Pending Home Sales +0.3% m/m but -4.7% y/y — Contracts Falling Faster Than Closings

Fundamentals · 2026-09-17

Pending home sales rose 0.3% m/m in August but fell 4.7% year over year, with gains in the South (+2.3%) and West (+3.0%) against declines in the Northeast (-4.2%) and Midwest (-1.6%) — and all four regions lower on the year. The forward read is worse than the monthly number suggests: contracts are falling 4.7% annually while closings, in our brief on the same month, fell only 1.2% — so the pipeline is thinner than recent completions imply. Among the 50 largest metros, the tenth-best market grew 0.2%, meaning at least forty were flatter than that. NAR's Lawrence Yun: contract signings are "roughly 30% below where they were in the years leading up to the pandemic." Pending sales lead closings by one to two months, so this is the September and October read. Next release 20 October.

What It Changes

• Contracts are falling faster than closings, and that is the whole point of this release. Pending sales are down 4.7% on the year; existing-home sales for the same month were down 1.2%. Because contracts lead closings by one to two months, the gap says the closings number still has room to fall.
• The monthly gain is real but tiny, and it happened into rising rates. +0.3% with the 30-year hitting 6.97% in the same window — its highest since May 2025 — is better than the rate move alone would predict. Yun's framing is fair: buyers "steadily entered into contracts... even though mortgage rates increased."
• The breadth is the alarming part. Among the fifty largest metro areas, the tenth-largest annual gain was 0.2%. To make the top ten in America you needed to be roughly flat. That is not a soft patch in a few markets; it is forty-plus large metros running flat to lower.
• Yun named a price mechanism worth testing. The Northeast and Midwest had the fastest price growth in August and posted the steepest declines in contract signings. Where prices rose fastest, buyers stopped signing.
• It rounds out a week in which every housing release said the same thing from a different angle — builder sentiment 32, mortgage applications -4.1%, single-family permits -1.8%, completions -27.1% year over year, and now contracts -4.7%.

Impact

• USD — Neutral — pending home sales has no dollar transmission, it publishes at 10:00 into a day already carrying five 08:30 releases, and it is a percentage-change release with no index level.
• US Indices (ES / NQ / YM / RTY) — Slight bearish, and only for one sector — this is a read on transaction volume, which is revenue for brokerages, title insurers, mortgage originators and moving-adjacent retail, not for the index.
• RTY carries it, as it has all week. Homebuilders are less exposed than they look — new-home sales are a separate channel and builders can buy volume with incentives, which 66% of them are already doing. The direct exposure is the existing-home transaction chain.
• Futures came in bid overnight — ES +0.85%, NQ +1.10%, YM +0.72%, RTY +0.79% — on 13% to 18% of average volume, ahead of Friday's triple witching.
• Gold (GC) — Neutral — no mechanism worth writing down.
• The honest note on all of the above: this release is second-tier at best, it lands 90 minutes after a five-release block, and it will not move an instrument. Its value is as a forward read on the housing transaction cycle.

Inside The Number

Pending home sales rose 0.3% in August from July and were down 4.7% against August 2025. NAR reports the release as percentage changes only; unlike its existing-home-sales report it does not publish the underlying index level, so there is no dollar figure or unit count to anchor against.

The monthly and annual numbers point in opposite directions, and the annual one is more informative. A 0.3% monthly gain on a sample covering roughly 40% of MLS data is well inside the noise of any single month. A 4.7% annual decline, with every one of the four regions lower, is not.

Now the comparison that gives this release its value. The Pending Home Sales Index measures contracts signed. Existing-home sales measures contracts closed, one to two months later. NAR states the relationship explicitly: monthly contract activity "parallels the level of closed existing-home sales in the following two months."

Set the two side by side for the same month. Closings fell 1.2% on the year. Contracts fell 4.7%. The pipeline feeding future closings is shrinking roughly four times as fast as the closings themselves. That is the single most useful fact in this report, and it means the existing-home-sales number — already at 3.98 million, its first print below 4.0 million since June 2025 and a third consecutive monthly decline — is more likely to keep falling than to stabilize.

The monthly gain deserves its due, though, because of what it happened into. August was the month the 30-year conforming mortgage rate reached 6.97%, its highest since May 2025, and the month builder confidence fell to 32. Contracts still rose 0.3%. Yun's reading is that "buyers steadily entered into contracts in August even though mortgage rates increased," and attributes the annual weakness to "higher mortgage rates offsetting the increased buying power created by job gains and income growth outpacing home price growth."

That last clause is worth separating out, because it is a genuine two-sided argument. Incomes are growing faster than home prices — the August median existing-home price rose 1.6% on the year against wage growth comfortably above that — so affordability should be improving. It is not, because the rate is doing more damage than the income gain is repairing.

Yun's regional explanation is testable and it fits. He said the Northeast and Midwest "saw the fastest home price growth in August, which is part of the reason that those same two regions posted the steepest declines in contract signings." And indeed those are the two regions that fell month over month, by 4.2% and 1.6%, while the South and West rose. Where prices moved up fastest, buyers stepped back fastest.

