Britain’s housing market has flipped: where UK house prices have risen and fallen most since 2020

The order of the regional table is the headline. Every region above the Midlands has appreciated faster than every region in the South. London sits at roughly half the national figure. Its rate of negative resales is roughly four times that of the East Midlands.

Why UK flats have lagged houses since 2020

The split by property type is sharp.

Houses have gained 25% to 28% across all three types, with very few negative resales. Flats have not. The median flat is up 10.66%, less than half what houses have done. Almost one in seven flats (14.66%) sold for less than its 2020 price.

The reasons for the flat lag are not mysterious. Service charges have risen 41% on average between 2019 and 2024 according to the Property Institute. Cladding remediation costs and EWS1 form requirements have made many flats unmortgageable for periods. Leasehold ground rents have come under regulatory scrutiny. Working from home shifted demand toward houses with garden space. These pressures compound, and they hit older flats hardest.

This is also why the "new build flats sold at a loss" finding from Hamptons earlier in 2026 was the visible part of a broader pattern. The flat market across the UK has decoupled from the house market.

London property prices since 2020: the capital is its own market now

The London cut is the sharpest single finding in the data.

London houses are growing reasonably, in line with the East of England and South West. London flats are not. Almost one in five London flats sold in the last 12 months sold for less than the 2020 price. The median London flat has appreciated 7.69% over five years, which is below cumulative inflation in the same period.

The cladding crisis, leasehold reforms, the post-pandemic demand shift to houses, and the buy-to-let cooldown have all weighed on capital flats. Investors who bought central London apartments in 2020 expecting steady appreciation are, in many cases, sitting on real-terms losses.

For buyers, this is a reset. London flats are no longer pricing in a "London premium" relative to actual capital appreciation. Whether they will resume that premium when interest rates settle is an open question and one the data does not answer.

If you are looking at a London flat, pulling the full sale history from HM Land Registry is the single most important check you can run. Many flats have multiple recorded sales over the last decade and the trajectory is often visible.

The UK towns with the fastest house price growth since 2020

This is the bigger story in the data. For three decades, the property press has treated London and the South East as the engines of UK house price growth. That is no longer the case.

The 30 UK towns with the highest five-year growth in same-property prices are concentrated in the North and Midlands. Greater Manchester, Merseyside, Yorkshire and the Welsh valleys dominate.

A few patterns stand out. Five of the top fifteen towns are in Greater Manchester. South Yorkshire features four times in the top 30. Wales appears with both Swansea and Bridgend. Liverpool, Manchester, Bradford, Sheffield, Leeds and Plymouth all crack the list as major cities. Oldham is the single most surprising entry: a town often used as a stand-in for "left-behind Britain" in property coverage now leads the country on five-year price growth.

The UK towns with the slowest house price growth since 2020

The inverse list is dominated by London and the southern commuter and retirement belt.

Woking is the standout entry. Even excluding London, no UK town has a higher rate of below-2020 resales than Woking's 15.69%. Woking is a Surrey commuter town with a fast train to London Waterloo, traditionally a buyer's-market darling. In the last twelve months, almost one in six Woking resales went below 2020 prices.

The Brighton, Bournemouth, Poole and South Coast cluster is unusual: median appreciation is positive but well below the national figure. These are towns that boomed during the work-from-home shift in 2020 and 2021 and are now consolidating from those highs.

UK postcodes with the biggest house price changes since 2020

Town-level data smooths over a lot of variation within a town. The postcode-district view is sharper. We required at least 30 repeat-sale pairs per postcode district to qualify, giving 48 districts with enough data.

Only 14 postcode districts have median five-year growth below the national figure of +24%. They are heavily concentrated in inner-south-west London, the Brighton and Hove area, and the Surrey, Essex and Kent commuter belt.

CR0 (Croydon) is the most extreme entry. One in five Croydon resales in the last year sold for less than the 2020 price. SW11 (Battersea), SW16 (Streatham) and SW17 (Tooting) follow closely. The pattern is unambiguous: south-west London plus the inner south-east commuter belt are showing genuine market weakness, not just slow growth.

The fastest-rising postcode districts tell the opposite story.

WA5, on the western edge of Warrington, leads the country at +44.81%. Greater Manchester features three times in the top 17 (M28 Worsley, SK6 Marple, M20 Didsbury). Nottinghamshire features four times. The Stoke and Newcastle-under-Lyme cluster appears twice. Cardiff makes the list. Sheffield and Holmfirth bring Yorkshire in.

What this means for buyers and sellers

A few practical things follow from this data.

For buyers, the assumption that property always appreciates needs an asterisk. Five years is a meaningful holding period. If you bought a London flat in 2020, the median outcome is a 7.69% gain on the sale price, before stamp duty, legal costs, and any service charge or cladding contributions in between. After all those, real-terms returns are likely negative for a substantial share of London flat owners. This is not an argument against buying. It is an argument for checking the recent sale history of any property you are considering. Properties that have changed hands repeatedly in a short window often have a reason.

