Risks and vulnerabilities in Cambodia’s economy
Not a catalogue of fears, but a transmission map covering external demand, fuel, tourism, remittances, banks, property, data quality and available buffers.
Period: 2026-Q2 · checked 22.07.2026
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.
Risk is a transmission path, not a single event
For an apartment owner, an external shock matters through a chain: it changes exports, tourism, fuel or remittances; then employment and household costs; then rent, arrears, bank credit and resale liquidity. A headline alone does not reveal the strength of each link.
A risk map should separate probability, impact and capacity to absorb the event. Public reports usually describe scenarios and vulnerabilities but rarely provide a precise probability. We therefore do not manufacture scores or treat a forecast as guaranteed.
Buffers also matter: reserves, bank capital, fiscal space, deposits, insurance, expense reserves and the quality of the building. Country vulnerability and asset vulnerability overlap but are not identical.
External demand and export concentration
The United States took 40.7% of Cambodia’s merchandise exports in 2025 [3]. This concentration provides scale but exposes sectors to tariffs, rules of origin and demand in one market. Both the IMF and World Bank identify trade uncertainty as a risk, though their 2026 growth forecasts were released at different dates [1][2].
The first quarter of 2026 combined resilient exports excluding gold with faster import growth and a slowing GTF contribution after earlier frontloading [1]. This is not proof of an inevitable crisis; it is a reason not to apply one strong export percentage to every income group and district.
At asset level, the most exposed properties are those dependent on one employer, factory cluster or corporate group. A diversified tenant base does not remove macro risk, but reduces reliance on one channel.
Fuel and imported inflation
The spring 2026 fuel-price surge showed how quickly an external energy shock can pass into transport and food [1]. Cambodia imports energy, so a stable exchange rate does not remove the price pressure.
At property level, the channel runs through utilities, generators, delivery, staff and repairs. Tenant income is also squeezed if transport and food rise faster than wages. The same shock can therefore raise owner costs while limiting rent increases.
Buildings differ in exposure. Energy efficiency, backup systems, longer service contracts and a transparent budget reduce uncertainty, but do not guarantee an unchanged service charge.
Tourism and geographic concentration
International arrivals in January–May 2026 fell 47.8% year on year, mainly through the land channel [5]. Siem Reap and border destinations experienced a different shock from air markets or domestic visits to Phnom Penh.
Exposure is highest when a property depends entirely on short stays and one origin market. Tourism may be less important for a long-term capital-city unit than employment and education. A single “Cambodia tourism risk” should not be applied equally to every asset.
The buffer is not an optimistic annual forecast but a realistic operating model: seasonal occupancy, commissions, variable costs, building rules and an alternative long-stay audience.
Remittances, worker returns and domestic demand
Remittance inflows fell 23.6% in 2025 [3]. The World Bank linked the fall to worker returns from Thailand and a collapse in quarterly inflows late in the year [1]. For households using remittances for consumption and debt service, this is a direct budget shock.
Returning workers increase labour supply but do not immediately create income. The June update estimated that about 230,000 returnees remained unemployed [1]. Population present and solvent demand can therefore move in different directions.
For an apartment owner, exposure depends on tenant and employer profile. A unit serving foreign or highly skilled workers may respond differently, but remains connected to the wider service and consumption economy.
Banks, non-performing loans and property
The National Bank reported a banking-system gross NPL ratio of 8.9% in the Financial Stability Review 2025 [3]. A separate supervisory report shows 8.6% gross and 2.4% net NPLs [4]. Coverage, cut-off date and methodology may differ; both series are retained rather than averaged.
The net ratio is lower after provisions and other adjustments. It does not mean 2.4% is the “real” problem and 8.6–8.9% can be ignored. Both are relevant to asset quality and banks’ capacity to absorb losses.
A buyer still needs to verify land and project security, bank priority, release conditions and strata-title mechanics. System-wide capital does not repair a weak individual contract.
Property as its own risk channel
FDI into construction and real estate fell 32.4% in 2025 [3], the official residential price index declined, and commercial reports showed substantial supply. These are different signals, but together support stricter liquidity testing.
For a new project, risk includes completion finance, unsold developer inventory, management quality and post-completion competition—not just market price. For a completed asset, actual occupancy, rental history, building budget and common-area condition matter more.