The long-run framing is the line to keep. Yun: "Nationally, contract signings today are running roughly 30% below where they were in the years leading up to the pandemic." Not below the 2021 peak — below the *normal* years that preceded it. Transaction activity peaked in 2021 when mortgage rates approached 3%, and NAR notes it "has not approached that level since." A market running a third below its pre-pandemic baseline five years later is a structural condition, not a cyclical dip.

Against the rest of this week, the housing data has been unusually consistent. Mortgage applications fell 4.1% with purchases down 1% and the 30-year at 6.97%. Builder confidence fell to 32, its lowest in a year, with buyer traffic at 23 and 38% of builders cutting prices. This morning single-family permits fell 1.8% and completions came in 27.1% below last year. Existing-home inventory has reached 1.62 million, a ten-year high in months' supply at 4.9. And now contracts are down 4.7% on the year in all four regions.

Every one of those is a different survey, a different agency and a different part of the transaction, and they agree. What they describe is not a collapse — it is a market where supply is building, prices are still rising modestly, and volume has simply stopped.

And for the Fed, this is the sector where the "not restrictive" claim looks weakest. Chair Warsh said on Wednesday he is "hard-pressed to call financial conditions restrictive," and that financial capital is not a constraint — a view 84% of Philadelphia manufacturers happen to share about their own businesses. Housing is the counterexample. A 6.97% mortgage has contract signings 30% below pre-pandemic norms and falling 4.7% a year.

The Internals

The national and regional picture, August 2026:

Region · Month over month · Year over year · Read

National · +0.3% · -4.7% · Up on the month, down on the year

West · +3.0% · -6.7% · Largest monthly gain, largest annual decline

South · +2.3% · -3.8% · Gained, and the smallest annual fall

Midwest · -1.6% · -4.9% · Fell on both horizons

Northeast · -4.2% · -3.9% · Steepest monthly decline

Contracts against closings, the comparison this release exists for:

Measure · August 2026 · Read

Pending sales, contracts signed, m/m · +0.3% · The forward pipeline

Existing-home sales, closings, m/m · -2.0% · Third consecutive monthly decline

Pending sales, y/y · -4.7% · All four regions lower

Existing-home sales, y/y · -1.2% · Falling far more slowly than contracts

Gap on the annual comparison · 3.5pp · Contracts shrinking roughly four times faster

Lead time NAR states · One to two months · August contracts feed September and October closings

The ten largest annual gains among the fifty biggest metro areas, per Realtor.com Economics:

Metro · Year over year · Metro · Year over year

Richmond, VA · +11.3% · Austin-Round Rock-San Marcos, TX · +4.2%

San Antonio-New Braunfels, TX · +6.6% · Birmingham, AL · +4.0%

Memphis, TN-MS-AR · +6.4% · Sacramento-Roseville-Folsom, CA · +1.7%

Virginia Beach-Chesapeake-Norfolk, VA-NC · +5.1% · Indianapolis-Carmel-Greenwood, IN · +0.9%

Cincinnati, OH-KY-IN · +4.7% · St. Louis, MO-IL · +0.2%

The week's housing data, all of it, in one place:

Release · Reading · Direction

Pending home sales, y/y · -4.7% · All four regions lower

Housing completions, y/y · -27.1% · Statistically significant

Mortgage applications, weekly · -4.1% · Purchases -1%, refinancing -9%

NAHB builder confidence · 32 · Lowest since September 2025

NAHB buyer traffic · 23 · Unchanged, at the floor

Builders cutting prices · 38% · Up from 35%

Single-family building permits · -1.8% · Year-to-date authorizations -2.5%

30-year conforming mortgage rate · 6.97% · Highest since May 2025

Existing-home months' supply · 4.9 · Ten-year high

Single-family housing starts · +7.6% · The one number going the other way

Regional And Metro Detail

The regional split is a clean North-South story and it inverted on the month. The South rose 2.3% and the West rose 3.0%, while the Northeast fell 4.2% and the Midwest fell 1.6%. On the year, though, every region is down, and the ordering reverses: the West is worst at -6.7% despite having the best month, and the South is least bad at -3.8%.

Yun's price explanation covers the monthly split but not the annual one. He attributed the Northeast and Midwest declines to those regions having "the fastest home price growth in August" — buyers balking where prices ran hardest. That works for the month. It does not explain why the West, with the largest monthly gain, carries the steepest annual decline, which points at affordability levels rather than recent price changes: the West starts from the highest price base in the country, so a 6.97% mortgage bites hardest there regardless of what prices did last month.

The metro table is the most revealing thing in the release, and it is revealing by accident. NAR lists the ten largest annual gains among the fifty largest metros. Richmond leads at +11.3%, and then the list decays fast: San Antonio +6.6%, Memphis +6.4%, Virginia Beach +5.1%, Cincinnati +4.7%, Austin +4.2%, Birmingham +4.0%. By eighth place Sacramento is at +1.7%, ninth is Indianapolis at +0.9%, and tenth is St. Louis at +0.2%.

