The New Market

For the first two decades of this record, a sale above $10 million was an event: Arizona closed between zero and three of them a year. 2019 closed six. 2026 has closed 51 by late August — already the largest year in Arizona history, running at 2.4 times last year’s pace — and it has done so in a year our own weekly index has read the broad luxury market below its five-year norm throughout. Something at the very top is no longer following the market beneath it. This study measures what.

First, the two honest corrections

Part of this curve is arithmetic, and we correct for it before claiming anything. Prices across the tier rose roughly 87% since 2019 by our repeat-sales index, so a fixed $10M line catches homes that were $5.3M-grade homes in 2019 — “$10M is the new $5M” is close to literally true. Held to a constant-quality bar, the growth from 2019 through 2025 largely disappears at every tier: the luxury pyramid inflated uniformly, its shape unchanged, only its denomination different. Second, thin tails are mechanically sensitive: with prices distributed the way housing prices are, small uniform inflation multiplies counts above high fixed thresholds faster than above low ones. Both corrections are real, and through 2025 they explain the top of the market almost entirely.

Then 2026 broke the pattern

Neither correction survives this year. January through August, apples to apples: 24 closings above $10M in 2024, 21 in 2025, 51 in 2026 — 2.4 times the prior year’s pace — while tier prices moved about 1% and generous tail-sensitivity multiplies that into perhaps a 10–20% count effect, not 140%. At 51 events against 21, the change clears statistical noise. Even at a constant 2019-$10M-grade bar (an $18.7M home today), 2026 has already matched 2019’s full-year count with four months remaining — the first constant-quality movement at the top in this record. And it is happening into a mild market: the weekly index below 100 all year, price cuts outnumbering raises across the broad tier. The very top has decoupled from the tape beneath it.

Different product

The tier is also physically distinct from the market below it, and increasingly so.

The typical $2–5M sale in 2023–26 is a 21-year-old home. The typical $10M+ sale is five years old, and 43% of the tier’s closings — 49% in 2026 — were built within two years of their sale date. In 2019 that new-build share was zero. Much of this market’s growth is not existing trophies trading faster; it is inventory that did not previously exist being delivered and absorbed — which is also how a tier can grow while constant-quality resale volume across luxury sits flat. Even its resales skew to last cycle’s deliveries. The ultra tier trades in new and near-new architecture; the rest of luxury trades the existing stock.

Concentrated geography

Since 2023, 64% of all $10M+ closings are in Paradise Valley, and Silverleaf alone accounts for a further 18% — two places, 82% of the market. The state’s price ceiling has moved with it: the record has been reset repeatedly in three years, culminating in a $40M Paradise Valley closing in 2026, with each of the last three record-setters standing within the same few square miles.

What makes it a category

A market segment earns separate treatment when it can be measured separately and behaves separately. This one now clears both bars on three independent measurements: volume — 51 closings in a partial year supports tracking and reporting where 0–3 a year supported anecdotes; dynamics — it accelerated 2.4× into a below-norm market it evidently does not take direction from; and product — new-and-near-new construction concentrated in two locales, physically unlike the tier below. By the standards that make “luxury” a segment of the whole market, ultra-luxury is now a segment of luxury — and we will report it as one.

Caveats

Prices here are final asking prices, not closed prices; the gap between the two is widest at the top of the market, in both directions, and does not affect year-over-year comparisons made on the same basis. The 2026 count includes a meaningful share of speculative new construction closing on delivery; we reviewed the year’s composition and report the new-build share rather than netting it out, because delivered-and-absorbed inventory is market activity — but readers should know the growth is roughly half new product. The repeat-sales index used for constant-quality bars is built on the broad upper market; if the very top appreciated faster than the tier, our constant-quality counts are overstated and the pre-2026 “no real growth” conclusion only strengthens. 2026 is a partial year; four months remain to confirm the pace. Counts at the $10M threshold are small in early years; era comparisons, not single-year readings, carry the conclusions.

The Arizona ultra-luxury guide →  ·  This week’s market reading →  ·  You Make Your Own Market →