How auction clearance rates are counted, and why two sources disagree

In this article

Written by

Kooky

Builder of Shaka, the payment router that pays every agent their commission on closing date.

About Kooky and Shaka →

The auction clearance rate is the most quoted number in Australian property. It arrives every week, it fits in a headline, and it looks like a simple score: the share of auctions that ended in a sale. Yet anyone who follows Brisbane closely will have seen two or three different rates for what seems to be the same weekend, sometimes far apart.

Nobody has made a mistake when that happens. Each data provider has to decide which auctions go into the count, what counts as a sale, and what to do with the auctions nobody has reported yet. Those decisions differ, and each is defensible. This guide sets out how the main providers count, using their own published methods and one real Brisbane week from June 2026, and explains what has to match before two rates can be put side by side.

Cotality releases of 8 June and 11 June 2026; Domain's Brisbane auction results page for 31 May to 6 June 2026, last updated 10 June 2026.

One Brisbane week, three published rates

Start with the week that ended on the first weekend of June 2026.

On Monday 8 June, Cotality published its preliminary figures for the week ending Sunday 7 June. For Brisbane it listed 138 auctions, with results collected for 94 of them: 30 sold and 64 not sold. That gave a preliminary clearance rate of 31.9%.

On Thursday 11 June, Cotality published its final figures for the same week. Brisbane now showed 135 auctions and a final clearance rate of 34.1%.

In between, Domain's Brisbane auction results page, last updated on 10 June, showed a clearance rate of 18%. Its page covered Sunday 31 May to Saturday 6 June and listed 135 auctions scheduled and 113 reported, of which 20 were sold, 17 withdrawn and 76 passed in.

Three figures, then: 31.9%, 34.1% and 18%. The first two come from the same provider at two stages of collection. The third comes from a different provider, with a different week and a different map. The rest of this guide takes those differences one at a time.

A clearance rate is a fraction

Every clearance rate is one number divided by another. On top sits the number of properties counted as sold. Underneath sits the number of auctions the provider has decided to count. Almost every disagreement between two published rates comes from one of those two lines.

A scheduled auction can end in several ways, and the vocabulary matters:

• Sold at auction: the property sells under the hammer on the day.
• Sold prior: a buyer and the vendor agree before the auction, and the auction does not go ahead.
• Passed in: the auction is held, but the property is not sold on the day.
• Sold after: a property that was passed in sells by negotiation in the hours or days that follow.
• Withdrawn: the auction is cancelled before the day, and the property is not sold.
• No result reported: the auction was scheduled, and the provider does not yet know what happened.

A provider must take a position on each of these. Is a sale agreed on Tuesday a sale for Saturday's auction? Is a withdrawn auction a failed auction, or an auction that never took place? Is an unreported result unknown, or probably unsold? Each provider answers these questions in its own published method.

How Cotality counts

Cotality, the property data company, describes its method on its auction results page. Rates are calculated for auctions scheduled during the week ending Sunday. The clearance rate is the total known number of properties sold before, at or after auction, set against the total known number of auction results, including passed in and withdrawn auctions.

Two things follow from that wording.

First, the top line is broad. A property sold prior counts as sold, and so does one sold after the auction, as long as Cotality knows about the sale when it calculates the rate.

Second, the bottom line is the number of known results, not the number of auctions scheduled. An auction with no result collected is left out of the calculation until its result arrives. Cotality says plainly that not all results are available or known to it at the time of calculation. Withdrawn auctions are in the bottom line, so a withdrawal lowers the rate.

Cotality also states that its combined-capitals rate is presented as a weighted average, and that where fewer than 10 auction results have been collected, a clearance rate should be treated as statistically unreliable.

How SQM Research counts

SQM Research, an independent property research firm, publishes a different method on its auctions methodology page.

Its bottom line is every property originally scheduled for auction, including those withdrawn. Properties sold prior and properties sold on the day count as sold. A property sold after the auction is not counted as sold. Withdrawn properties, and those switched to a private treaty or tender listing, are treated as passed in. A result that has not been reported is also treated as passed in until it is confirmed. An auction that has been rescheduled is excluded from that day's count.

SQM Research also says it does not publish a preliminary figure. It waits several days and follows changes to the advertisements, and it says it normally publishes on Tuesday afternoons, covering the previous week through to the Sunday. Its stated reason is that results reported by agents in the first hours can leave out a large share of auctions.

The table below puts the two published methods side by side.

The two methods agree on most rows. They part on two: the sale that comes after a pass-in, and the auction nobody has reported.

The same auctions under two rules

A worked example shows how far those two rows can move a rate. The figures below are illustrative, not market data.

Take 100 scheduled auctions. Of these, 38 sell under the hammer, 6 sell prior, 5 are passed in and sell afterwards, 26 are passed in and stay unsold, 10 are withdrawn, and 15 have no result reported when the rate is calculated.

Under a known-results method of the kind Cotality describes, the bottom line is the 85 auctions with a known result. The top line is 49: the 38 hammer sales, the 6 sold prior and the 5 sold after. The rate is 49 divided by 85, or 57.6%.

Under an all-scheduled method of the kind SQM Research describes, the bottom line is all 100 auctions. The top line is 44: the 38 hammer sales and the 6 sold prior. The rate is 44 divided by 100, or 44.0%.

Nothing about the auctions changed between the two bars. The gap of 13.6 points comes entirely from the rules.

The example also shows why the first method moves over time. If the 15 missing results later arrive and all of them turn out to be unsold, the known-results rate becomes 49 divided by 100, or 49.0%. That is closer to the all-scheduled figure, and the 5 points that remain are the sales made after the auction. The two methods answer slightly different questions. One asks how many scheduled auctions ended in a sale by auction day. The other asks how many of the known campaigns produced a sale.

