UK Home Sales Fell 2% in July: Why It's Already Old News

By Lisa Hayes · 30 August 2026

HMRC published its latest property transaction figures on Friday 28 August, and they show 96,710 homes changed hands across the UK in July 2026 on a seasonally adjusted basis. That is 2% down on June and 1% below July last year. Read the headline on its own and you would think buyers had gone cold. They haven't. These are completions, and a completion in July tells you what somebody decided to do in the spring.

I want to walk you through what the numbers actually say, because this release is one of the most misread in the property calendar. It is also one of the most useful, once you understand what it is measuring. If you are thinking about selling this autumn, the lag built into these figures is the single most important thing to grasp.

What did HMRC actually report?

The headline numbers from the 28 August release, all provisional:

• 96,710 UK residential transactions in July 2026, seasonally adjusted. Down from 98,390 in June, a fall of 2%. Down 1% on July 2025.
• 106,620 UK residential transactions in July 2026, non-seasonally adjusted. Up 3% on June and up 5% on July 2025.
• 10,350 non-residential transactions, seasonally adjusted, broadly flat on June and 2% below July 2025.

HMRC compiles these from Stamp Duty Land Tax returns in England and Northern Ireland, Land and Buildings Transaction Tax returns via Revenue Scotland, and Land Transaction Tax returns via the Welsh Revenue Authority. This is not a survey, and it is not built from asking prices. It is a count of transactions on which tax has been declared. That makes it about as close to the truth of what is happening as the UK housing market gets.

The catch, and HMRC says so plainly in the release, is that the latest month is provisional. Not every return has landed when the figures are compiled. HMRC expects revisions and says the numbers usually settle after roughly three months. So treat 96,710 as a good first draft rather than a final answer.

Why is the same month down 1% and up 5% at the same time?

This is where a lot of coverage falls over, and it is worth two minutes of your time.

The non-seasonally adjusted figure is the raw count: 106,620 homes sold in July 2026 against 101,650 in July 2025. More houses genuinely changed hands this July than last July.

The seasonally adjusted figure strips out the predictable rhythm of the housing year. July is always busy. Conveyancers push to complete before the school holidays swallow August. Statisticians adjust for that so you can compare July with February without the seasonality drowning out the signal. Once adjusted, July 2026 comes in at 96,710 against 97,820 for July 2025, a shade lower.

Neither number is spin. They answer different questions. The raw figure answers "how many homes sold last month?" The adjusted figure answers "was last month strong or weak for the time of year?" The honest summary is this: a busy July in absolute terms, an average one once you account for the calendar.

So when you see one outlet running "home sales jump 5%" next to another running "home sales fall", you now know why. Neither is lying. One picked the number that made the better headline.

The most important line in the release is the one nobody quotes

Buried in HMRC's executive summary is this: "These statistics represent completions which are on average two to four months after an initial offer is made on a property. They do not necessarily represent the current strength of the property market."

That sentence should be printed at the top of every article about this data.

A completion in July 2026 is the tail end of a decision made in March, April or May. The buyer viewed the house, made the offer, had it accepted, got the survey done, waited on searches, waited on the chain, and finally exchanged and completed. July's figure is a photograph of the spring market, developed four months late.

And what was happening in spring 2026? Five-year fixed rates were still close to 4%, on Zoopla's tracking. Sentiment was reasonable. A buyer who agreed a deal in April was borrowing on materially better terms than a buyer applying today. The number we are looking at was baked before the summer rate rises had bitten.

Here is the uncomfortable implication. If completions from the good months are only running level with last year, the completions that reflect the summer have a harder job. Those will show up in the September, October and November releases. The next one lands on 30 September and covers August. That is the one worth watching.

How does July 2026 compare with a normal July?

Context matters more than month-on-month wobbles. Here is HMRC's own ten-year run of July figures.

July	Non-seasonally adjusted	Seasonally adjusted
2017	105,610	103,750
2018	103,690	98,900
2019	104,780	98,050
2020	80,660	71,990
2021	83,520	76,210
2022	110,030	103,710
2023	85,510	86,170
2024	97,320	92,590
2025	101,650	97,820
2026	106,620	96,710

Look at that column properly and the panic drains out of it. On the adjusted measure, July 2026 sits comfortably above the miserable July of 2023, above July 2024, and a whisker below July 2025. It is below the pre-pandemic run of 2017 to 2019, when the market routinely cleared 98,000 to 104,000 a month.

Not a crash. Not a boom. A market running at roughly 95% of its late-2010s pace, which has been the honest description for most of the past two years.

The unadjusted column is the highest July since 2022 and the third highest in the whole table. Plenty of people are still moving house. If your worry is that nobody is buying, this data says otherwise.

