Phnom Penh's Retail Glut: 450,000 Square Metres Vacant, an 18-Year Absorption Clock

Cambodia’s real estate consultancies have a habit of finding the silver lining in bad supply numbers, and Advantage Property Services’ H1 2026 outlook is a clean example: it frames Phnom Penh’s retail oversupply as “one of the most attractive leasing environments in years” [S-855]. That’s true, and it’s also only half the story. The other half is that Phnom Penh now has roughly 981,000 square metres of completed retail space, occupancy has fallen to 57 percent from 67 percent a few years ago, and at current absorption rates it would take about 18 years to fill the space sitting empty today [S-855]. Whether that’s an opportunity or a warning depends entirely on which side of the lease — or the mortgage — you’re standing on.

The numbers APS reported

From the H1 2026 outlook [S-855]:

• Total completed retail stock: approximately 981,000 square metres across Phnom Penh.
• Occupancy: down to 57% in H1 2026, from 67% during 2018–2021 — a ten-point decline in citywide occupancy.
• Vacant space: roughly 450,000 square metres currently unlet.
• Absorption timeline: APS projects it would take approximately 18 years to fully absorb the existing vacant inventory at current leasing activity, if supply and demand trends continue unchanged.
• Rents, H2 2025 to H1 2026: asking rents fell 4.2% for community malls and 2.4% for retail podiums; shopping-mall rents were broadly stable; prime high-street locations rose 3.1%, the one segment moving against the grain.

The headline driver is straightforward and one this site has already documented on the residential side: years of retail development outpacing tenant demand, adding shopping centres, podiums, and community malls faster than the market could absorb them.

Read the format breakdown before the average

The citywide 57 percent occupancy figure hides more than it reveals, because the pain isn’t evenly spread. Community malls and retail podiums — the lower-tier, often borey- or condo-attached retail format that has proliferated across the city’s newer residential belts — are the segments losing rent and, by implication, tenants. Shopping malls (the larger, anchor-tenant format) are holding rents flat, which in a shrinking-occupancy market is itself a sign of relative resilience, not strength. Prime high-street frontage is the only segment actually gaining pricing power, up 3.1 percent, because well-located street-front retail is functionally a different, much scarcer asset than a podium unit inside a half-empty complex. If you’re holding or evaluating retail property in Phnom Penh, which of these three buckets an asset sits in matters more than the citywide average — a podium unit and a prime BKK1 high-street shopfront are not the same investment, even though both get filed under “retail.”

This format split rhymes with what this site has already found on the residential side. Our counter-cyclical developer piece found that the developers still building through Cambodia’s downturn are concentrated in affordable and mid-market housing, not luxury — demand has narrowed to the segments with genuine end-user need, and supply is following it there. Retail is showing the same pattern from the other direction: demand is holding for scarce, well-located space and draining from generic, commodity retail boxes attached to residential developments.

The 18-year number deserves to be taken literally, and skeptically, at once

Eighteen years is a mechanical extrapolation — APS is holding current leasing velocity constant and dividing it into current vacancy, not forecasting that nothing will change for two decades. Leasing pace, new supply, and demand all move, and a number like this is best read as “the current trajectory is badly out of balance,” not as a literal countdown. But it’s also not a number to wave away. Even if the real absorption period turns out to be a third of that, or half, it still describes a market where new retail supply has structurally outrun demand for years, not quarters — closer in shape to Phnom Penh’s now-familiar condo oversupply story than to a normal cyclical dip. Our mid-2026 market outlook already documented condo stock roughly tripling from about 20,000 units in 2019 toward 85,000 expected by 2027, with citywide occupancy in the high-70s to low-80s percent even years into the cycle. Retail’s 57 percent occupancy and multi-year absorption math is the same structural story — oversupply built faster than end-demand — playing out in a different asset class.

Two audiences this cuts in opposite directions

APS’s framing — “attractive opportunities for tenants and investors” — is accurate for exactly one side of this market and incomplete for the other:

• If you’re a tenant, operator, or retailer evaluating Phnom Penh expansion, this is genuinely a buyer’s market. More available space, more landlord competition, and real negotiating leverage on rent, lease terms, and location quality than at almost any point in the recent cycle. A brand that couldn’t previously afford a good location can likely afford one now.
• If you’re a landlord, developer, or considering buying an existing retail asset as an income investment, the same data reads as a warning. Falling occupancy and falling rents in the community-mall and podium segments mean cash flow assumptions underwritten during the building boom are less and less defensible, and the pressure Phnom Penh’s distressed asset landscape has already documented in other property segments is a live risk in commodity-grade retail too. A vacant unit in a half-full community mall is a genuinely different risk profile than a shopfront in a scarce high-street location, even if both show up as “retail” in a portfolio.
• If you’re evaluating a borey or satellite-city development with retail components — the ground-floor shops and small malls that anchor many of the projects in this site’s satellite borey belt guide — treat the retail component’s income projection with real skepticism. The residential units may sell on genuine end-user housing demand; the attached retail is competing in a market where the median format is losing rent.

