Home Sales Just Had Their Best Month of 2026. Zillow's Own Economist Says That Was the Top.

The Best Housing Headline of 2026 Came With Its Own Expiration Date

Home sales jumped 7% in July compared to a year ago — Zillow's strongest annual gain of 2026. Zillow published that number and then, in the same report, told you not to trust it: "July was a strong month for existing home sales, but unfortunately it may represent the peak of what we can expect for the rest of the year," said chief economist Mischa Fisher.

That's not a hedge. That's the company that generated the bullish headline telling you it's also the ceiling. When your source for the good news is the same source calling it a top, you don't get to repeat the headline and skip the warning attached to it.

Here's why it matters today: every headline-skimming investor in your feed is about to see "home sales up 7%" and read it as a recovery signal. It isn't one — it's a snapshot of offers accepted back in June, when 30-year rates were still hanging around 6.5% before an oil-price shock pushed them higher. The forward-looking number in the same Zillow report — newly pending listings, the contracts that become September and October closings — grew just 0.3% year over year and fell 7.7% from June. The pipeline behind July's good month is close to empty.

Three Reports, Three Different Stories, Same Month

This is where it gets useful instead of just interesting: three of the biggest names in housing data looked at the same few weeks and each told a different story, because each was measuring a different thing.

Zillow measured closings and got a headline number: sales up 7% year over year, the best of 2026 — driven by contracts signed a month or more earlier.

NAR measured closings too, a few days later, and got a calmer story: existing-home sales fell 1.7% month over month but rose 0.7% year over year, with Chief Economist Lawrence Yun calling the market "remarkably stable" and year-to-date sales up 2.4%.

Redfin measured what's actually moving through the pipeline right now, and it's the least reassuring of the three: pending home sales are down 1.6% year over year and sitting at their second-lowest level since March, even after a small 0.4% weekly uptick to start August. The weekly average mortgage rate hit 6.69% — its highest level in over a year — and pushed the median monthly housing payment to $2,626, up 1.7% year over year.

Put them in order and you get a timeline, not a contradiction: June contracts closed strong in July (Zillow), the broader closing data confirms "stable, not booming" (NAR), and the newest contracts being signed right now are shrinking under a one-year-high mortgage rate (Redfin). The freshest data is the worst data. That's not noise — that's a market decelerating in real time, and the lag between "contract signed" and "sale reported" is hiding it from anyone who only reads headlines.

The Numbers

Read left to right and the story writes itself: prices are still climbing (37 straight months, per NAR), inventory is quietly building (+1.5% YoY), and the newest wave of buyers is signing contracts into a rate that's the highest it's been in a year. That combination — rising supply, rising rates, flat demand growth — is what a ceiling looks like from the inside, before it shows up in a headline sales number.

I'd put a date on it: when NAR reports August existing-home sales on Thursday, September 10, expect the year-over-year comparison to go flat or negative, and expect Redfin's pending-sales tracker to still show a year-over-year decline. If I'm wrong, it'll be because rates dropped meaningfully in the next four weeks — and at 6.69% with a Fed that's shown no urgency to cut, that's not the way to bet.

Common Mistakes Investors Make Here

• Citing the report that agrees with them and ignoring the other two. Bulls quote Zillow's 7%. Bears quote Redfin's -1.6%. Both are real numbers measuring different things — you need all three to see the shape of the market, not the one that confirms what you already believed.
• Confusing "prices still rising" with "market still tight." Prices climbing for a 37th straight month while inventory grows and pending sales shrink isn't strength — it's a market where sellers haven't gotten the memo yet. That gap is where negotiating leverage lives.
• Underwriting deals off last month's closed comps instead of this month's pending activity. Closed sales tell you what buyers were willing to pay when rates were lower. Pending listings tell you what's happening at today's rate. If you're only pulling comps, you're underwriting a market that already changed.
• Treating a one-year-high mortgage rate as a temporary blip instead of the current price of capital. 6.69% has been the trend, not the exception, for over a year. Plan your offers and your DSCR math around it, not around a rate cut that hasn't happened.

Where I Land

Zillow's own economist told you July was the top, and I believe him more than I believe the headline his own company put out. I'm underwriting every deal I look at right now assuming flat-to-declining transaction volume through year-end, not a recovery — the pending-sales data backs that up in every market I track, not just the national number. If I'm buying this quarter, I'm buying on price and terms that work at today's 6.69% rate, not a rate I'm hoping shows up by Thanksgiving, and I'm using the wider spread between rising inventory and slowing pending sales to push for concessions sellers weren't offering in June. The "market's turning around" crowd is reading one number out of three reports. Don't be the guy still quoting July's headline in October.

How to Use PropGPT for This

"Pull the pending-to-active listing ratio for [ZIP code / metro] over the last 90 days and flag whether it's decelerating faster than the national trend." This tells you if your target market is drying up ahead of or behind the national curve — critical for timing an offer versus waiting for a price cut.

"Compare closed sales volume to newly pending listings for [metro] month over month for the last 6 months and highlight any growing gap between the two." This is the local version of the Zillow-vs-pending divergence in this article — it flags markets where the "hot" headline number is about to catch down to a cooling pipeline.

"Screen for properties in [metro] with 45+ days on market and at least one price cut, then rank by estimated seller motivation." Rising inventory plus slowing pending sales means more sellers are sitting past their expected timeline — this finds them before they show up in a broader market report.

"Run a DSCR and cash-flow analysis on [property address] at a 6.69% rate and again at 7.25%, and flag if the deal only works at the lower number." If a deal needs a rate cut to pencil, it's not a deal right now — this stress-tests it against the actual current cost of capital instead of a hoped-for one.

"Set a weekly watch on new pending listings and price-cut share for [metro] and alert me if pending listings drop two consecutive weeks." Turns this article's "watch the leading indicator" advice into an actual recurring check instead of a one-time read.

The Bottom Line

July's sales number was real, and so was the warning attached to it. Zillow, NAR, and Redfin all looked at the same few weeks and handed you three different stories because they measured three different stages of the same pipeline — and the newest stage, pending sales, is the weakest one. Rates at a one-year high, inventory quietly climbing, and a company's own chief economist calling the good headline a ceiling isn't a setup for a fall selling season rally. Underwrite this quarter like the peak already happened, negotiate like the leverage is shifting your way, and don't let anyone's cherry-picked headline talk you out of the deal terms the pending-sales data says you can actually get.