The year so far looks strong, and that needs a caveat

HMRC also publishes a financial-year-to-date comparison covering April to July. On the seasonally adjusted measure:

• 2022–23 421,980
• 2026–27 393,800
• 2024–25 368,100
• 2025–26 342,900
• 2023–24 332,410

On the face of it, a strong start to the financial year: 393,800 against 342,900 for the same four months last year, an increase of roughly 15%.

Now the caveat, and it is a big one. April 2025 was artificially awful. The stamp duty nil-rate thresholds fell on 1 April 2025, so an enormous number of transactions were dragged forward into March 2025 to beat the deadline. HMRC flags this explicitly: a large peak in March 2025 and a subsequent drop in April 2025. April 2025 is a hole in the base, and comparing this year against it flatters 2026 considerably.

Strip out the distortion and the fair reading is that 2026 is running ahead of 2025 and 2024, but still short of 2022. Improved, not transformed. Anyone telling you the market has come roaring back is quoting a comparison against a stamp-duty-shaped crater.

That episode also left a lasting mark on what moving costs. If you have not run your own numbers recently, our guide to how much it costs to sell a house covers the full picture: agent fees, conveyancing, removals and tax.

Why are volumes holding up when prices aren't?

This is the genuine puzzle of the 2026 market, and it is where the transaction data becomes useful to you as a seller.

Prices have gone flat. Nationwide reported annual growth of 1.8% in July, down from 2.2% in June, with the average home at £277,542 and monthly movement of just 0.1%. Lloyds, which runs the index formerly branded Halifax, had the average at £299,253 in July with annual growth of 0.1%, its slowest since November 2023. The ONS put annual growth at 2.0% in the year to June, down from 3.0% the month before. Zoopla's August index, published on 27 August, has annual growth at 0.9%, down from 1.3% in June.

Four indices, four methodologies, one direction: growth has drained away to somewhere between zero and two per cent.

Yet transactions are broadly steady. How?

Because the constraint is not desire. It is arithmetic. Zoopla's research puts it starkly: the average five-year fixed rate at 75% loan-to-value has moved from below 4% in January to around 4.8% in August. Zoopla calculates that a buyer who could support a £200,000 mortgage in January can now support roughly £182,000 on the same monthly payment. That is a 9% cut in buying power in eight months. To buy the same house without paying more each month, that buyer needs to find an extra £18,200 in deposit.

Rightmove's August index pointed the same way, reporting that the average mortgage rate it tracks rose from 4.95% to 5.09% over the month.

So buyers are still buying. They are simply buying less house. And when a large number of buyers all lose 9% of their firepower at once, the adjustment does not usually arrive as fewer sales. It arrives as lower prices. Sellers meet the market, or they sit.

That is exactly what Rightmove recorded in August: new sellers cut asking prices by 2%, or £7,360, to an average of £364,999. The biggest August drop in eight years. Rightmove also downgraded its 2026 forecast for asking prices from 2% growth to somewhere between flat and a 2% fall, and noted the number of homes for sale was near a twelve-year high for the time of year.

Put the two datasets side by side and the picture resolves. Transactions are being sustained by sellers accepting less. That is the trade being made across the country right now, month after month.

What about where you live?

HMRC's monthly release is a UK-wide count with no regional breakdown, so for geography you have to look at the price indices. Zoopla's August data shows the split clearly.

Prices are flat or falling across most of southern England. The South East is 0.3% below where it was a year ago and London is down 1%. Head north and west and the picture inverts: Yorkshire and the Humber up 1.7%, the North West up 3.1%, Northern Ireland leading at 5.4%, although even that is slowing.

The reason is affordability, not fashion. Where prices are high relative to incomes, a rate rise does more damage. Zoopla estimates a London buyer needs roughly £35,500 more deposit to hold their position, against about £10,200 in the North East. Southern buyers, and first-time buyers in particular, are the ones being priced out of their own plans.

One encouraging signal in the same report: for the first time since August 2025, buyer searches are higher than a year ago in every region and nation of the UK. Searches are up 7% nationally, the strongest annual rise in twelve months, with the South East (+8.9%) and East of England (+8.5%) leading. Agreed sales remain 6% below last year, but the gap is closing.

People are looking again. They are just looking with less money.

What does this mean if you're selling right now?

Five practical conclusions I would draw from this week's data.

One: the buyer pool is real. More than 106,000 households completed a purchase in July. This is not 2008. If your house has been on the market for three months without an offer, the problem is almost certainly price or presentation, not a missing market. Our guide on how to sell a house that won't sell works through the diagnosis honestly.

Two: price to today's borrower, not last year's. The person viewing your home in September has around 9% less to spend than the equivalent buyer had in January. A valuation anchored to what the neighbours got in 2024 will not survive contact with a lender's surveyor. Start with a realistic view of what your house is actually worth today.

Three: expect the process to take time. Two to four months from offer to completion is HMRC's own average, and an average is not a worst case. If you need to be out by Christmas, an offer accepted in October is already tight. Checking how long properties are sitting on the market in your area before you list is time well spent.

