Cambodia construction and real estate in numbers

Indicators and their vintage

Each figure keeps two dates: the period it measures and the date it was published. A 2025 result released in mid-2026 was not available at the end of 2025, and a forecast is never shown as an observed outcome.

One market, three layers of evidence

Construction and property cannot be described by one table. The first layer is administrative: approvals, permits and stated capital. These show intent and procedural status, not construction start or completion. The second layer is completed stock and additions under a consultancy method. The third is prices, occupancy, vacancy and transactions, where public evidence is weakest.

The National Bank of Cambodia publishes a residential property price index and summaries of newly approved projects. Knight Frank estimates condominium, office, retail and landed-housing stock. The sources complement each other but cannot be mechanically merged because official indicators and commercial samples use different definitions, geography and observation dates.

A buyer can move from national and city supply to segment, district, project and unit. The closer the analysis gets to the asset, the thinner the public data and the more important primary documents, actual contracts and physical inspection become.

What the official price index shows

In December 2025, the National Bank’s residential property price index was down 3.8% year on year for Cambodia, 4.3% for Phnom Penh and 1.2% for the provinces ^[2]. The index describes direction and magnitude within the regulator’s model, not an average apartment price in dollars.

The index does not mean every Phnom Penh apartment lost 4.3%. The sample spans different assets, while a transaction depends on layout, view, floor, condition, title, furniture, seller urgency and payment terms. The index change cannot be applied automatically as a discount to a price list.

Period and frequency matter. A December year-on-year value compares December 2025 with December 2024; it is not the average change across the year. If the method or base is revised, the historical series should be updated with a note.

Approval is not construction is not completion

Fifty-five new residential projects were approved in 2025, up from 40 in 2024. Within that total, the National Bank reports 41 landed-housing projects versus 27 a year earlier and 11 condominium projects versus 12 ^[2]. Remaining categories and classification rules should be checked in the source table.

More approvals do not prove more completions. After approval, a scheme can rephase, slow, enter presales, seek finance or fail to begin. Future-supply analysis requires site evidence, contractor information, a programme and subsequent permits.

Stated capital for residential development projects reached US$982.4 million, 22.1% higher year on year ^[2]. This is an administrative value, not verified expenditure, sales or bank financing. It should not be labelled “property investment” without qualification.

Phnom Penh condominiums: stock and completions

Knight Frank estimated existing Phnom Penh condominium supply at 63,334 units in H2 2025, up 9.6% year on year ^[4]. This is a consultancy estimate under its market boundaries and inclusion rules, not an official register of every title or occupied unit.

A total of 1,770 units completed in the half-year: 895 at Vue Aston and 875 at La Vista One ^[4]. The concentration in two schemes shows how completions can change competition in particular locations and segments even when the citywide figure appears manageable.

Existing stock is not the same as units available for sale or rent. Some homes are owner-occupied, leased, held off market or retained by developers. For an asset, the number of genuinely comparable competing units and their absorption matter more.

Segments and new-launch pricing

Knight Frank divided existing condominium stock into Prime at 4%, High-end at 17%, Mid-tier at 56% and Core at 23% in H2 2025 ^[4]. These categories belong to the firm’s method and cannot be mapped automatically to “luxury” or “affordable” language on a project page.

The average asking price of new launches was US$676 per square metre of net saleable area ^[4]. It covers H2 2025 launches, not all new homes, resales or completed transactions. The low average can reflect a launch mix concentrated in Core and Mid-tier rather than a uniform citywide price fall.

Price per square metre depends on the area convention. Net saleable, internal and gross floor area produce different denominators. A comparison with US$676/m² first requires matching the area definition and price status; otherwise the figure creates false discount precision.

Offices: large stock and low average occupancy

Phnom Penh office supply reached about 1,040,255 square metres of net lettable area in H2 2025, up 2% ^[4]. Average occupancy was 58.7%, down 2.7 percentage points, while Prime occupancy was 76%, down 5 points ^[4].