Read what that implies. These are the *best* ten of the fifty largest markets in the country. If the tenth-best grew 0.2%, then at least forty of the fifty largest metros grew less than that, and given the national figure of -4.7%, most of them fell. Only seven of fifty managed growth above 4%. The national weakness is not concentrated in a few troubled markets; it is nearly everywhere, and the handful of exceptions are mid-sized Southern and Midwestern metros.

Where the gains are tells its own story. Richmond, San Antonio, Memphis, Virginia Beach, Birmingham, Austin, Indianapolis, St. Louis, Cincinnati — nine of the ten are Southern or Midwestern, and most are markets with median prices well below the national figure. The single coastal entry is Sacramento, at +1.7%. Affordability is doing the selecting: contracts are being signed where housing is cheap relative to income, and are not being signed anywhere else.

On the data itself, two limitations worth stating. The index is built on a sample covering roughly 40% of multiple listing service data each month, which makes single-month national moves of a few tenths essentially meaningless. And NAR does not publish the index level in this release — only percentage changes — and asserts redistribution restrictions on the underlying series, so there is no published level here against which to rank August historically.

The lead-time caveat NAR itself prints is also worth repeating, because it is the reason this release is watched at all and the reason it sometimes misleads. Contracts usually close within one or two months, but the gap varies, and NAR names the causes: "buyer difficulties with obtaining mortgage financing, home inspection problems, or appraisal issues." In a month when the 30-year hit 6.97% and refinancing applications fell 9%, financing failures are the most likely of those three to widen the gap — which would mean fewer of August's contracts reach closing than the usual relationship implies.

Against This Morning's Open

• The Open tabled this correctly and flagged honestly that it had nothing to forecast against — The Open listed pending home sales at 10:00 with consensus and prior both marked n/a, hitting USD and RTY.
• That n/a was the right call rather than a gap. NAR publishes this release as percentage changes with no index level, and it is not widely forecast — unlike the NAHB row yesterday, where a 34 consensus existed and was left blank.
• The Open's housing thesis was that a second soft month would be the first hard data against Warsh's "not restrictive" claim. Housing starts did not deliver that this morning. This release comes closer: contracts down 4.7% in all four regions, and running 30% below pre-pandemic norms.
• The calendar was complete for a second day — nine rows, including both Thursday fixed-cadence items and this 10:00 release, against three rows and five omissions on Wednesday.

What This Sets Up

• Next release — Tuesday 20 October at 10:00 ET, covering September, the same morning as the next housing starts report.
• The existing-home-sales print is the one this predicts. August contracts feed September and October closings. With contracts down 4.7% on the year against closings down 1.2%, the base case from this release is that closings keep falling from 3.98 million rather than stabilizing.
• Whether the South and West gains hold. Those two regions carried the entire monthly increase. If they roll over in September while the Northeast and Midwest stay negative, the national figure goes negative with them.
• Whether the metro breadth improves. The tenth-best of the fifty largest metros grew 0.2% this month. A list where tenth place is meaningfully positive would be the first sign the weakness is narrowing rather than spreading.
• Whether affordability or rates break first. Yun's own framing is that income growth is outpacing price growth, so the rate is the binding constraint. October's report is the first with a full month of the post-FOMC rate path behind it.

What Is This?

• What it is: The Pending Home Sales Index, published monthly by the National Association of Realtors, measures contracts signed on existing single-family homes, condominiums and co-ops — homes where a deal has been agreed but the sale has not yet closed. It is built from a sample covering roughly 40% of multiple listing service data each month, and indexed so that 100 equals the average level of contract activity during 2001, a year NAR chose because existing-home sales then ran at 5.0 to 5.5 million, considered normal for the US population. This release reports percentage changes nationally and for the four Census regions, without publishing the index level.
• Why it matters: It is the closest thing to a leading indicator that residential real estate produces. Existing-home sales — much the larger share of the housing market — records closings, which happen one to two months after the contract. So pending sales tell you what the closings data will say next. It is also the cleanest read on how buyers respond to mortgage rates, because the contract is signed at the moment the rate decision is actually made.
• How to read it: Four things. The annual change matters far more than the monthly, because a 40% MLS sample cannot resolve a 0.3% national move. The lead to closings is one to two months but it is not fixed — NAR names financing failures, inspection problems and appraisal issues as reasons contracts do not convert, and all three get more common when rates are rising. Compare it against existing-home sales deliberately: when contracts fall faster than closings, the closings number has further to go, and when they fall more slowly, closings are about to stabilize. And there is no index level in the release, so historical context has to come from the percentage changes and from NAR's own commentary rather than from a published series.

Want the full explainer? Economic News Events, Explained breaks down this release and every other one we cover, in plain English.

_For informational purposes only. Not investment advice._