In September 2026, Redfin estimated 928,892 active homebuyers in the United States — the fewest since its records began in 2013 — against 1,510,784 sellers. On paper, that is the strongest buyer's market in well over a decade: more choice, more negotiating power, less competition. And yet buying a home has rarely been harder to pull off.
Both things are true at once, and holding them together is the key to understanding this market. Housing demand depends on more than the desire to own a home. A household may want to buy but lack the down payment. Another may need to move but be locked into a 3% mortgage it cannot afford to give up. A third may be able to buy but hesitate because its income outlook feels uncertain. Wanting, needing, and being able are three different populations — and right now, the gap between them is the story.
1. The current imbalance
Redfin's September 2026 estimates show sellers outnumbering buyers by 581,892 — a 62.6% seller advantage, the largest in the series and the fourth consecutive monthly record (August was +56.9%). Buyers fell 2.4% in the month, a fifth straight decline; sellers rose 1.2% to their highest level since March 2020. Of the 50 largest metros, 41 are buyer's markets, 4 are balanced, and only 5 remain seller's markets.
The gap keeps setting records
Seller advantage — how many more sellers than buyers there are, in percent
One country, two housing markets
Seller advantage by metro, September 2026 — of the 50 largest metros, 41 are buyer's markets, 4 balanced, 5 seller's
The remaining seller's markets are concentrated in the Northeast and Midwest, where little gets built: Nassau County, NY (−28.6%), Montgomery County, PA (−22.7%), Newark (−18.5%), Milwaukee (−16.1%), and San Francisco (−10.6%). One country, two housing markets.
What these numbers actually measure
Redfin's seller figure is a count: the number of active listings in the MLS — including newly built homes listed there, and listings that have sat unsold for months. Its buyer figure is a model, built from Redfin's own data on how long buyers typically take from first tour to closing, combined with MLS listings and pending sales. A market counts as balanced within ±10%.
Crucially, the model only sees people who are actively participating. A renter who would love to buy but never books a tour — because the monthly payment math fails before they start — does not appear as a "buyer" at all. That blind spot is exactly why a record buyer's market can coexist with record unaffordability.
2. The primary explanation: rates and the math
The clearest explanation for the deterioration is the combination of high mortgage rates, elevated prices, and a shrinking pool of buyers able or willing to transact. The financing shock has been fast: Freddie Mac's benchmark 30-year fixed rate climbed from 6.66% in late August to 7.40% by the week ending October 15, 2026 — up from 6.27% a year earlier. Home prices did not offset it: the median existing-home price was $417,500 in September, up 1.3% year over year and a record for the month (NAR).
The rate shock, week by week
Freddie Mac 30-year fixed mortgage rate, weekly readings
Run the arithmetic on a representative purchase — a $400,000 home, 20% down, a $320,000 loan (illustrative, using typical tax and insurance estimates):
The same house, two rate worlds
Monthly cost of a $400,000 home with 20% down (a $320,000 loan), illustrative
The rate alone adds roughly $867 a month on the identical house and loan. By the Atlanta Fed's affordability measure, a household needs an income of roughly $125,000 to afford the median-priced home, against a median household income of about $86,500. (Affordability indexes disagree in level — they use different down-payment and cost assumptions — but every version of the math points the same direction.)
Buyers are responding exactly as that math predicts. Mortgage purchase applications were down 16% year over year in the week ending October 9 (MBA) — a sixth straight weekly decline — and refinance applications were down 56%. Redfin's buyer count has fallen for five straight months as rates climbed. No more complicated explanation is required for the current downturn.
What is weakening the market
High mortgage rates, expensive homes, and reduced purchasing power. These forces align directly with the observed decline in purchase applications and buyer estimates.
What could weaken it further
Narrower entry-level hiring, persistent graduate underemployment, and uncertainty about future income — delaying purchases among younger households. Section 5 takes this up.
3. The blocked pipeline: renters who want to buy
If demand had simply evaporated, the story would end there. It hasn't. The Federal Reserve's Survey of Household Economics and Decisionmaking (2024) found that 30% of renters were trying to buy a home. What stops them is money, not desire: asked why they rent, 68% cited being unable to afford a down payment, 49% an unaffordable monthly mortgage payment, and 42% an inability to qualify for a mortgage (renters could select multiple reasons). To be fair to the data, renting is also a genuine choice for many — 58% cited convenience and 39% said they simply prefer to rent. Not every renter is a frustrated buyer. But tens of millions are blocked buyers.
