Kite Property Index — methodology
What each series measures, where it comes from, what is suppressed and why.
v1-2026-11
Sources
- REGA — registered Ejar rental contracts, published quarterly per region.
- REGA Real Estate Indicators — every sale deal registered by Ministry of Justice notaries and by the Real Estate Registry, deal by deal, from February 2023. Used for derived figures only; no deal is republished.
- Kite — listings aggregated from 20+ Saudi sources, deduplicated across them.
Contracted rents from REGA's rent index are used under the Saudi Open Data License v2.0. Sale prices are Kite's own figures, derived from REGA's Real Estate Indicators. https://open.data.gov.sa/en/pages/policies/license
Coverage
| Value | Scope | From | To |
|---|---|---|---|
| Apartments and floors | City | 2023-04-01 | 2026-04-01 |
| Apartments and floors | National | 2023-04-01 | 2026-04-01 |
| Land | City | 2023-04-01 | 2026-04-01 |
| Land | National | 2023-04-01 | 2026-04-01 |
| Asking rent | City | 2026-03-01 | 2026-08-01 |
| Asking rent | National | 2026-03-01 | 2026-08-01 |
| Contracted rent | City | 2025-01-01 | 2026-01-01 |
| Contracted rent | National | 2025-01-01 | 2026-01-01 |
| Contracted rent | Neighbourhood | 2025-01-01 | 2026-01-01 |
How each series is built
Contracted rent, by city: a chained index over neighbourhoods. Each quarter is compared with the previous one using only the neighbourhoods REGA published in both, weighted by the earlier quarter's contracts, and the steps are multiplied together. The level is anchored on the first quarter's contract-weighted mean, so it still reads in SAR a year. A neighbourhood that appears or disappears does not move the figure. Neighbourhoods are matched by town and name, because the same name recurs in different towns of the Eastern Province and Qassim.
Land and homes, by city: the same chained comparison, over cells of one neighbourhood, one register (notary or Real Estate Registry) and one building use, each needing at least 8 deals in the quarter. Each cell's median price per square metre is compared with its own median the quarter before; cells are weighted by their deals, with the weights refreshed once a year, and the steps are multiplied together. The result is an index, 100 in the first quarter. Keeping the register in the cell matters: a notary deed that calls a property a plot may carry a building, while the Registry records vacant land as such, so a district that moved to the Registry leaves one cell and enters another instead of printing the difference as a price move. A city's own land or home series is published only where it rests on at least 500 deals in every quarter, moves less than 5% a quarter on average, and its quarterly changes do not simply reverse each other, which is the signature of sampling noise: for land Riyadh, Jeddah and Madinah, for homes Riyadh, Jeddah and the Eastern Province. Every city still counts in the national series.
National rent (contracted and asking): the same chained method over cities. A city counts from its second consecutive period, and one that stops reporting leaves the next step without moving the level. National land and homes: the cities with an index value in every quarter, Riyadh included, each weighted by its deals in the first quarter, held fixed for the whole series. A city missing any quarter is left out of the national series entirely.
Missing periods are shown as gaps, never filled in. The sale series begin in the second quarter of 2023, the first whole quarter of the source, and end at the last quarter that has closed. Where a quarter is missing, the next one is compared directly with the last one observed, and no change is quoted across the gap.
How to read these figures
- Mean rent on registered Ejar contracts, as published by REGA. City and national figures compare each quarter with the previous one over the same neighbourhoods and cities, so contracts moving between cheaper and dearer areas do not read as a rent change.
- Median asking rent for apartments listed on Kite. Monthly, so it leads the contracted series.
- Price index for land plots in registered sale deals — Ministry of Justice notaries and the Real Estate Registry, from REGA's Real Estate Indicators. Each quarter is compared with the previous one over the same neighbourhoods, within the same register, so a change in which plots sold does not read as a price move. The first quarter of each series is 100.
- Price index for apartments and floors in registered sale deals, per square metre of the unit's deed area, from the same source. Each neighbourhood is compared with itself, apartments with apartments and floors with floors, so a shift in what sold does not read as a price move. Villas and houses are left out: their deed area is the plot, not the home. The first quarter of each series is 100.
- Sale prices include Riyadh in every period. Riyadh's sales moved from notaries to the Real Estate Registry district by district from 2024, and wholly on 19 May 2026; the source carries both, and each neighbourhood is compared only with itself within the same register, so the move does not read as a price change. Deals covering several properties, and bulk registrations (20 or more deals in one district on one day), are left out.
- Asking and contracted rents are not directly comparable in level: asking rents are what landlords list, contracted rents are what tenants signed, and the two series cover different sets of places. Read each series against its own history.
Any cell with fewer than 5 observations is suppressed, not estimated.
Every dated release is kept as published: the page renders a frozen copy of its figures and never recomputes them. A correction is issued as a new revision carrying a visible note; a published figure is never edited in place.