Cambodia's FX Reserves Fell 8.1% — But This Is Not the AMRO Liquidity Story

Two Cambodian banking-sector headlines landed within 48 hours of each other this week, both involving falling numbers, both from credible sources, and both easy to mentally merge into a single “Cambodia’s dollar reserves are shrinking” story. They are not the same story. Understanding why is worth the ten minutes it takes, because the two figures point in genuinely different directions once you separate them.

Two Different Numbers, Reported Two Days Apart

On 28 July 2026, AMRO — the ASEAN+3 Macroeconomic Research Office — reported that Cambodian commercial banks’ immediately accessible US dollar liquidity held at the National Bank of Cambodia had fallen to 10.4% of total USD deposits, the lowest level on record, as banks redirected roughly $10.5 billion into overseas assets during 2024 and 2025. This platform covered that finding in detail, including its direct read-through for property and construction financing.

On 30 July, the NBC’s own Semi-Annual Report disclosed a different figure: Cambodia’s international reserves — the central bank’s own foreign exchange holdings, not commercial banks’ deposits at the central bank — stood at $25.3 billion at the end of H1 2026, down 8.1% from $27.5 billion at the end of 2025. The NBC attributed the decline to lower deposits, mandatory reserves, and guarantees held by banking and financial institutions, plus gold price movements, and explicitly framed the decline as manageable: reserves still cover roughly eight months of imports, comfortably above the three-month benchmark economists generally treat as adequate for a developing economy.

Both figures fell. Both involve the NBC. Both were published in the same week. It would be easy — and wrong — to read them as two data points confirming the same crisis.

Why These Are Not the Same Metric

AMRO’s figure measures a ratio: how much of commercial banks’ own USD deposits they choose to keep as immediately accessible liquidity at the central bank, versus how much they deploy elsewhere, including overseas. A bank moving capital offshore in search of better returns lowers this ratio without touching the country’s official reserves at all — it is a statement about private bank balance-sheet allocation, not about the state’s own foreign-currency holdings.

The NBC’s figure measures a stock: the central bank’s own pool of foreign exchange assets, managed under its own investment policy, used to defend currency stability, finance imports in a crisis, and backstop the financial system in aggregate. This stock can decline for reasons entirely disconnected from private bank behaviour — gold price movements, valuation effects, or the central bank’s own portfolio rebalancing.

Here is the connection the NBC’s own report makes explicit, and the one worth sitting with: it attributes the reserve decline partly to lower deposits, mandatory reserves and guarantees held by banking and financial institutions — meaning some of the same underlying phenomenon AMRO flagged (banks holding less at the NBC) shows up on both sides of the ledger. Banks choosing to keep less liquidity parked at the central bank simultaneously lowers AMRO’s liquidity-coverage ratio and contributes to a lower reserve balance at the NBC, because mandatory reserves and deposits held by banks are a component of what the central bank counts as reserves in the first place. They are two views of a related shift, not two independent problems — but they answer different questions, and only one of them (AMRO’s) is genuinely at a record-low, historically unusual level. The reserve figure, even after the decline, sits within a normal, healthy range by international standards.

Reading the NBC’s Own Framing

The NBC’s Semi-Annual Report is, unsurprisingly, a self-reported document from the institution being assessed, and its framing is reassuring throughout: reserves “remain comfortably above international benchmarks,” the decline does not reflect “pressure on Cambodia’s external accounts or balance of payments,” and reserve levels “reinforce Cambodia’s ability to withstand potential external shocks.” None of that framing is unreasonable given the actual numbers — eight months of import cover is a genuinely strong position, more than double the standard adequacy threshold, and the report is transparent about naming the specific drivers of the decline rather than obscuring them.

It is still worth noting what this report does not address: it says nothing about the NPL trajectory, nothing about APD Bank’s withdrawal restrictions announced the same week, and nothing about bank-level liquidity distribution across smaller versus larger institutions — the specific concentration risk AMRO’s separate stress test flagged. A healthy aggregate reserve position at the country level coexists comfortably with real stress at individual institutions; the two are not mutually exclusive, and this report was never designed to speak to the second.

Reserves Follow Two Consecutive Years of Accumulation

Context matters for how alarming an 8.1% decline should sound. Reserves rose from $24.8 billion in H1 2025 to $27.5 billion by end-2025 — a substantial build-up over the preceding period — before this half’s pullback to $25.3 billion. Even after the decline, reserves remain above where they stood a year earlier. A single-period decline following several periods of accumulation reads differently than a sustained multi-period drawdown would, and the current data describes the former, not the latter.

