Are UK House Prices Really Rising or Falling? Why the Headlines Don’t Agree

By early September 2026, you could find almost every version of the UK house-price story simply by choosing a different reputable index.

Rightmove reported that average new asking prices had fallen 2% in August and were 1% lower than a year earlier — the largest August monthly drop it had recorded since 2018. Nationwide, working from a different dataset entirely, had annual price growth at 1.6%. Zoopla put annual growth at 0.9%. Lloyds — the index formerly known as Halifax — recorded something more striking still: prices down 0.4% on the year in August, its first annual fall since November 2023. The official UK House Price Index, meanwhile, had average prices up 2.0% in the twelve months to June.

Five reputable sources. Five different numbers. That isn’t, by itself, evidence that any of them is wrong.

If it sounds contradictory, it’s because the question “are house prices rising or falling?” doesn’t have one single statistical answer. The major indices are observing different properties, at different stages of the selling process, using different methods and, sometimes, describing different months. Understanding that is one of the most useful pieces of market literacy a property reader can have, because it makes almost every future house-price headline easier to interpret.

The indices aren’t measuring the same event

Start with the basic split: some indices capture what sellers are asking, some use mortgage valuations and approvals, some incorporate prices at which sales have recently been agreed, and the official index ultimately works from completed transactions.

Rightmove’s index is built from newly marketed asking prices. It’s a property portal, not a lender or a land registry, so it catches a property at the point it comes to market, before negotiation has happened and before anyone knows whether it will sell for that amount. That makes the Rightmove House Price Index an unusually early read on seller sentiment and competitive pricing. It also means a movement in the index should not automatically be translated into “homes sold for this much less”.

When Rightmove reported August’s 2% monthly fall, part of the explanation was straightforwardly seasonal and competitive: there were more homes available than is usual for the time of year, and new sellers were having to work harder to attract buyers. That’s a real signal — about the price at which sellers and agents believe a property needs to enter the market — but it isn’t the same signal as a fall in completed sale prices.

Nationwide and Lloyds sit further along the transaction. Both use their own mortgage-lending data at the approval stage, after a valuation has taken place but before the final transaction is registered as complete. The Nationwide House Price Index is based on Nationwide’s owner-occupier house-purchase mortgage business; cash purchases and buy-to-let transactions are outside that sample. Lloyds’ index uses mortgage approval data from Lloyds Banking Group. The Halifax name disappeared in July 2026, when it became the Lloyds House Price Index, but the underlying methodology did not change.

That makes lender indices substantially more immediate than completed-sales data, but necessarily narrower. A cash buyer does not appear in either lender’s mortgage approvals, and neither lender sees every mortgaged purchase in the country. Two lender indices can therefore disagree even when they are describing the same calendar month — as Nationwide’s annual rise of 1.6% and Lloyds’ annual fall of 0.4% did in August 2026.

Zoopla tries to bridge parts of that gap. Its index uses sold prices, mortgage valuations and information from recently agreed sales. Zoopla describes the measure as tracking achieved sale prices rather than asking prices, and the index is revisionary as later information arrives. Its August 2026 House Price Index put annual growth at 0.9%.

Then there’s the UK House Price Index, jointly produced using data from HM Land Registry, Registers of Scotland, Land and Property Services Northern Ireland and the ONS. It uses residential transactions whether bought with cash or a mortgage and is based on actual sale prices rather than advertised prices or mortgage valuations. That gives the official UK House Price Index much broader market coverage, but at the cost of speed: registrations arrive after completion and the newest estimates are revised as more transactions feed into the dataset.

The differences are easier to see side by side:

The useful question, then, isn’t which organisation you trust most. It’s which stage of the market you are trying to understand.

Why the same period can tell two true stories

The clock is the part readers most often miss.

An asking-price index reflects decisions sellers are making now. Mortgage-approval indices come later. Agreed-sale data comes later again. A completed-sales index is describing deals that began their journey through the market weeks or months earlier.

