London's property market is sending contradictory signals. Headline indices suggest moderate softening, but our quality-adjusted factor model — which strips out compositional changes in what's actually transacting — reveals a far more dramatic picture. The Baseline Market factor fell 35.7% over the twelve months to December 2025, the steepest year-on-year decline we have recorded. Meanwhile, new-build homes surged by 10.0%, reversing years of discounts.
These are not contradictory data points. They are two sides of the same story: a market repricing around quality, tenure type, and energy efficiency, while aggregate volume remains thin. Here is what the data actually shows.
Our model regresses log(price/sqm) against eight property characteristics every rolling quarter. The intercept — what we call the Baseline Market factor — captures the market-wide price movement after controlling for size, type, build period, energy rating, and tenure. It is the closest thing to a "like-for-like" London price index.
Over the twelve months to December 2025, that Baseline Market factor fell by 35.7%. To put that in context: the Brexit-to-COVID period (May 2016 to October 2019) saw the Baseline Market rise by roughly 10.9%, and the rate-rise correction from mid-2022 to mid-2025 produced a more modest shift. A 35.7% single-year decline is exceptional.
Why might the headline indices show a smaller fall? Because the mix of properties transacting has shifted. Fewer small flats are selling; more large family homes and new-build units are completing. When higher-value stock dominates the transaction ledger, the raw median price holds up even as the underlying market deteriorates. Our model adjusts for exactly this distortion.
The table below shows the period return for each factor over the twelve months to December 2025. These are computed as (1 + val_Dec25/100) / (1 + val_Dec24/100) − 1, expressed as a percentage.
| Factor | 12-Month Return (%) | Direction |
|---|---|---|
| Baseline Market | -35.7 | 🔴 Down |
| Floor Area | -9.4 | 🔴 Down |
| Floor Area (non-linear) | -1.1 | 🔴 Down |
| New Build Premium | +10.0 | 🟢 Up |
| Freehold Premium | -0.1 | ⚪ Flat |
| Flat Premium | -0.1 | ⚪ Flat |
| Energy Score | +0.3 | 🟢 Up |
| Build Period | +0.0 | ⚪ Flat |
| Room Count | -0.0 | ⚪ Flat |
Three factors dominate the story: the Baseline Market collapse, the Floor Area discount widening, and the new-build reversal. Everything else — freehold vs leasehold, flat vs house, energy rating, build period — was effectively static over the year.
The new-build factor's 10.0% gain in twelve months is the most striking reversal in our dataset. Since we began tracking this factor, new builds have almost always traded at a discount to equivalent second-hand stock. By mid-2025, the cumulative new-build factor sat at roughly +2.3% — already unusual. By December 2025 it had reached +12.5% cumulative, implying that new-build homes now command a meaningful premium over otherwise identical resale properties.
Several forces explain this. First, Help to Buy's withdrawal in 2023 initially depressed new-build demand, but the resulting supply contraction has now fed through: fewer new completions mean less competition among developers, and those units that do complete sell at firmer prices. Second, tighter Energy Performance Certificate regulations have made older stock less attractive, particularly for buy-to-let investors who face minimum EPC requirements. New builds — almost universally A or B rated — sidestep this entirely.
Third, mortgage lenders have begun to offer differential rates for energy-efficient properties, which disproportionately benefits new construction. The Energy Score factor itself only gained 0.3% over the year, but this likely understates the indirect benefit that flows to new builds through cheaper financing.
The Floor Area factor fell by 9.4% over the year, extending a trend that has been underway since mid-2022. This factor captures the per-square-metre premium or discount associated with property size. A negative return means that larger homes are now trading at a steeper discount per square metre relative to smaller ones.
This is consistent with a market where affordability constraints are binding. When mortgage rates sit above 4%, buyers simply cannot stretch to the total price that larger properties demand, even if the per-sqm rate looks reasonable. The result: larger homes see their per-sqm prices compressed, while smaller properties — more accessible on monthly payments — hold their per-sqm values better.
Combined with the Baseline Market's decline, this creates a double hit for owners of large family homes. The market-wide adjustment reduces all property values, and the floor area factor compounds it for bigger properties. A 150 sqm house in South London faces both the 35.7% baseline decline and the 9.4% floor area headwind. These do not simply add together — they interact multiplicatively in the model — but the directional message is clear: larger homes bore the brunt of 2025's correction.
Perhaps the most surprising finding is what didn't move. The Flat Premium factor was essentially flat at -0.1% year-on-year. After years of steady erosion — from roughly +10.0% in 2016 to close to zero by early 2025 — the flat discount appears to have stabilised. Flats are no longer losing ground relative to houses, though they remain far from recapturing the premium they once enjoyed.
Similarly, the Freehold Premium barely moved at -0.1%. The anticipated impact of leasehold reform legislation, which has been debated since 2023, has not yet materialised in the transaction data. Either the reforms are priced in, or — more likely — the practical implementation has been slow enough that buyers and sellers are still transacting on much the same terms as before.
For our earlier analysis of how these factors evolved through Brexit, COVID, and the rate-rise correction, see our three-part series: 2016–2019, 2020–2022, and 2022–2025.
The scale of the Baseline Market decline — 35.7% in a single year — warrants scrutiny. Some of this may reflect thin transaction volumes in late 2025 amplifying the model's estimates. We deliberately exclude the final two to three months of data because Land Registry reporting lags produce incomplete samples, but even the October-to-December window should be treated with some caution at this stage.
Three indicators will determine whether this decline deepens or stabilises:
London-wide averages mask enormous local variation. Our postcode tool shows rolling median £/sqm and factor exposures for every postcode area in London — from W1 to RM and DA.
Analyse a Postcode →Our factor model runs a rolling three-month cross-sectional OLS regression of log(price/sqm) on eight property characteristics for every Land Registry transaction matched to an Energy Performance Certificate. The intercept captures quality-adjusted market-wide price movements; the coefficients capture how the market prices each characteristic over time. Period returns are calculated as (1 + val_B/100) / (1 + val_A/100) − 1 and expressed as percentages. We exclude the most recent two to three months of data due to Land Registry reporting lags which produce incomplete transaction samples and noisy estimates. All data covers Greater London. Full methodology is available on our factors page.