Twenty, Fourteen, Thirteen: What the Ranking of Three Growth Rates Says

Three segments, three inventory growth rates, published on the same day for the same region. The temptation is to comment on each in turn. There is more to be had from their order, provided you know exactly what an order contains and what it loses along the way. For how condo supply is spread geographically, see our piece on condos by sector across the CMA.

The three rates, and nothing else

The QPAREB release of August 6, 2026 gives three year-over-year inventory changes for the Montreal census metropolitan area in July 2026: 20% for condominiums, 14% for plexes, 13% for single-family homes. The region held 19,790 active listings in total, up 17%, which is 9% above the historical average for a month of July.

This piece deals only with those three numbers and with putting them in order. It deliberately sets aside what each segment is doing on other measures, because mixing a ranking with other indicators is the fastest way to make it say anything at all.

What an active listing counts

The active listing count is a state, taken on a date: every property on offer at that instant. It is not a running total over the period, nor a tally of arrivals or departures. Comparing July 2026 with July 2025 therefore compares two photographs, taken a year apart.

What the change measures, then, is how much the queue of sellers standing there at the same time has lengthened in that segment. It is a simple, well-defined quantity, so long as you do not ask more of it: it says nothing about who joined the queue, who left it, or at what pace.

Why the order is sturdier than the values

A single value is vulnerable to anything affecting its measurement. A change in counting scope, a calendar effect, an unusual comparison base last year: each of those shifts the published rate without the market having moved by as much.

Those disturbances usually hit all three segments together, since they are counted by the same body, on the same method, over the same territory, on the same date. A shared bias moves the three values the same way and leaves their order untouched. That is the property making a ranking sturdier than its components: it survives what makes the values arguable.

First blind spot: a ranking flattens distances

Here is the costliest limitation, and it is visible at a glance. Between 20% and 14% there are six points. Between 14% and 13% there is one. A three-place ranking presents these segments as three distinct, evenly spaced positions, which they are not.

The real structure of the data is different: one segment clearly apart, and two others effectively tied. A one-point gap between plexes and single-family homes supports no claim of differing behaviour between them; it sits at the same order of magnitude as measurement noise. The accurate phrasing is not "condos first, plexes second, single-family third" but "condos on one side, the other two together on the other".

Second blind spot: growth and level are two different rankings

A growth rate measures a move against a starting point specific to each segment. The segment whose inventory lengthens fastest is therefore not necessarily the one holding the most listings, nor the one whose supply is most abundant relative to its demand.

Those two rankings, of changes and of levels, are logically independent: any combination of the two is possible. Conflating them produces false claims about the relative abundance of supply, delivered with the confidence a published figure lends. When a ranking of growth rates circulates, the question to ask is simply: and what does the ranking of levels look like?

Third blind spot, and the use that survives

An order taken on one date says nothing about how long it has held. It may have stood for twelve months or formed last month, and nothing in the three published numbers settles which. A ranking stable across successive readings and a brand-new one call for very different interpretations, while looking exactly alike on the page.

With those three caveats stated, one honest and useful application remains: knowing which queue you are standing in. A condo seller is in the segment whose queue has lengthened most, and is better off knowing that before preparing a listing. This ranking sets no price, describes no neighbourhood and replaces no comparable. It directs the questions worth asking, which is already more than a single rate commented on alone will give you.