An index decline does not mean every unit fell equally, just as a higher launch quote does not prove a higher transaction price. Asset liquidity needs comparable transactions and realistic marketing time.
Buffers: why vulnerability is not a crisis
International reserves stood near US$29 billion in February 2026—about eight months of imports according to the World Bank [1]. Public debt remained around 26% of GDP, and the IMF assessed debt-distress risk as low [1][2]. These buffers provide policy room.
Banks also retained capital and provisions, although rising NPLs required supervisory attention [3][4]. The accurate wording is “elevated vulnerability and a need for timely asset recognition”, not an automatic claim that the system is insolvent.
An owner’s buffer is liquid savings, insurance, limited leverage, clean title and the ability to absorb vacancy. National reserves do not pay a private apartment’s service charge.
Data quality is a risk category
Some series arrive with delays, are revised and cover different universes. Estimates of 2025 GDP ranged around 5.0–5.3%, while 2026 forecasts changed as new shocks appeared [1][2][3].
The property market lacks a public register of every realised transaction price, vacancy and project rent. An excessively precise model can hide weak inputs. Ranges and scenarios are more honest than a single decimal.
Every source needs a date and critical parameters should be refreshed before a decision. An old forecast can be correctly cited as a historical expectation, but not as a current fact.
How to build a risk profile for a specific apartment
Start with four layers: country, city, building and contract. Country covers external, currency and banking channels; city covers employers, transport and competition; building covers condition and budget; contract covers title, security, payments and responsibility.
For each risk record three elements: evidence, damage mechanism and buffer. A tourism decline matters only if the model depends on short stays; a low break-even occupancy or long-stay audience may provide resilience.
The output should not be a “Cambodia 7/10” score, but a list of conditions under which the asset can or cannot withstand a scenario. That format separates manageable risk from unknowns.
What this means for a property buyer
This page does not answer “buy or do not buy”. It identifies which external events can change income or liquidity and through what mechanism. A specific asset needs its own exposure map rather than the entire national risk list copied onto it.
The most dangerous combinations include weak title plus project security, high leverage plus vacancy, short-stay dependence plus one visitor market, or heavy unsold supply plus a need for quick resale.
Buffers also combine: completed title, low break-even cost, expense reserves, diversified tenants and transparent management reduce exposure. They do not guarantee returns, but make scenario consequences testable.
What we do not know
Open sources do not assign probabilities to most risk scenarios. A World Bank or IMF reference to downside risk is not a numerical probability of crisis.
NBC NPL ratios differ between publications. Without full technical reconciliation of coverage, we neither select one nor average 8.6% and 8.9% [3][4].
There is no single public series of realised transactions, vacancy and rents for all condominiums. Asset liquidity remains a matter for local due diligence.
Geopolitical and trade conditions can change faster than statistical releases. The last-check date is part of the content, not a formality.
Frequently asked questions
Does an 8.9% NPL ratio mean banks are insolvent?
No. It measures loan quality and must be read with capital, provisions and net NPLs. It raises risk attention but is not a diagnosis of the entire system.
Why are two gross NPL values shown?
They come from two NBC publications and may differ in coverage or cut-off. We show 8.9% and 8.6% separately rather than inventing an average [3][4].
Does a stable riel remove economic risk?
No. A stable exchange rate reduces one volatility channel but does not remove inflation, credit risk, tourism exposure, employment risk or asset liquidity.
What is the most important risk for a new project?
There is no universal answer. The combination of land rights, security, completion finance, sales pace and title issuance is often critical. Macro risk acts through that structure.
Does public debt near 26% guarantee safety?
No. It is a comparatively moderate buffer and the IMF assessed debt-distress risk as low, but it does not remove external, banking or private-debt risks [1][2].
How often should a risk profile be updated?
After a material source or event: a new NBC report, macro update, tenant change, building budget, security change or project milestone. Transaction-critical facts should be rechecked immediately before signing.
Where this leads next
Apply this to a specific property
Tell us the project and the goal — we will say which of these numbers actually bears on that decision and what still has to be confirmed in the building’s own documents.
WhatsApp Contact formInformational material based on public, dated sources. It is not a public offer and not individual investment, tax or legal advice, and no forecast here is a promise of price or yield. Figures carry the period and the release date of their source and may be revised by the issuing body. A decision on a specific property requires document, price and ownership-cost checks with an independent Cambodian lawyer.