Saudi Real Estate Price Index: What GASTAT's REPI Measures

The Saudi real estate price index rose 1.3% year on year in Q2 2026, to 106.34 on a 2023 = 100 base, and it is mostly a land number. Residential plots carry 45.849% of GASTAT's basket and rose 6.3%, to the highest level the series has printed; villas carry 10.455% and fell 9.7%, to 94.89. Land of all three kinds is 70.62% of the basket (GASTAT, Q2 2026). Rebuild the print from GASTAT's own weights, as we do below, and the land block accounts for more than the whole move while built property subtracts. The release contains no price level, no city and no transaction count.

Key figures at a glance

What the Saudi real estate price index measures

REPI tracks the price change of registered transactions on a 2023 = 100 base: quarterly, three sectors, eight property types, 13 administrative regions, latest print Q2 2026. GASTAT's methodology report, last updated 14 April 2025, names the statistical unit as "the registered real estate transaction" (GASTAT). A listing that never sold does not enter it, and neither does a valuer's opinion.

The formula is Laspeyres, aggregated from strata of one property type at one location, each holding "a minimum number of transactions" that GASTAT never states. Records arrive weekly, in coordination with the Real Estate General Authority and the Ministry of Justice, enriched with census data and satellite imagery through a geospatial AI model, and the commercial or residential split is inferred from street width rather than declared. REGA publishes its own real estate indicators platform off the same flow. Two field values on the methodology page matter more than their wording suggests: "Adjustment: Not applicable" and "Micro-data access: Not available".

One definition on that page decides how the whole product reads. A weight is "the ratio of the total value of transactions for one real estate type to the total value of transactions for all real estate types in the base year" (GASTAT). A share of 2023 transaction value, not of the housing stock. Villas sit at 10.455% because villa money was a tenth of the money that changed hands that year, not because villas are a tenth of Saudi housing. Read the weight as a stock share and everything drawn from the index afterwards is wrong.

Why the Q2 2026 headline is a land number, not a house price

Residential plots rose 6.3% year on year in Q2 2026 to 107.65, the highest reading in the published series, while villas fell 9.7% to 94.89: sixteen percentage points apart inside one +1.3% headline (GASTAT, Q2 2026). Plots outweigh villas by more than four to one, and even the line marked "residential" is mostly land, since plots are 45.849 of that sector's 72.661 weight points, or 63%.

Omran estimate: what moved the Q2 2026 headline. Multiply each component's published weight by the change in its published index level from Q2 2025 to Q2 2026, in index points of the general index. Residential plots (45.849%, 101.30 to 107.65) contribute +2.91; villas (10.455%, 105.03 to 94.89) take away 1.06; commercial plots (22.836%, 116.56 to 112.44) take away 0.94; agricultural land adds 0.22, apartments 0.17, and buildings, shops and floors net to -0.01. The three land lines sum to +2.19 index points against a published headline move of +1.34: land is 164% of the move and built property subtracts. The total comes to +1.29 rather than +1.34 because GASTAT publishes levels to two decimals. Only GASTAT weights and GASTAT levels enter.

Omran estimate: how often the headline is a land story. Run that decomposition across all 18 quarters for which the index has published a year-on-year change. Land moved the headline by more than built property in 16 of 18, carried at least 95% of the move in five, and carried more than the whole of it in three: Q1 2024 at 151%, Q3 2025 at 122%, Q2 2026 at 164%.

A Saudi price headline read as news about houses is, sixteen quarters out of eighteen, news about land. In Q2 2026 the two halves moved in opposite directions.

Two other GASTAT releases say what the villa number does not mean. Villa rents rose 2.65% year on year in the July 2026 CPI and apartment rents 5.02% (CPI); construction costs rose 2.26% in the year to July 2026, 2.17% in the residential sector (CCI). Residential land is up 6.3%, the cost of building is up 2.26%, villa rents are up 2.65%, and only the sale price of the finished villa fell. Nobody is paying less to live in a villa; what repriced is the transaction. A feasibility model that takes +1.3% as its escalation assumption gets the land line, the cost line and the exit price wrong at once. The Saudi residential market pillar takes the demand side apart.

