One number, and what it actually tells you

What it measures

It’s a ratio, and ratios are worth a second look. Each week we count how many $2 million-plus homes went under contract against how many are sitting for sale, then compare that pace to the market’s own average over the previous five years. 100 means a normal pace. Above 100, the market is clearing faster than its recent normal; below, slower.

This week it reads 116.2 — a touch above normal. See the full reading →

A detail worth knowing

Because it’s a ratio, the Index is lopsided by nature, and the lopsidedness is informative. It can’t fall below zero—you can’t have fewer than no sales—so a cold market bottoms out around 30. But there’s no ceiling on how fast homes can clear relative to normal, so a genuine frenzy runs to 400 or 500. That asymmetry is why the number spikes dramatically at the top and merely sags at the bottom: enthusiasm has no limit, but a slow market can only get so slow. When you see the Index at 110, you’re a tenth above normal. When you saw it at 500 in 2021, the market was clearing five times its own recent pace—a reading that, by construction, only extraordinary demand can produce.

How to read it

Three zones, and the middle one is the point:

• Below 70 — cold. Rare. Buyers have stepped back and most homes wait. Last seen in the 2008–09 crash.
• 70 to 250 — normal. Where the market sits almost all the time. It’s a deliberately wide band, because within it the market is not the variable that decides your sale. Your price is.
• Over 250 — frenzy. Rare. Homes clear fast, at or over asking. Last seen in 2021–22.

What 25 years of the data actually show

Here is the finding that surprised us, drawn from the full record behind the Index: the market sets your odds, but it almost never sets them to certainty. Even in 2021—the fastest market Arizona has ever recorded—roughly one luxury listing in three still failed to sell. In the depth of the 2009 crash, four in five failed. So the Index is not telling you whether homes are selling; some always are, and many always aren’t. It is telling you your margin for error.

That reframes what a market reading is for. A firm market forgives a pricing mistake—there are enough buyers that a slightly-high number still finds one. A soft market punishes the same mistake without mercy. But the mistake itself is never the market’s; it’s a decision made when the home is priced and listed. The Index tells you how much room you have. It has never, in 25 years of data, told anyone they could skip pricing carefully.

Here is the part worth sitting with. If you sort every luxury listing of the last 18 years by how hot the market was the week it launched, and ask how many failed to sell, the answer barely moves across the entire normal range:

Look at the middle three. A strong market at 200 — the kind everyone calls a seller’s market — failed 55% of the time. A merely normal market at 100 failed 56%. Across the whole wide band, from soft to strong, the failure rate holds near 55–60%. The market’s temperature barely touches it. The needle only breaks in the two rare extremes: a true crash, where almost nothing sells, and a genuine frenzy, where the tide lifts even mispriced homes. Everywhere in between — which is to say, almost always — the outcome was decided at the list price, not by the market. The full study, with method and data →

What it can’t do

It measures the market, not your home. It won’t tell you what a specific house is worth or what one particular buyer will pay—it’s the climate, not the forecast for your street. And it looks backward, not forward: it reports what happened last week, not what buyers are about to do. That’s deliberate. A measurement you can verify is worth more than a prediction you can’t, and the number is built to be the former.

Where it comes from

Every Friday the week’s data is pulled and capped at the previous Thursday, so the reading is the same whatever hour it’s run. Two counts—homes for sale, homes newly under contract—are adjusted for the season, compared to the market’s own five-year norm, and smoothed over four weeks. Once a week publishes, its number is frozen and never revised. The full construction, the 18-year back-history, and the complete data set—free to download and check—are all published for anyone who wants to reproduce a single number of it: how the Index is built, in full →