Three publishers print three different “UK house price” annual numbers in the same window. ONS says 0.0% (12 months to March 2026). Halifax says +0.4% (April 2026). Nationwide says +1.7% (May 2026). None is wrong — each answers a different question, for a different reference month. This brief decomposes what a single national figure bundles together.
The ONS UK HPI (0.0%, 12 months to March 2026) counts all completed sales including cash, priced at Land-Registry registration. Nationwide (+1.7%, May 2026) counts owner-occupier mortgaged purchases only, priced at mortgage approval. Different populations, different pipeline stages, different reference months — the 1.7 pp spread is structural. RAOSCAFF arithmetic: 1.7% − 0.0%.
In early June 2026 a reader asking “what are UK house prices doing?” meets three live answers. The ONS UK House Price Index says 0.0% annual growth, an average of £268,000, for the 12 months to March 2026, published 20 May 2026. Halifax (Lloyds) says +0.4% annual, average £299,313, for April 2026, published 8 May 2026 (primary-attributed/not-direct-primary). Nationwide says +1.7% annual for May 2026, published 1 June 2026.
Three publishers, three numbers, three reference months. The headline depends entirely on which question is being asked — and each of the three indices answers a different one.
The average-price gap between Halifax (£299,313, April) and ONS (£268,000, March) is £31,313 — approximately 11.7% of the ONS figure (RAOSCAFF arithmetic: £299,313 − £268,000 = £31,313; £31,313 ÷ £268,000 × 100 ≈ 11.7%). This gap is structural: it reflects three different buyer populations, not three different readings of the same houses.
A single “UK house prices” question fans out into three answers anchored to March, April, and May 2026 — each correct, each describing a different population at a different pipeline stage.
Axis 1 — Transaction stage. The GOV.UK methodology comparison places the three indices at different points in the conveyance. ONS uses “transaction data at the end of the conveyance process, calculated based on completed sales” at registration — the page notes the period between sale and registration “can exceed 2 months, particularly for new builds.” Halifax and Nationwide price “at the approval stage but after the corresponding valuation has been completed.” The ONS March figure and the Nationwide May figure sit at opposite ends of the pipeline: one looks backward to completed deals, the other forward to approved-but-not-yet-registered mortgages. Neither is wrong.
Axis 2 — Population and coverage. ONS is “based on both cash and mortgage transactions” — it includes cash purchases, a substantial share of the market; it is mix/quality-adjusted and completion-based (lagged). Halifax and Nationwide “will not include any cash transactions.” Within the two lender indices the coverage still differs: buy-to-let is “included in Halifax,” while “owner occupied properties only are used by Nationwide — buy to let properties are excluded,” and Nationwide is also mix-adjusted. ONS and Nationwide therefore represent two different defined quantities measuring one mix-adjustment idea — applied to different populations at different pipeline stages. Three different buyer crowds produce three different averages. The £31,313 average-price gap between Halifax and ONS (≈11.7% of the ONS figure, RAOSCAFF arithmetic) is structural — not a dispute about the same houses.
Axis 3 — Reference moment. The three “latest available” readings as of 4 June 2026 are anchored to March, April and May 2026 — a two-month spread (RAOSCAFF arithmetic: May 2026 minus March 2026). Because each index also captures a different pipeline stage, a truly shared reference period never exists across all three. A comparison of ONS March with Nationwide May is a comparison across both time and method.
| Publisher | Annual change | Average price | Reference month | Transaction stage | Coverage |
|---|---|---|---|---|---|
| ONS UK HPI | 0.0% | £268,000 | 12 months to March 2026 | Completed sale at Land-Registry registration (2+ month lag; March provisional on ~46% of GB sales) | All completed sales incl. cash; most comprehensive population |
| Halifax (Lloyds) (primary-attributed/not-direct-primary) | +0.4% | £299,313 | April 2026 | Mortgage approval stage, after valuation; earlier in pipeline than ONS | Own mortgage lending; includes buy-to-let; excludes cash buyers |
| Nationwide | +1.7% | Not disclosed | May 2026 | Mortgage approval stage; most forward-looking of the three | Owner-occupier mortgaged only; mix-adjusted; excludes cash and buy-to-let; narrowest population |
Methodology: GOV.UK — Comparing house price indices in the UK (direct fetch). Figures: ONS + Nationwide direct primary fetch; Halifax primary-attributed/not-direct-primary via named-spokesperson trade-press reproduction.
If you read one thing: whenever you quote a UK house-price figure, carry the publisher and the month with it — the headline depends on both.
In June 2026 the UK house-price headline depends on which publisher you read and which month they report. ONS (0.0%, 12 months to March 2026) prices all completed sales at Land-Registry registration — the broadest population, the most lagged stage. Halifax (+0.4%, April 2026) and Nationwide (+1.7%, May 2026) price at mortgage approval — earlier in the pipeline, narrower populations. The 1.7 percentage-point spread between the top and bottom annual figure is structural, driven by transaction stage × coverage × reference month, not by one index being closer to the truth than the others. The honest reading is not “which number is right” but “which question am I asking” — and the answer changes with every publisher and every month. RAOSCAFF Mirror Brief UK-01 decomposes the triangle and makes one claim: carry the publisher and the reference month whenever a UK house-price figure is quoted. Every figure here appears in FACTS.md §H; derived figures are labelled RAOSCAFF arithmetic. This brief critiques metric definitions, never a publisher; each index is stated to be correct under its own methodology; no ranking is implied. Politically neutral; no party, government, or minister named; no causal-on-policy market claim; no buy/sell/hold position; no forecast.
Mirror format — RAOSCAFF anchors on primary published figures and decomposes the divergence. ONS UK HPI (0.0%, £268,000, 12 months to March 2026, −0.4% NSA monthly) and Nationwide HPI (+1.7%, −0.6% MoM, May 2026) are direct primary fetch. The GOV.UK methodology comparison (Comparing house price indices in the UK) is direct primary fetch, supplying the transaction-stage and coverage distinctions. Halifax (+0.4%, £299,313, −0.1% MoM, April 2026, full regional table) is primary-attributed/not-direct-primary via named-spokesperson (Amanda Bryden, Halifax head of mortgages) reproduced consistently across trade press (Financial Reporter + Mortgage Solutions), and cross-checked against the Halifax historical series; every Halifax figure is labelled primary-attributed/not-direct-primary at each use. Sources 5–7 are corroborating cross-checks for Halifax, not direct-primary fetches. All derived figures — the 1.7 pp rate spread, the £31,313 average-price gap, the ≈11.7% gap-as-share, the two-month reference spread, the 9.6 pp regional range — are RAOSCAFF arithmetic from the published figures and are labelled as such throughout.
Predict-not-recommend. Defamation-disciplined: the brief critiques the metric definitions (stage, coverage, reference month), never a publisher; each index is stated to be correct under its own methodology; none is named the accurate one. Politically neutral: no party, government, or minister named; no causal-on-policy market claim; base-effect context reported descriptively as publishers state it. Reference-period precise: every figure carries its own publisher and month; the three numbers are never presented as the same moment. Figures are the latest available from each publisher as of 4 June 2026 and are superseded the moment any publisher issues a newer release. This brief is general information, not financial, investment, property, or mortgage advice; it recommends no course of action and takes no position on whether UK house prices will rise or fall.
ONS + Nationwide + both GOV.UK pages = direct primary fetch. Halifax = primary-attributed/not-direct-primary (named-spokesperson quotes reproduced consistently across trade press + historical series); labelled as such at each use. Sources 5–7 are corroborating cross-checks for Halifax figures. Full citations and arithmetic register in the companion FACTS.md, §H.