The 2024 methodology break, and the basket it replaced

Two official dates govern the revision and they are not the same date. The Q1 2026 bulletin states that "data according to the updated methodology has been published starting from Q3 of 2024" (GASTAT); the methodology report says "these updates have been applied to the data from the year 2021". Publication start and data start. Conflating them is the standard error in Saudi price commentary.

The index that was replaced barely contained a house. As published for Q1 2024 on the 2014 base, the basket was 96.64% land (residential plots 62.16%, commercial plots 30.87%, agricultural 3.61%), villas carried 0.40%, and there were ten property types rather than eight (GASTAT, pre-revision tables). For Q2 2022 the old index reads +0.72% year on year where the current one reads +16.5% (GASTAT, Q2 2026 tables). The old release also carried an index for every property type in every region, which no release since Q3 2024 does: the record gained a credible villa weight and lost any way of asking what villas did in the Eastern Province.

Since the rebase the history has held still: no index level moved by more than 0.073 points between the Q3 2024 release and Q2 2026, and none has been revised across the Q4 2025, Q1 2026 and Q2 2026 files. The weights are what moved. Residential plots read 45.7178 in the Q3 2024 file against 45.849 in the Q2 2026 file, so "fixed 2023 weights" means fixed within a vintage, not across them.

Three official numbers for Saudi property prices in 2025

Saudi prices in 2025 rose 2.0%, or 2.9%, or fell 0.7%, depending which government publication you open. All three come off this one index and all three are correct. GASTAT's own annual-average change for 2025 is +2.0%, 104.32 against 102.27 (GASTAT, annual-average rows). The Ministry of Finance reports that "the average Real Estate Price Index (REPI) increased by 2.9% until Q3 of FY2025", commercial 7.0%, residential 1.5% (MoF FY2026). The third covers a quarter rather than a year: Q4 2025 printed -0.7% year on year, the first negative annual reading in the series (GASTAT, Q2 2026 tables).

The ministry's number travels furthest and is misread most often, because it looks like a full-year rate. MoF averages three quarterly index levels against the same three of the prior year: mean(104.93, 105.00, 103.88) over mean(100.59, 101.79, 102.56) gives +2.91%, printed as 2.9. The prior year's statement proves the method. Averaging the published quarterly rates for FY2024 returns 1.9% where the ministry printed 1.8%, and 1.2% for agriculture where it printed 1.0%; averaging the levels reproduces all four of its lines (MoF FY2025). Nine months, not twelve, and an average of levels rather than of growth rates.

The central bank uses a third window again, compounding annual averages: "as of 2024, the real estate index has increased by 20.8 percent since 2021", residential 19.3% and commercial 25.7% (SAMA). Naming your window and your method is the whole discipline, and most published commentary names neither.

What GASTAT does not publish, and what we build from it

The Q2 2026 release contains no occurrence of SAR, riyal, sqm or square metre, in the bulletin or the workbook, and no transaction count of any kind. It reports how a basket of registered transactions moved. It will not tell you what anything costs, and three further absences matter as much:

• No denominator. GASTAT promises "a minimum number of transactions within each stratum", never states it, and publishes no suppression floor or standard error. A regional print may rest on very few deals and the file will not say so: the Eastern Province swung from +6.9% year on year in Q1 2026 to -2.5% in Q2 2026 (GASTAT). Volumes sit in the ministry registry instead (see SREM transactions).
• No city, no district. Thirteen regions is the finest geography, and the Riyadh region is not Riyadh city. Levels also do not compare across regions, since each is rebased to 100 in its own 2023: Al Baha at 77.91 is not cheaper than Riyadh at 115.98, it is 22% below its own base while Riyadh is 16% above its own (GASTAT, Q2 2026). Growth over a common window does compare: since Q1 2021, Riyadh +55.9% and Al Baha -32.0%, which is where the regional series picks the story up.
• No rent, and no seasonal adjustment. REPI is built on sale transactions, so rents belong to the CPI and to Ejar, and the five-year Riyadh rent freeze in force since 25 September 2025 will never show up here (REGA). Adjustment is "not applicable" on a 22-quarter series, so nobody, ourselves included, can say how much of the +3.0% quarterly print is seasonal. No property type carries an age, quality or off-plan dimension either.