So when Rightmove says new asking prices have fallen sharply in August while the UK HPI says completed prices were still 2.0% higher in the year to June, the two measures aren’t really arguing about one event. They’re looking at different points on the same chain.

A market that is cooling in real time can therefore show exactly this pattern: weaker new asking prices alongside still-positive completed-sale data. If that weakening persists, it may eventually appear in the later indices — but there is no rule saying it must, because negotiations, changing mortgage rates, the mix of homes sold and regional differences can alter the picture before those transactions reach completion.

The newest UK HPI figures deserve particular caution for another reason. The first estimate for June 2026 was based on sales volumes representing only around 47% of the sales in Great Britain that were ultimately expected for that period. More transactions are incorporated later and the recent figures are revised. That isn’t a flaw hidden in the data; it’s the trade-off involved in publishing an official transaction index before every relevant registration has arrived.

Northern Ireland adds another timing wrinkle. Its component of the UK HPI is available quarterly rather than monthly, so figures are carried forward between quarterly updates. Even within one “UK” index, not every part of the country is necessarily being refreshed on precisely the same timetable.

Monthly, annual, and the trap of the high base

A second, entirely separate source of apparent contradiction is the comparison period.

“Prices fell this month” and “prices are higher than a year ago” can both be true simultaneously. Nationwide’s revised July 2026 numbers are a neat example: its seasonally adjusted monthly index fell 0.1%, while its annual rate remained positive at 1.4%. One describes the movement from June to July; the other compares July with the same month a year earlier.

The annual figure also carries a complication that is easy to miss: the base effect. An annual growth rate can slow simply because the month dropping out of the twelve-month comparison was unusually strong.

The official UK HPI provided a good real-world example in summer 2026. Annual growth slowed from 3.0% in May to 2.0% in June, but the ONS specifically noted that part of this was because price growth a year earlier had been stronger following the April 2025 Stamp Duty Land Tax changes in England and Northern Ireland. In other words, the annual rate weakened partly because the comparison month moved.

Whenever a headline says annual growth is “accelerating”, “slowing” or has “turned negative”, it’s worth asking two questions rather than one: what are prices doing now, and what happened in the month a year ago that has just entered the comparison?

Same sample, different shape

Every index also draws from a different slice of the housing market, and that changes what the headline number represents.

Mortgage-lender indices exclude cash purchases by definition. Nationwide goes further and bases its index on owner-occupier purchases, excluding buy-to-let. The UK HPI, by contrast, includes market-value purchases whether funded with cash or a mortgage, and its transaction data covers a much broader section of the market.

Rightmove sees the extraordinary breadth of homes advertised through its portal, but an advertised property is not a completed transaction and some listings will never sell at the original price — or at all. Zoopla’s blend of recent agreed sales, valuations and sold-price information gives it a different shape again.

This also explains something else that regularly confuses readers: the average house price itself is not directly comparable between indices.

In August 2026, Rightmove’s average new asking price was £364,999, Nationwide’s average was £275,465 and Lloyds’ was £298,468. That doesn’t imply that somebody has misplaced £90,000 of value. The figures relate to different stages of the transaction and are standardised or modelled using different samples and methodologies.

For the same reason, averaging the five headline numbers together does not produce a more accurate “true” UK house price. It produces a number with no coherent methodology behind it. The movement within a consistent index over time is generally much more informative than comparing the nominal average price published by one index with the nominal average published by another.

None of this makes one index “correct” and the others “wrong”. It means each is a genuinely different cross-section of a market that does not behave uniformly.

Seasonally adjusted versus raw

A smaller but persistent source of confusion is seasonal adjustment.

Housing has a strong seasonal rhythm. Spring tends to bring greater activity; summer holidays can interrupt it; Christmas produces another predictable slowdown. If an index doesn’t compensate for that, part of a monthly rise or fall may simply be the calendar behaving normally.

