You Make Your Own Market

Agents repeat a piece of folk wisdom: you make your own market. Sellers repeat its opposite: wait for the market to improve. Both are testable. We linked every Arizona $2M+ launch from 2005 through 2024 — 21,978 selling campaigns with known resolutions — to the reading of our weekly market index in the week each campaign launched, then followed every one to its outcome. If the market decides, sell rates should climb steadily with launch-week conditions. They don’t.

The table

Index units: 100 = the tier’s trailing five-year norm (methodology published here). “Sold” means the campaign ever closed, at any price. Days and discount are medians among those that sold; the discount is measured from the original asking price.

Read the middle four rows. From an index of 60 to an index of 250 — from a market a third below its norm to one running at two and a half times normal, a span four times as wide as ordinary year-to-year variation — the share of launches that ever sold moved from 47% to 51%. Days on market barely moved. The negotiated discount barely moved. Roughly three of every four campaigns in the record launched inside this band, and for all of them the launch-week market carried almost no information about whether the home would sell.

The honest check: inside a single era

Pooled tables can smuggle in regime effects, so we re-ran the linkage within eras. Within 2012–2019 — eight years of ordinary markets — the correlation between monthly launch-cohort sell rates and launch-month conditions falls to 0.28, and the spread from the weakest quarter of months to the strongest was about eight points (37% vs. 45–47%). Across the 2005–2011 regime transit the correlation is 0.74, and in 2020–2024 it is 0.68 — but in both cases the power comes from crossing a regime boundary (bubble to bust, ordinary to frenzy), worth twenty to fifty points. Stated as a rule: normal market fluctuation rarely changes a launch’s odds; regime changes change them enormously, but arrive perhaps twice a decade; the asking price matters every day.

What we think it means

The interpretation we place on this is deliberately conservative. Sellers, collectively, adapt their asking prices to conditions — in strong markets, ambition rises to absorb the extra demand; in soft ones, asks eventually concede. The market’s state is spent on the price rather than banked as probability, which is why better conditions show up in this record as modestly better speed and negotiation, not as better odds of selling at all. Within the wide normal band, the variable that separated sold from failed was not the week the campaign launched. It was the ask — the subject of One Shot, where the launch price’s association with outcomes is measured at every tier.

We have believed a version of this for years — you make your own market — as a working conviction from practice. This is what the record says when the conviction is put to 21,978 tests: within everything that counts as a normal market, the outcome belongs mostly to decisions the seller controls.

Caveats

All associations here are observational; launches are not randomized to market conditions. Cohorts launched into strong markets may differ from those launched into weak ones. Market conditions persist for months, so a launch week’s reading is correlated with the conditions the campaign subsequently lived through; this inflates, rather than deflates, the apparent power of launch-week conditions — making the flat middle more striking, not less. Campaigns launched near the end of the window have had less time to resolve; restricting to pre-2024 launches does not materially change the table. The index is measured in rebased units (each week relative to its own trailing norm), which is what permits comparison across eras.

This week’s index reading →  ·  One Shot: the pricing study →