Omran estimate: the land index and the built-property index GASTAT does not publish. Split the eight components into land (residential plots 45.849, commercial plots 22.836, agricultural land 1.935, together 70.62%) and the five built types (29.38%), then re-aggregate each group as a weighted mean of its published index levels. Validation first: re-aggregated this way, all eight components reproduce the published general index across all 22 quarters, mean absolute deviation 0.06 index points, worst miss 0.19. The land index then reads 109.38 in Q2 2026, up 2.9% and at a series high; the built-property index reads 99.52, down 3.0%, a second consecutive quarter below its own 2023 base and a fourth consecutive negative annual print (-0.5%, -1.3%, -2.6%, -3.0% from Q3 2025). Land dipped to -1.0% in Q1 2026 and turned. Built property has not.

Two markets are moving apart inside one number, and the weighting settles which of them the headline reports. Policy is aimed at the larger half: more than 16 million sqm of developed but unbuilt white land meeting the criteria for the second phase of the white land fees was registered across 21 Riyadh neighbourhoods, per MOMRAH on 8 June 2026 (MOMRAH). This index will report the result as a price and never as a volume.

Omran rebuilds these series from the primary files, GASTAT workbooks vintage by vintage, REGA and Ejar, Ministry of Justice registrations, SAMA and Tadawul filings, and labels every derived number as ours (start with our map of Saudi public property data). Ask for the REPI tables with our land and built-property split, scoped to the sectors and regions you actually price.

FAQ

Why did the index rise 1.3% when villa prices fell 9.7%? Because residential plots carry 45.849% of the basket and villas 10.455%. On the decomposition above, the three land components contributed +2.19 index points in Q2 2026 against a headline move of +1.34, while villas subtracted 1.06.

Is the REPI a house price index? No. Land is 70.62% of the weight and built residential property 26.8%, so it behaves as a land price index with a housing component attached. That is how it printed +1.3% in Q2 2026 with villas at -9.7% and apartments at +1.1%.

Does GASTAT publish a Riyadh city price index? No. The finest geography is the administrative region, and the Riyadh region takes in Al Kharj, Al Majma'ah and everything else inside its boundary. The Riyadh region index reads 115.98 in Q2 2026, up 4.2% year on year.

What does the index say a square metre costs in Riyadh? Nothing. The Q2 2026 bulletin and workbook carry no currency and no area unit anywhere, so there is no price level in the product to read. For SAR per sqm you need transaction records, not an index.

Sources

• GASTAT, REPI Q2 2026 release page: link
• GASTAT, REPI Q2 2026 tables (levels, rates, weights, regions): link
• GASTAT, REPI Q2 2026 bulletin (headline and villa prose): link
• GASTAT, REPI methodology and quality report (unit, weights, Laspeyres, 2021 restatement): link
• GASTAT, REPI Q1 2026 bulletin ("published starting from Q3 of 2024"): link
• GASTAT, REPI Q3 2024 tables (first post-revision vintage, weights column): link
• GASTAT, REPI Q4 2025 tables (third vintage in the back-revision test): link
• GASTAT, pre-revision REPI tables Q1 2024, 2014 base (96.64% land, ten types): link
• GASTAT, CPI July 2026 tables (villa, apartment and imputed rents): link
• GASTAT, Construction Cost Index July 2026 (+2.26% YoY): link
• Ministry of Finance, Budget Statement FY2026 (REPI +2.9% through Q3 FY2025): link
• Ministry of Finance, Budget Statement FY2025 (the four lines that identify the method): link
• Saudi Central Bank, Financial Stability Report 2025, Box 2.1 (+20.8% since 2021): link
• REGA, Riyadh five-year rent freeze, effective 25 September 2025: link
• MOMRAH, White Land Program, Riyadh registrations (8 June 2026): link