The Green Bonds Detail

One detail buried in the report deserves a brief mention on its own terms: as of H1 2026, 6.3% of Cambodia’s international reserves were invested in green bonds and ESG-related instruments, which the NBC frames as aligning reserve management with evolving global investment practice. This is a small but genuine data point in the broader story this platform has tracked around Cambodia’s green-finance push — from LOLC’s solar financing to CMA-ARDB green lending partnerships — and it signals that ESG integration has reached even the central bank’s own reserve-management policy, not just commercial lending products. It has no direct property read-through, but it is a useful marker of institutional direction.

What This Means for Property, Precisely

The property-relevant conclusion here is narrower than either headline alone might suggest, and it is worth stating precisely rather than generically.

What the reserve figure supports: continued confidence in Cambodia’s currency stability and its ability to defend the dollarized system during external shocks — the structural backdrop that makes USD-denominated property pricing, financing, and long-horizon investment planning viable in Cambodia at all. A country with a genuinely depleting reserve position would be a much more serious macro risk to the entire property market’s pricing convention than anything in this report describes.

What the reserve figure does not address: bank-level lending capacity, developer financing risk, the deposit-insurance gap, or institution-specific stress like APD Bank’s this week. Those risks live at the commercial-bank level, and a healthy central-bank reserve position does not resolve them — it simply means the country-level backstop is not, on this evidence, the source of the current banking-sector strain. The strain, where it exists, is happening bank by bank, which is exactly why this platform’s guidance has consistently pointed toward diligencing your specific counterparty bank rather than taking system-level reassurance as a substitute.

Why Reserve Adequacy Matters More in a Dollarized System

Reserve adequacy is a more load-bearing concept in Cambodia than it would be in a country with its own freely floating, independently managed currency. This platform’s coverage of Cambodia’s dollarization and what it means for property buyers has established the core mechanic: the NBC cannot create US dollars the way a conventional central bank can create its domestic currency, which means its ability to act as a lender of last resort during a genuine dollar-liquidity crisis is structurally limited. In that context, the central bank’s own stock of foreign exchange reserves is not just a macro-stability indicator — it is one of the few tools the NBC actually has to intervene directly in a dollar-denominated financial system if conditions deteriorate sharply.

This is precisely why the distinction this article draws matters practically, not just semantically. If Cambodia’s official reserves were genuinely depleting toward the three-month adequacy floor, that would represent an erosion of the NBC’s own crisis-response capacity — a meaningfully different and more serious problem than commercial banks individually choosing to hold less liquidity at the NBC in pursuit of better returns. The current data describes the second situation, not the first. Reserves at eight months of import cover, even after this half’s decline, still leave the NBC with a substantial buffer relative to the adequacy floor most economists apply to developing economies — roughly two-and-a-half times the standard benchmark.

The Risk of Over-Reading Either Direction

There are two equally mistaken ways to read this week’s cluster of banking-sector news, and both are worth naming explicitly. The first is to treat the AMRO liquidity warning, the reserve decline, and APD Bank’s restrictions as three independent confirmations of a single unfolding crisis — a reading that overstates the connection between genuinely distinct metrics and risks generating alarm disproportionate to what any single data point actually supports. The second, equally mistaken reading is to let the NBC’s reassuring reserve framing paper over the separate, real concerns raised by AMRO’s stress test and APD Bank’s institution-specific stress this same week — treating “reserves are fine” as though it answers questions about bank-level liquidity distribution or individual-institution solvency, which it does not.

The accurate position sits between those two readings: Cambodia’s aggregate external financial position, measured by official reserves, is genuinely sound and not the source of this week’s other banking-sector news. But that soundness coexists with real, separately documented stress at the bank level — smaller institutions with thinner buffers, rising property-sector NPLs, and at least one bank now restricting customer withdrawals. Both things are true at once, and treating either fact as resolving the other is the error to avoid.

What to Watch

• Whether the H2 2026 reserve figure continues declining or stabilises — a second consecutive half-year drop would be a more meaningful trend than this single-period decline, and would warrant revisiting the “not a BOP problem” framing.
• Whether AMRO’s bank-level liquidity ratio and the NBC’s reserve figure continue moving together, confirming they share a common driver, or diverge — divergence would suggest the reserve decline has a second, independent cause worth investigating.
• APD Bank and any similar institution-specific developments, since those remain the more property-relevant, more immediate risk signal than the aggregate reserve position this report describes.

Sources

• [S-900] Cambodia Investment Review — Cambodia’s Foreign Exchange Reserves Fall 8.1% in H1 2026, But Remain Strong Enough to Cover Eight Months of Imports (30 July 2026) — the NBC’s Semi-Annual Report reserve figures, the stated decline drivers, and the green-bond reserve allocation.
• [S-901] Research Cambodia — AMRO Warns Cambodia Banks’ Dollar Liquidity at Record Low — What It Means for Property Financing — the commercial-bank liquidity metric this article distinguishes from the reserve figure above.
• [S-902] Research Cambodia — APD Bank Restricts Withdrawals to Appointments and $20,000 — the same-week institution-specific stress signal that aggregate reserve health does not resolve.