Preliminary and final are two different numbers

The second source of confusion sits inside a single provider's series. A weekly rate is published more than once.

The Brisbane figures show how incomplete the first count can be. In Cotality's preliminary release of 8 June 2026, results had been collected for 94 of 138 Brisbane auctions, which is 68.1%. A week later, in the preliminary release of 15 June, results had been collected for 112 of 144, or 77.8%. In both weeks, between a fifth and a third of Brisbane's auctions were not yet in the calculation on Monday morning.

Cotality says it collects, on average, 99% of results by the time of the final. The final is therefore a different measurement from the preliminary, taken on a fuller sample. This is also where SQM Research's choice makes sense on its own terms: by treating an unreported result as passed in and publishing once, on Tuesday, it avoids a first figure that later moves. Cotality's choice gives readers an early reading on Monday and a settled one on Thursday. Both are coherent, and they produce different numbers.

Revisions go both ways

SQM Research writes on its methodology page that preliminary rates from other providers are often revised lower some days later. Often is not always. Two consecutive Brisbane weeks in June 2026, both from Cotality, show that the direction is not fixed.

In the first week the final came in 2.2 points above the preliminary. In the second it came in 2.6 points below. The number of auctions also changed between the two stages, from 138 to 135 and from 144 to 142, as the list of scheduled auctions was itself corrected.

The lesson is narrow but useful. A preliminary rate is an early estimate with its own margin, and the late results can push it either way. Comparing one week's preliminary with another week's final mixes two kinds of measurement.

Domain's page: another week and another map

Domain, the property portal, publishes auction results by city. Its Brisbane page for the week discussed above showed a clearance rate of 18%, against Cotality's final of 34.1%. Before reading that gap as a disagreement about the market, three differences are worth setting out.

The week is different. Domain's page ran from Sunday 31 May to Saturday 6 June 2026. Cotality's week ended on Sunday 7 June. An auction held on Sunday 31 May sits in Domain's week and not in Cotality's, and one held on Sunday 7 June does the reverse.

The map is different. Domain's Brisbane page also lists suburbs on the Gold Coast and the Sunshine Coast, while Cotality publishes Brisbane and those two regions in separate lines. That both sources arrive at 135 auctions does not mean they counted the same 135.

The published breakdown is different. Domain's page reports 135 scheduled and 113 reported, with 20 sold, 17 withdrawn and 76 passed in. Those three outcomes add up to the 113 reported, and 20 out of 113 is 17.7%, which rounds to the 18% shown. The page itself does not set out a formula, so that is an observation about the published numbers and not a statement of Domain's method.

Cotality's final figures for the same week add one more piece of context. Brisbane's pass-in rate was 56.3%, the highest of the capitals. Both sources, in other words, describe a week in which most Brisbane auctions did not sell on the day. They differ on how many sales to count, over which days and across which suburbs.

Small samples move in big steps

Queensland adds a further complication: outside the Brisbane total, the weekly counts are small.

Cotality's sub-region tables for the week ending 7 June 2026, in final form, listed 50 auctions on the Gold Coast with 16 sold, a rate of 32.0%, and 23 auctions on the Sunshine Coast with 10 sold, a rate of 43.5%.

With 23 auctions, each result carries a lot of weight. Had one more Sunshine Coast property sold that week, the rate would have been 11 out of 23, or 47.8%: a move of 4.3 points from a single sale. Across the combined capitals, where Cotality's final count for the same week was 1,175 auctions, one more sale would have moved the rate by less than a tenth of a point.

This is why a regional rate that jumps 10 points from one week to the next is often a story about arithmetic. Even in Brisbane, where Cotality counted between 135 and 144 auctions in the two weeks above, a handful of results still shows in the rate.

The combined capitals figure

The national number that leads most weekly reports is built differently again. Cotality presents its combined-capitals clearance rate as a weighted average, so the cities that hold the most auctions carry the most weight. Brisbane, with 135 of the 1,175 auctions in the final count for the week ending 7 June 2026, is a small part of it.

For that week, Cotality's final combined-capitals clearance rate was 47.3%. The same release counted 455 auctions passed in, or 38.7%, and 164 withdrawn, or 14.0%. The three shares add up to 100%, which is the fraction laid out in full: every auction in the final count was sold, passed in or withdrawn.

Cotality's release of 11 June 2026 also gave a point of reference: the decade average for the combined capitals is 64%. A reference like that only holds within one provider's series. Setting a rate from one provider against a long-run average from another compares two different fractions.

What two rates need in common

Put together, the published methods suggest five things that have to match before two clearance rates describe the same thing.

1. The provider. Cotality, SQM Research and Domain each publish their own count. A rate belongs to the series it comes from.
2. The week. A week ending Sunday and a week ending Saturday contain different auctions.
3. The stage. A preliminary rate and a final rate are measured on different samples, even within one series.
4. The map. "Brisbane" may or may not include the Gold Coast and the Sunshine Coast, depending on the page.
5. The sample. A rate on 23 auctions and a rate on 1,175 do not carry the same weight.

When all five match, a change in the rate says something about the auctions. When one of them differs, part of the change, and sometimes all of it, comes from the counting.

None of this makes the clearance rate less useful. It remains one of the few weekly readings of how buyers and vendors are meeting each other. It simply comes with a label, and the label is part of the number.

A clearance rate is a count before it is a verdict. Two rates disagree most often because they did not count the same auctions, on the same days, in the same way.

Kooky, from Shaka

Kooky edits Queensland Estate and builds Shaka, the payment router he made for Queensland property professionals. One payment comes in, and every agent, agency and party in the deal receives their signed share on closing date.

About Kooky and Shaka →