United Kingdom Edition
Data Method|

UK HPI Methodology for London

How the official UK House Price Index measures London house prices in 2026 — what the source data is, how hedonic regression and mix adjustment work, why borough-level figures are smoothed with a three-month moving average, and why every release is provisional.

Overview

The UK House Price Index (UK HPI) is the official measure of UK residential property prices, jointly produced by HM Land Registry and ONS. The London series within it is the headline reference for London average price, year-on-year change and first-time buyer values used across PropertyWiki's London cluster.

UK HPI uses completed-sale price data from the four UK land registries, cleaned and combined into a hedonic regression model. The model controls for the changing mix of property types, sizes and geographies selling in any given month, so the published average is comparable over time even when the mix of sales swings.

Methodology at a glance

ItemValue
Index publisherHM Land Registry / ONS jointly
Underlying sourceCompleted-sale price-paid data from the four UK land registries
MethodHedonic regression with mix adjustment
Geographies publishedUK, country, region, county, local authority (borough)
FrequencyMonthly
Typical reporting lagApproximately 2 months
Revisions windowBackdated revisions for several months as data arrives
Borough series smoothingThree-month moving average
February 2026 London average price£542,000 (-3.3% year on year)
February 2026 London first-time buyer average£463,000

Sources: GOV.UK About the UK House Price Index; ONS House Price Index QMI; GOV.UK UK HPI England (Feb 2026).

How UK HPI is built — step by step

  1. Land registries record completed sales. Solicitors lodge completed sales at HM Land Registry (England & Wales), Registers of Scotland and Land & Property Services (NI). Recording takes a few weeks after legal completion, which is itself usually a few weeks after exchange.
  2. Cleaning and exclusions. ONS / HM Land Registry exclude commercial sales, transfers between connected parties and clearly anomalous prices. New-build prices are included once enough transactions accumulate; cash sales and mortgaged sales are both included.
  3. Hedonic regression. A hedonic regression model assigns implied prices to property attributes (size, type, location, age) so the index can compare a basket of like properties over time, instead of being moved by a change in the mix of houses that happen to sell.
  4. Mix adjustment and weighting. Modelled prices are then weighted by the standard property mix to produce an average price for each geography. For London this prevents a heavy month of central flat sales (or a quiet month of family-house sales) from distorting the published average.
  5. Borough smoothing. Borough-level series apply a three-month moving average to reduce the impact of small monthly samples. Single monthly movements at borough level should always be read as directional rather than precise.
  6. Publication and revision. UK HPI publishes monthly via the GOV.UK statistical bulletin and the UK HPI data service. Each release revises earlier months as more land-registry data arrives, so historical figures can move slightly between releases.

PropertyWiki editorial rules

We apply the following rules across every London borough page so UK HPI figures are presented consistently and never overstate precision.

  • Always quote the reference month (e.g. February 2026), not the publication month, and label values as provisional when they are inside the typical revision window.
  • Treat single-month borough movements as directional only — compare against the rolling 6- and 12-month change for the same borough before drawing conclusions.
  • Never compare a UK HPI sold-price figure with a portal asking-price figure. They measure different things at different points in the transaction lifecycle.

Frequently Asked Questions

The UK House Price Index (UK HPI) is the official measure of average UK residential property prices, jointly produced by HM Land Registry, ONS and the devolved nations. It uses completed-sale price data from the land registries, weighted by property type, age and region, and is published monthly with a roughly two-month lag.

UK HPI uses completed sales recorded at the land registries. Recording typically happens weeks after exchange of contracts, and the index publishes about two months after the reference month. As a result, the latest UK HPI release shows the position roughly two months ago and is also subject to revision as more data arrives.

Mix adjustment uses hedonic regression to control for the changing mix of property types, sizes and locations selling in any given month. For London this matters because the share of central-borough flat sales versus outer-borough family houses can swing the simple average sharply, even when underlying prices haven't moved.

Borough-level UK HPI figures use a three-month moving average to smooth small monthly samples, but they are still noisier than the London-wide series. They are best read directionally over six- and twelve-month windows rather than month to month, especially in low-volume boroughs.

This page reflects UK HPI methodology as published by GOV.UK and ONS as of May 2026, including the methodological treatment of cash sales, new builds, repossessions and revisions described in the official quality and methodology information (QMI) documents.

PT

PropertyWiki Team

Editorial Team

Published: May 9, 2026

Updated: May 9, 2026

PropertyWiki's editorial team curates UK property data using HM Land Registry, ONS and GOV.UK methodology sources.