Detailed Methodology

Kuwait Real Estate Indices

Tracks the real price of a standardized neutral plot against the 2019 baseline.

Detailed Methodology — Price Index

“Are prices higher or lower than normal, and by how much?”

Definition & purpose

Definition

The AlHisba Price Index (PI) measures residential price movement by tracking the price of a neutral plot per area — the area’s default size, full ownership, genuine sale, no premium features and no defects — and comparing it with the 2019 reference year. It answers: “are prices higher or lower than normal, and by how much?”

Purpose

Simple price averages move whenever the mix of deals changes: a quarter dominated by small premium plots looks like a price rise even if nothing became dearer. By fixing the measured object (the neutral plot), the index isolates true price movement, following the constant-quality principle used by official price statistics worldwide.

The Neutral Plot Standard

Plot Size
Area’s default size
Ownership
Full ownership
Transaction
Genuine sale
Premium Features
Excluded
Property Defects
Excluded
Base Year
2019 = 100

The index tracks this neutral plot per area and compares it with the 2019 reference year.

Price Index Formula

PI = Current Neutral Plot Price ——————— ÷ ——————— Neutral Plot Price in 2019 (Base Year) × 100
Current Neutral Plot Price
KD 165,000
2019 Neutral Plot Price
KD 150,000
(Base Year)
Price Index Result
110
+10% above the 2019 baseline

Why 2019? The last complete year of normal market activity before COVID-19, and the reference already used across AlHisba indices. The annual average is used — never a single quarter — for two reasons: a single quarter is one draw of thin-quarter noise that would permanently tilt every future value, and averaging the four quarters cancels seasonality so no season is baked into the base.

Publication levels

Kuwait (national)

One national residential Price Index: a value-weighted geometric composite of the covered areas’ indices. Published with an evidence-coverage statistic showing what share of the national weight was backed by areas’ own deals that quarter.

Areas

Unlike the Volume Index, the PI is published at area level. This is possible because the neutral price is a constant-quality value supported by the neighborhood trend in quiet quarters — it does not suffer the wild percentage swings that make small-sample activity indices unreliable. Quarters where an area had fewer than 5 own deals are marked ‘estimated’ (trend-following).

Categories

Private residential only at launch. Investment, commercial, and coastal segments are outside scope until their own neutral-value definitions are established.

Not indexed / labeled

Sabah Al-Ahmad Sea City and areas outside the residential register are excluded entirely. Three areas with no expert-appraisal history (Jahra, Ahmadi, Farwaniya) are published with an ‘estimated’ label and excluded from the national weighting until calibrated.

Data cleaning & exclusions

Source
Registered sale transactions published by the Ministry of Justice (MoJ), at individual deal level, covering 2010 to date — land and built homes together.
Quality filters
A deal is kept only if all hold: ownership share = 100%; contract is a genuine sale (بيع) — inheritance, gifts, auction awards and court transfers are excluded; property class is land, or land-with-house; price and size are positive and valid. Each deal is counted once.
Outlier prices
No manual deletion of unusual prices. Robustness is built into the method: the neutral value is computed from order statistics (price percentiles), which a single extreme deal cannot move, and quiet quarters are additionally supported by the neighborhood trend. Extreme index readings that survive are kept visible with a review marker — never silently removed.
Appraisal reference rule
The expert appraisal series used for calibration keeps one value per area-quarter: the last recorded entry; earlier entries in the same quarter are cancelled. An appraisal value proven wrong is removed entirely — never numerically edited.
Universe rule
The period value and the 2019 base always contain the same universe: every covered area has its own 2019 base, exclusions apply to both sides equally, and the national weight table may only contain areas that exist in the model’s area register.

Reading & comparing

> 100
Above the average
Prices above the 2019 average
< 100
Below the average
Prices below the 2019 average
≠ 0
Never reads zero
In quarters with no deals the area’s value follows its neighborhood trend and is labeled accordingly
1

All change is year-over-year. Published change compares each period with the same period of the previous year, never with the immediately preceding period, so that seasonal cycles do not masquerade as price moves — the same rule used across AlHisba indices.

2

Significance bands. Moves under 2% are within the method’s measured noise and read as stable; 2–5% is an indicative move; above 5% is a significant move. These thresholds come from the method’s validated accuracy (about ±5%), so readers are never alarmed by statistical noise.

3

Base effects & annotations. When history itself was abnormal, readings can look extreme — e.g., premium districts peaked in Q2 2023 then corrected by 30% or more; such movements are published unaltered with an explanatory annotation. Values flagged for review remain visible with a marker.

4

Smoothing. No smoothing line is applied to the PI and no filter touches the data. Stability in quiet quarters comes from the method itself (neighborhood-trend support proportional to how little evidence exists) — evidenced movements pass through in full.

Preliminary values & revision

1

Running quarter

During a quarter, the shown value is a Preliminary Index: it recalculates as registrations arrive, naturally starting from the neighborhood trend and converging to the area’s own deals as they accumulate, on the same 2019 base.

2

Finalization

A quarter’s value is declared final after quarter end plus a registration grace period, and the final value is frozen: it becomes the fixed comparison base and is never restated silently unless proven wrong.

3

Revision policy

Historical values change only at the annual chain-linking point (weight refresh), with a public log. Confirmed corrections to the calibration reference are applied by removal, never by silent editing.

Limitations

What PI does not measure
The Price Index measures price levels, not the number of deals. Deal counts are the separate Volume Index — and the two can move in opposite directions: prices can fall while trading volumes rise, and vice versa. Read them together.
Not a property valuation
The index tracks the neutral plot. Your specific property may carry premium features (corner, frontage, finishing) priced above neutral, or defects priced below — the index does not value individual properties.
Resolution
The method’s validated accuracy is about ±5%. Movements smaller than that need a second quarter to confirm; quarter-to-quarter wiggles under 5% are ordinary noise.
Quiet quarters
Where an area records very few deals, its value leans on the neighborhood trend — labeled ‘estimated’, never hidden. In market-wide quiet spells the national index leans more on trend, which the published evidence-coverage statistic makes visible.
Coverage
Registered private-residential transactions only. Unregistered arrangements, investment, commercial, and coastal segments are outside scope of first version. Minor differences from MoJ raw statistics stem from the quality filters above.