For sellers in the slow-growth South, asking-price benchmarks from 2021 and 2022 may now be misleading. Anchor on actual recent neighbour sales rather than the 2021 peak. Properties priced for the 2021 market are now sitting on the books.

For sellers in the high-growth North and Midlands, comparable evidence may understate where the market actually is. A Land Registry sale records a price six to eight weeks after offer agreement. In a fast-moving market, today's offers are ahead of the published comparables.

Across the whole market, some checks matter more than others. If you are considering a flat, especially in London, pull the full sale history. Service charge information should be available before exchange. EWS1 form status should be on file. If you are considering a house anywhere, check planning applications nearby (the median UK property has 12 within 500 metres in any 12-month window) and look at crime data and school provision before making an offer.

Move Insights covers all of this in a single £29 property report for any UK address.

How can I check what my own home is worth compared to its 2020 price?

The cleanest check for a single property is the HM Land Registry Price Paid Data lookup, which lists every recorded sale of a property. If your home sold in 2020 (or any year) you can see the price. Comparing it to comparable nearby sales in the last few months gives a defensible estimate of where the market sits today. We surface this for any UK address in the Move Insights sale history view.

Methodology

Source data: HM Land Registry Price Paid Data. The full national dataset covers transactions from 1995 to early 2026. The most recent transaction in this analysis is dated 27 February 2026.

Approach: Repeat-sales analysis at the address level. For each property in the dataset, we identified pairs where one sale fell in 2020 and another fell in the trailing 12 months ending 27 February 2026. Properties with no qualifying pair are excluded. Properties with more than two sales in the windows are matched to the latest 2020 sale and the most recent recent-window sale.

Match key: Internal canonical address ID, equivalent to PAON + SAON + street + postcode after normalisation. We verified a sample of pairs by full address: all matched pairs are confirmed same-property transactions.

Filters applied:

• New-build sales excluded on either side of the pair (is_new_build = 0 required on both)
• Leasehold flats below £100,000 excluded on either side, to remove retirement leases that depreciate by design
• Time gap between sales required to be 2 to 7 years inclusive, to keep the comparison roughly in a five-year window and exclude short-term flips

Filters not applied: The HM Land Registry ppd_category_type flag (B for non-standard transactions) is not stored separately in our dataset. The new-build filter covers a large share of category-B transactions in practice.

Geographic aggregation: Town label resolved via the town_city field from the Land Registry PPD CSV, which has 100% coverage in our enriched address table. Region inferred from postcode area via a fixed mapping covering the nine English regions and Wales. Some boundary postcode areas (CH, DN, HP, WD, EN) are assigned to a single region for consistency. Scotland and Northern Ireland are not covered by the Land Registry PPD and are excluded.

Town count notes: Some towns appear under both their unitary authority name and their county name in the source data (Southampton appears under both "Hampshire" and "Southampton"; Nottingham appears under both "City of Nottingham" and "Nottinghamshire"). For the rankings above, we have either combined the two entries or noted them as variants.

Statistical thresholds: Towns require at least 50 qualifying repeat-sale pairs to be ranked. Postcode districts require at least 30 pairs. Below these thresholds the median is not stable enough to publish.

Headline statistics use medians, not means. The mean of repeat-sale percentage changes is +131%, pulled up by extreme outliers including post-renovation pairs and probate flips. The median is +24%, which is what a typical UK home has experienced.

The data is not inflation-adjusted. Cumulative CPI inflation over the same period is roughly 22% (depending on month), which means a substantial share of the homes that show "growth" in nominal terms have actually moved sideways or backwards in real terms.

Total qualifying pairs: 21,722 nationally, after filters. 484 pairs were dropped for the low-value-flat exclusion. No pairs were dropped for the time-gap exclusion within the 5-year window.

Caveats and limitations

This analysis tracks transactions that completed and were registered. Properties that were marketed but did not sell, or where deals fell through, do not appear. The dataset is a complete record of sold homes, not a complete record of market sentiment.

Repeat-sale analysis at the local level requires a property to have changed hands twice in the windows. Towns with low transaction volume have fewer qualifying pairs and are excluded from rankings. Some towns we would expect to see (parts of rural Wales, much of the Highlands and Islands which sit outside this dataset entirely) are simply too small to qualify.

The dataset does not include Scotland or Northern Ireland. The UK House Price Index covers all four nations. This analysis covers only England and Wales because that is what the HM Land Registry Price Paid Data covers.

Most importantly, repeat-sales medians describe what happened to the average property in each area. Individual properties can move materially differently from their local median, in either direction. Use the medians as context for a specific decision, not as a forecast.

Data and source

Built from HM Land Registry Price Paid Data, covering all residential property sales registered in England and Wales. We refresh our copy of the dataset monthly. The figures in this analysis use data registered up to 27 February 2026.

For any UK address you can check the full Land Registry sale history, comparable nearby sales, planning applications, flood risk, crime, schools, broadband and EPC data on Move Insights from £29.