Why renters rent: the wall is money, not desire
Share of renters citing each reason for renting, 2024 — 30% of renters were trying to buy
And their intentions are now collapsing under the math. The Philadelphia Fed's LIFE survey found the share of renters planning to take out a mortgage within six months fell from about 15% in January 2025 to 6.4% in January 2026. Longer-term belief is eroding too: in Redfin's February 2024 renter survey, 38% of renters said they didn't believe they would ever own a home, up from 27% a year earlier.
Intentions are evaporating
Renters planning to take out a mortgage within the next six months
The pipeline shows the strain at the entry point. In NAR's annual Profile of Home Buyers and Sellers, first-time buyers accounted for just 21% of purchases in the year through June 2025 — the lowest share since that series began in 1981 — with a median first-time buyer age of 40. (NAR's separate monthly survey read 31% for September 2026; the two instruments cover different periods and should not be compared directly — the annual figure is the structural trend.) When the first rung of the ladder empties out, the whole chain of move-up transactions above it slows too.
4. The supply puzzle: why owners keep listing
If buyers are frozen by 7.40% rates, why do sellers keep coming? Part of the answer is that the famous mortgage "lock-in" effect — owners refusing to give up cheap pandemic-era loans — is powerful but steadily eroding. ICE's Mortgage Monitor (July 2024 report, May 2024 data) found 76% of mortgage holders still had rates below 5%, with 5.8 million fewer sub-5% mortgages than two years earlier. The Federal Housing Finance Agency's National Mortgage Database shows the erosion continuing every quarter since:
Lock-in is eroding, quarter by quarter
Share of outstanding fixed-rate mortgages with rates below 5%
The average owner still holds a rate roughly three points below today's market — moving still hurts. But life does not wait for rates: new jobs, growing families, divorce, and retirement keep generating moves regardless of financing costs. Redfin describes sellers who have "simply given up waiting" for lower rates. And part of the rising seller count is arithmetic rather than enthusiasm: Redfin notes the number of homes on the market is rising "largely because many homes are lingering without finding buyers." Slow sales stack up.
Two smaller contributors deserve honest sizing. New construction is roughly 17–19% of listings nationally — and 25–37% of sales in major Texas metros — which helps explain the Sun Belt's extreme surpluses; MLS-listed new builds are counted in Redfin's seller figure, while unlisted builder inventory is not. And the largest investors (1,000+ homes) have been net sellers for ten consecutive quarters — but the volume is trivial, about 1,700 net homes in Q2 2026, while investors overall bought 3.6 times what they sold. There is no investor flood.
Myth-check: the ARM reset wave that isn't
A common theory holds that adjustable-rate mortgages resetting higher will force a wave of selling. The numbers do not support it. ARMs are just 3.1 million loans — 5.6% of the 55.4 million active mortgages (ICE, October 2026). First resets run at roughly 150,000 loans in 2026 and 186,000 in 2027 — under 1% of all mortgages per year. Overall serious delinquency is 2.03% (MBA, Q1 2026) and distressed sales are just 2% of transactions (NAR). The pain is real for a narrow cohort — about 74,000 seven-year ARMs written in 2020 reset in 2027 with a median payment increase around $1,066 a month — but concentrated pain is not a market-wide supply event.
5. The Next Risk: Not Just Whether Buyers Can Afford a Home, but Whether They Feel Secure Enough to Buy
The housing market's current weakness does not require an AI-driven employment shock to explain it. Mortgage rates have climbed sharply, home prices remain elevated, and mortgage purchase applications have fallen. These forces already provide a compelling explanation for why buyers are pulling back.
But there is a longer-term risk worth watching: the changing economics of early-career employment.
Recent college graduates face a difficult labor market, with underemployment reaching 42% in the New York Fed's June 2026 data. Research from Stanford's Digital Economy Lab also finds a widening employment gap among younger workers in occupations highly exposed to AI. The researchers point to weaker hiring as a possible mechanism, while cautioning that the findings do not establish AI as the cause.
These trends do not prove that automation is weakening housing demand today. In fact, youth unemployment improved year over year in September, even as housing activity weakened. That divergence suggests mortgage rates and affordability are more immediate explanations for the current downturn.