Rightmove does not seasonally adjust its house-price index, which is why its own commentary compares movements with what it normally sees in the same month. The fact that asking prices tend to soften in summer therefore matters when interpreting an August fall — and the interesting part of the 2026 figure was that the 2% decline was considerably larger than Rightmove would usually expect for August.

Nationwide takes the opposite approach for its headline month-on-month movement: it seasonally adjusts the monthly index specifically to strip out the normal seasonal pattern. Its published average house-price figure, however, is not seasonally adjusted. Lloyds also standardises its series and applies seasonal adjustment, while Zoopla states that its index is non-seasonally adjusted. The headline UK HPI statistics are also published as non-seasonally adjusted estimates unless otherwise stated.

So “house prices fell 0.2% this month” is still not a complete piece of information. Before attaching too much significance to a small monthly move, check whether the seasonal pattern has already been removed.

The number that matters is rarely the national one

There’s a final layer, easy to miss amid the index-versus-index debate: even a perfectly consistent UK-wide measure can describe a national average that bears little resemblance to the market where you are actually buying or selling.

The regional split in 2026 has been substantial. Lloyds’ August index, for example, had annual prices rising 6.9% in Northern Ireland and 3.5% in Scotland, while the South East of England was down 1.6% and Greater London down 1.5%. Other indices draw their regional boundaries and samples differently, but the broad point appears repeatedly: stronger growth in parts of northern Britain and Northern Ireland has been sitting alongside much weaker conditions in several higher-priced southern markets.

That means a national annual figure of 1% or 2% can be perfectly consistent with one part of the country rising strongly and another falling at the same time.

Go more local again and the divergence can become greater. Different towns, property types and price bands within the same region can behave differently, particularly where the stock available for sale is concentrated in one part of the market. A national index is useful for understanding direction and context. It is not a valuation of your house.

That regional divergence is substantial enough to deserve its own explanation — which is where the next article in this series picks up.

A working framework for reading the next headline

Before treating any house-price story as decisive, it’s worth running it through a short checklist:

1. Which index? Asking-price portal, mortgage lender, blended achieved-price measure or completed-sales index — each answers a different question.
2. Asking, approved, agreed or completed? Prices can move differently at each stage of the transaction.
3. Monthly or annual? Both can tell different but simultaneously true stories, and the annual rate is influenced by what happened a year earlier.
4. Which geography? “UK” and “Great Britain” are not interchangeable, and a national average can conceal large regional differences.
5. Which property sample? Cash purchases, buy-to-let, new builds and different types of mortgaged buyer are not represented identically in every index.
6. Seasonally adjusted or not? Especially important when interpreting small monthly movements.
7. What period does the data actually describe? A headline published today may be describing decisions made several months ago.
8. Is the latest figure provisional? Recent completed-sales data can change as more transactions are registered.

For someone trying to understand the direction of the wider market, the best approach is usually not to choose a favourite index and ignore the rest. Watch the same handful consistently and give more weight to the pattern over several releases than to a single dramatic month.

If new asking prices soften, lender indices subsequently weaken and achieved or completed sale prices eventually follow, that is a much stronger signal of a market trend than any one headline on its own. If the indices continue to diverge, that divergence is information too: it may be telling you that seller expectations, buyer affordability and completed transactions are moving at different speeds.

For an individual buyer or seller, go narrower still. Recent comparable sales, current competing listings, how quickly similar homes are finding buyers and the balance between supply and demand in the immediate area will generally tell you more about one property than a national percentage ever can.

None of that makes house-price reporting untrustworthy. The major indices are produced using serious, published methodologies and each gives a genuinely useful view of a market that is, in reality, thousands of overlapping local markets rather than one national one.

The contradictions in the headlines aren’t usually errors. They’re the housing market being measured from several honest but different angles at once. Reading past the top-line number to what is actually being measured is what turns a confusing headline into useful information.

Figures and index methodologies in this article were checked against the relevant publishers and official UK HPI information and were current at publication in September 2026. House-price data is regularly revised, and later releases may supersede figures quoted here.