Four: chains are the biggest single risk. In a market where buying power shrinks mid-transaction, a buyer who was comfortable in June can be short in September when their mortgage offer needs renewing at a higher rate. That is how chains break. If yours does, our guide to repairing a broken property chain sets out the options.

Five: know what speed is worth to you. Some sellers genuinely cannot wait four months, whether because of a job move, a divorce, probate or a repossession clock. That is what cash house buyers and quick-sale firms exist for. They pay below market value, and you should know roughly what percentage of market value cash buyers pay before you speak to one. Sometimes that discount is a fair price for certainty. Often it isn't. The only way to tell is to hold it against a realistic open-market figure and a realistic timescale.

Three pricing moves that work in a 96,000-a-month market

Knowing the market is flat is not the same as knowing what to do about it. Here is what I would actually change.

Set the asking price at the number you would accept, not above it. The old game was to add 5% for negotiating room. That worked when demand outran supply and buyers expected to haggle upwards from an offer. With stock near a twelve-year high, a buyer with a shortlist of eight houses does not haggle. They skip yours and view the one that is priced properly. The negotiating room you built in is the reason they never walked through the door.

Take the hit in the first three weeks or not at all. Listing interest peaks almost immediately and decays fast. If the first fortnight brings viewings but no offers, the price is close but not close enough, and a modest reduction in week three usually converts. If it brings no viewings at all, the price is not close and a small trim will not rescue it. Sellers who shave 1% every six weeks spend nine months arriving at a figure they could have set on day one, and by then the listing looks tired.

Remove the reasons to discount before you list. In a buyer's market, every open question becomes a price reduction: a missing FENSA certificate, an unresolved boundary, a lapsed guarantee, an old survey nobody has read. Sorting the paperwork before listing costs very little and takes away the leverage buyers would otherwise use in week eight, when you have the least appetite to argue.

Is now a bad time to sell?

No. But it is a bad time to be optimistic about price.

The market is functioning. Transactions are near their five-year norm, buyer searches are rising across every region, and roughly a hundred thousand families a month are getting keys. What has gone is the cushion. When prices were rising 5% a year, an over-ambitious asking price got rescued by the market catching up. At 0.9% growth, nothing catches up. An over-priced house in September is still over-priced in January, and by then it carries the extra handicap of a long listing history that every buyer can see.

There is also supply to reckon with. Zoopla reports 5% more homes for sale than a year ago, and Rightmove has stock near a twelve-year seasonal high. Your house is competing against more alternatives than it would have been last autumn, in front of buyers with less money. Realistic pricing is not defeatism in that environment. It is the strategy.

If you want the longer view of where prices have moved and where forecasters think they are heading, our 2026 UK house prices analysis pulls the major indices together.

What should you watch next?

Three dates and one theme.

30 September. HMRC publishes August transactions. That release starts to capture deals agreed in early summer, when rates were climbing. If the seasonally adjusted figure holds above roughly 95,000, the market has absorbed the rate shock without losing volume. If it drops through 90,000, the slowdown has moved from prices into activity.

The Bank of England's next decisions. Base rate has been held at 3.75%. Mortgage pricing does not track base rate directly. It tracks swap rates and lender competition, which is why fixed rates have risen this summer despite the hold. Watch the swap market, not the headline.

The autumn Budget. Property taxation is under active discussion again, and uncertainty alone slows a market. Buyers hesitate when they think the rules might change. Whatever eventually gets announced, the weeks beforehand tend to be sluggish.

The theme underneath all three: affordability is the whole story in 2026. Not confidence, not supply, not sentiment. What buyers can borrow determines what they can pay, and what they can pay determines what you get. Every meaningful number in this market now runs through the mortgage market first.

The bottom line

July's 2% dip is not the warning it looks like. It is a mildly soft reading of a market that was, when these deals were struck, in reasonable shape. The financial year is running ahead of last year, more homes completed this July than last July in raw terms, and buyer interest is up 7% in every corner of the UK.

But volumes are being held together by price concessions, and that concession is running at something like 9% of buying power. If you are selling this autumn, that is your number. Not the index, not the forecast, not what your neighbour achieved eighteen months ago. Price against what today's buyer can actually borrow and you will sell. Price against memory and you will join the growing pile of homes that quietly withdraw unsold in the new year.

If speed and certainty matter more to you than squeezing the last few thousand out of the open market, see what the alternatives actually pay before you commit to a six-month listing. Compare your options and get offers side by side, then decide with the real numbers in front of you.

Sources: HM Revenue & Customs, UK monthly property transactions statistics, published 28 August 2026. Supporting data from the Zoopla House Price Index (August 2026), Rightmove House Price Index (August 2026), Nationwide House Price Index (July 2026), Lloyds House Price Index (July 2026) and the Office for National Statistics.