The concern is what happens next. Buying a home requires more than sufficient income today; it requires confidence that income will remain dependable for years to come. If entry-level opportunities become harder to secure, career progression less predictable, or household income more uncertain, some would-be buyers may postpone a purchase even when they remain employed.
That would create another barrier to homeownership—not necessarily a sudden wave of layoffs, but a gradual reduction in the number of households willing to make a large, long-term financial commitment.
For now, this is a risk to monitor, not an established explanation for the housing market's weakness. Mortgage rates and affordability remain the primary story. But if employment insecurity deepens while borrowing costs remain high, the next stage of the housing downturn could become harder to reverse.
The employment evidence, briefly
For readers who want the receipts behind section 5: BLS unemployment for September 2026 was 8.0% for ages 20–24 and 4.3% for ages 25–34 — both improved from a year earlier, even as housing weakened. The NY Fed's Survey of Consumer Expectations (September 2026) put the average perceived job-loss probability at 13.5%, the lowest since December 2024 — encouraging — while the perceived probability of finding a job if unemployed, 46.1%, remains far below its 56.0% reading of October 2024. Those two measures describe different things — fear of losing a job versus confidence in replacing one — and they can move in different directions. Stanford's updated analysis (August 2026) finds a roughly 19% kept-pace employment gap for 22–25-year-olds in the most AI-exposed occupations since late 2022, driven by reduced hiring rather than separations; the Yale Budget Lab, by contrast, finds no clear aggregate AI footprint on employment so far. Both can be true: a narrow junior-hiring squeeze inside a broadly stable labor market.
The next risk, in numbers
Two slow-moving series to watch alongside rates
If you lost your job, your chance of finding another
NY Fed Survey of Consumer Expectations — the average perceived job-finding probability. Fear of losing a job actually fell to 13.5% in September 2026, its lowest since December 2024; confidence in replacing one has not recovered.
Underemployment, ages 22–27 vs. all graduates
NY Fed, Labor Market for Recent College Graduates (June 2026). Recent graduates also faced 5.7% unemployment vs. 4.2% for all workers — an inversion of the historical pattern. Stanford's August 2026 update puts employment for 22–25-year-olds in the most AI-exposed occupations roughly 19% below the path of less-exposed peers — a descriptive pattern its authors say is not a causal estimate.
6. What would change the story
Three forces could rebalance this market. First, rates: the buyer count fell for five straight months as the 30-year rate climbed — a sustained move back toward the low-6% range is the single fastest lever. Second, prices: surplus metros like Miami, Nashville, and Austin are where adjustment pressure concentrates, because that is where the sellers are. Third, time: lock-in erodes mechanically every quarter as life events move people and loans turn over — roughly four percentage points of mortgages crossed above a 6% rate during 2025 alone.
Until one of those gives, expect the paradox to persist: the best negotiating conditions for buyers in years, available mainly to the shrinking minority who can afford to use them. The want is there. The need keeps listing homes. The ability is what's missing.
Sources and notes
Redfin, Homebuyers vs. Sellers, September 2026 and Buyers & Sellers data center (next update November 12, 2026; figures are seasonally adjusted, subject to revision, and drawn from the current published series throughout). Freddie Mac Primary Mortgage Market Survey via GlobeNewswire. NAR, Existing-Home Sales, September 2026, and NAR 2025 Profile of Home Buyers and Sellers. MBA, Weekly Mortgage Applications Survey, October 14, 2026. Federal Reserve, Economic Well-Being of U.S. Households in 2024, Housing. Philadelphia Fed, LIFE Survey renter brief, June 2026. ICE Mortgage Monitor, July 2024 (May 2024 data), via BusinessWire; ICE ARM data, October 2026, via MBA Newslink. FHFA National Mortgage Database via Realtor.com (Q4 2025) and Calculated Risk (Q3 2025). BLS, Employment Situation, Table A-10. Federal Reserve Bank of New York, Survey of Consumer Expectations, September 2026, and Labor Outcomes of Recent College Graduates. Stanford Digital Economy Lab, "Canaries in the Coal Mine" (updated August 2026). Yale Budget Lab, Tracking the Impact of AI on the Labor Market. Atlanta Fed Home Ownership Affordability Monitor. The $400,000 purchase example is illustrative, computed at the stated rate with estimated taxes and insurance; your figures will differ. Data as of October 20, 2026.
Educational content only — not financial, tax, or investment advice.
Run your own housing numbers
See how much house your income actually supports at today's rates — and what a one-point rate change does to the payment.