real one, which is the one direction this error runs.
Computed from seeded demonstration sales, not from a PropTx feed. The segmentation, the sample thresholds and the arithmetic are real; the sold prices are not, so read this page as a description of the method rather than of the market. Nothing on this page is an appraisal or a valuation of any specific property, and a median is not a price anybody paid.