The retail-podium-as-investment-product problem

One specific product deserves its own callout, because it’s the point where this data stops being abstract market commentary and starts being a decision readers of this site actually face: the ground-floor or podium retail unit sold as a standalone investment inside a condo or mixed-use development, often marketed with a fixed rental-return guarantee for the first two or three years. This is a genuinely common product in Phnom Penh, and it sits squarely inside the segment APS shows losing rent — retail podiums, down 2.4 percent in the year to H1 2026 [S-855]. A guaranteed-return structure can paper over that reality for exactly as long as the guarantee period lasts. Our guaranteed rental returns piece already sets out why these guarantees are usually funded out of the purchase price rather than real operating income — this data adds a second, independent reason for skepticism: even before asking whether the guarantee is real, ask whether the unit sits in a format (podium, community mall) that is currently losing rent citywide. A guarantee that expires into a segment with falling asking rents and years of vacant competing supply is a materially worse bet than the same guarantee attached to a scarce prime-frontage unit. Concretely, before buying a retail-podium or mall-unit investment product:

• Ask what format the unit is — podium, community mall, shopping mall, or street-front — and weight your rent expectations by that segment’s current trend, not the citywide average.
• Get the building’s actual current occupancy and comparable achieved rents, not the developer’s projected or asking figures, before the guarantee period ends.
• Model the post-guarantee income at the segment’s current trend rent, not the guaranteed rent, to see what the unit is actually worth once the subsidy runs out.
• Check how much competing retail GLA is under construction nearby — more supply arriving into an already-oversupplied format compounds the problem rather than resolving it.

A companion office story, not yet fully public

APS’s own H1 2026 report title groups office in with retail and condo as facing “years of oversupply” [S-855], though the figures released to press focused on retail specifically. The office market’s own trajectory, from the firm’s most recent fully detailed release, is worth noting as directional context rather than as an H1 2026-dated figure: APS’s Q1 2026 outlook put total Phnom Penh office stock on track to reach roughly 1.3 million square metres in 2026, with more than 144,000 square metres of new supply expected this year, even as Grade A buildings continued to outperform older, non-core-district stock on occupancy [S-856]. That’s the same pattern as retail — new supply concentrated in commodity-grade stock, pricing power concentrated in the scarce, well-located, well-managed segment — and it’s consistent with treating “years of oversupply” as a citywide description that masks a widening quality split rather than a uniform decline. This site will cover the office market’s H1 2026 detail in full once APS’s specific occupancy and rent figures for the period are public; for now, treat the office headline as directionally aligned with retail’s story, not as independently verified at the same level of detail.

What this doesn’t tell you

APS’s own framing is worth taking at face value on one point: this isn’t a story about Cambodia’s retail sector lacking long-term potential, and prime high-street rent growth is real evidence that well-positioned, well-curated retail can still command pricing power even inside an oversupplied market. The failure mode here isn’t “retail is dying in Phnom Penh” — it’s “undifferentiated retail floor space, added faster than the city could use it, is now competing hard for a limited pool of tenants.” That’s a supply discipline problem, not a demand collapse, and it’s a familiar one to anyone who has already read this site’s coverage of the condo cycle. The lesson transfers directly: match the specific asset and format to specific, verifiable demand before you underwrite it, and don’t let a citywide average — in either direction — do the work that format- and location-specific diligence should be doing.

The takeaway

Phnom Penh’s retail sector has too much space and not enough tenants to fill it at anything like the pace it was built — 57 percent occupancy, roughly 450,000 vacant square metres, and a multi-year absorption overhang that APS itself puts at close to two decades on current trends. That is a genuine opportunity for tenants and operators with the patience to negotiate well in a soft market, and a genuine warning for anyone holding or underwriting commodity-grade retail income property, particularly the community-mall and podium formats absorbing the brunt of the rent declines. Prime high-street locations are the exception that proves the rule: scarcity still commands a premium even in a glutted market. Judge any specific retail asset by its format and location, not by the citywide number — the two stories inside this data point in opposite directions, and only one of them is your story depending on which side of the lease you’re on. None of this is investment advice; verify current occupancy, rent comparables, and tenant covenant strength for any specific asset before committing capital.

Sources

• [S-855] Cambodia Investment Review — APS H1 2026 Outlook: Phnom Penh Office, Retail and Condo Markets Face Years of Oversupply, Creating Attractive Opportunities for Tenants and Investors — total retail stock (~981,000sqm), occupancy decline (67% 2018-2021 to 57% H1 2026), ~450,000sqm vacant, ~18-year absorption estimate, rent changes by format (community malls -4.2%, retail podiums -2.4%, shopping malls stable, prime high street +3.1%).
• [S-856] Cambodia Investment Review — APS Phnom Penh Q1 Real Estate Outlook 2026: Uneven Recovery as Office Supply Rises and Affordable Housing Leads Demand — Q1 2026 office stock trajectory (~1.3M sqm forecast for 2026, 144,000+ sqm new supply), Grade A vs. non-core occupancy split.