Office indicators matter to housing through employment and amenity demand, but the link is not automatic. A new office can redistribute existing tenants between buildings without creating many jobs. Remote work, corporate consolidation and building quality also affect occupancy.

Future office supply was estimated at 746,680 square metres NLA ^[4]. A pipeline is not guaranteed completion: timing can move, and schemes can be rephased or repurposed. District analysis should identify projects with actual construction and a credible programme.

Retail space: vacancy as a local signal

Phnom Penh retail supply was estimated at 875,464 square metres NLA with occupancy of 61.8% in H2 2025 ^[4]. The corresponding 38.2% vacancy within the same method signals pressure in parts of the market, but does not mean 38.2% of every shop in the city is empty.

Retail vacancy can mean two different things for a residential project. In a weak complex, empty units reduce convenience and footfall. In a new district, vacancy can be a temporary result of early-stage occupation. Building age, tenants, operating history and observed traffic help distinguish the scenarios.

The retail pipeline was 272,491 square metres NLA, and the source explicitly warned about possible postponements ^[4]. A future mall should not be treated as an existing apartment amenity until construction, tenants and opening timing are credible.

Landed housing and industrial construction

Knight Frank estimated existing landed-housing supply at 95,051 units in H2 2025, up 1.5% year on year. The half-year added or expanded 446 units, while the average asking price of new launches was US$1,547 per square metre GFA ^[4]. That area convention is not comparable with condominium NSA.

The National Bank reported a 124.9% increase in approved industrial construction area and an 84.2% increase in project count in 2025 ^[2]. Administrative momentum can point to a larger industrial pipeline, but housing demand requires evidence of actual construction and jobs.

Landed housing and condominiums serve partly different households. An apartment buyer should not treat all 95,051 homes as direct competitors, although more affordable peripheral landed housing can affect family choices as transport and schools improve.

Testing a specific project against the market data

For an off-plan project, build a competitive set of assets with similar location, price, stage and format. Compare not only prices but area on one convention, payment schedule, title, service charge, furnishing, parking, timing and actual progress. A citywide average does not replace the set.

For a completed building, check occupation in the evening and at weekends, active comparable units, price reductions, signed rents, time to lease and common-area condition. Management information is useful but should be supported by leases, invoices and observation.

When the developer still sells identical units, it may be the owner’s strongest resale competitor. Check remaining stock, current discounts, instalment terms and agent commissions. A fully paid resale can compete through immediate transfer as well as price.

Macro data are a reasonableness check on a sales presentation. If a project promises rapid appreciation during a period of a falling official index and substantial comparable completions, that does not prove the opposite outcome—but it does justify asking for the calculation, source and scenario behind the claim.

What we do not know

There is no single open register of every completed, sold, occupied and listed Phnom Penh apartment. Stock estimates depend on consultancy methods, while administrative data do not capture actual use.

Building-level transaction prices, discounts, time on market and net rent are not systematically public. Listings show supply intentions, not completed deals.

Approved-project counts do not provide completion probability or timing. Without subsequent site monitoring, 55 approvals cannot be converted into a completion forecast.

Office and retail pipelines can be delayed or changed. Published area does not mean the project will open on schedule or in the announced format.

The official index does not reveal the price of a particular unit and may be revised. Without detailed microdata, the roles of sample composition, negotiation, asset characteristics and the wider cycle cannot be separated precisely.

Frequently asked questions

Where this leads next

Sources

The sources cited on this page, numbered in order. Each one is named with its issuing body and release date, because a figure without a vintage cannot be checked for staleness. We do not publish outbound links — the document name and the institution are enough to find and verify it yourself.

• [2] Financial Stability Review 2025 — National Bank of Cambodia — 28.03.2026
• [4] Cambodia Real Estate Highlights, H2 2025 — Knight Frank Cambodia — 02.2026