Why Central Banks Keep Buying Gold: the Counterparty-Risk Case

The United States government carries 8,133 tonnes of gold on its official books at a statutory price of $42.22 per ounce, a valuation last touched, according to official records, at a date not independently confirmed in available research. As of mid-September 2026, the spot price sits above $4,300 per ounce.

That gap is not a rounding error. It is a structural signal, and understanding it is the starting point for understanding why central banks keep buying the metal.

Consider two facts that are both true at once. Central banks globally purchased 863 tonnes of gold in 2025, extending a decade-long accumulation run. Yet the country holding the world’s largest declared reserves has not conducted a formal, independent audit of those reserves since 1953.

Neither fact is fringe. Both come from official records, and both belong in the same analytical frame.

What follows here is not the point; the framework is. After reading, you will have a clear lens for what central bank gold buying actually signals, why the audit question matters well beyond conspiracy territory, and what these trends together imply for gold’s monetary role from here.

The sovereign debt backdrop that makes gold buying rational

Start with the number that anchors everything else. The US Treasury’s Debt to the Penny dataset recorded total public debt outstanding at roughly $40.11 trillion as of 15 September 2026.

That figure did not appear overnight. Its origins trace to 1971, when the United States severed the dollar’s final link to gold, the so-called Nixon Shock, ending the gold exchange standard.

The sovereign debt vulnerabilities now embedded across major economies did not emerge from a single policy failure; they accumulated across five decades of monetary expansion, and the feedback loop between rising yields, higher debt service costs, and reduced fiscal flexibility is precisely what makes gold’s no-counterparty property increasingly significant to reserve managers.

What followed was five decades of floating fiat currency, expansive monetary policy, and the tools that condition produced: quantitative easing, and interest rates pushed to zero and, in some economies, below it.

The structural indicators worth holding in view are these:

• Total public debt of approximately $40.11 trillion as of 15 September 2026
• A Treasury buyback programme reportedly at least doubled in scale to help contain rising yields
• Five decades of fiat monetary expansion since the 1971 departure from gold convertibility

The buyback expansion is the tell. A programme scaled up to suppress yields is a containment mechanism, not a marker of a system operating comfortably within its means.

Some analysts reach for a historical parallel to frame the trajectory.

Analyst commentary compares the gradual debasement of modern fiat currency to the slow dilution of silver content in Roman coinage in the centuries before imperial collapse. This is an interpretive framing, not an established economic equivalence.

Here is what this debt level changes for a reserve manager. When a sovereign borrower’s obligations climb to this scale, the risk calculus on holding that sovereign’s bonds shifts. Gold carries no default risk, and in a $40 trillion debt environment, that property stops being theoretical and becomes operationally significant.

That is the condition. It reframes central bank gold accumulation as a rational institutional response to a documented structural situation, not a speculative or ideological bet.

What 863 tonnes a year actually tells you about institutional confidence

For roughly a decade, central banks have absorbed about 1,000 tonnes of gold a year. The consistency is the story. This is not a one-off crisis reflex; it is a sustained pattern of official-sector demand.

The 2025 figure, per the World Gold Council’s Gold Demand Trends Full Year 2025 report published 29 January 2026, was 863 tonnes in net purchases. That is a 21% drop from the 1,092.4 tonnes bought in 2024, yet it remains far above pre-2022 norms.

The inflection point sits in 2022.

In 2022, roughly $300 billion in Russian central bank reserves were frozen. For every other reserve manager watching, that event reframed a core assumption: sovereign assets held in another country’s system carry counterparty risk that can be activated overnight.

The survey data sharpens the read. Invesco found that 60% of respondents in 2023 cited asset-freeze concerns as a primary catalyst, and in 2025, 39 out of 46 responding central banks named the Russian sanctions event as a driver.

That tells you something important about what this trend actually is. Central bank gold buying is not primarily an inflation-hedge trade. It is a counterparty-risk trade, and that distinction determines whether the trend is durable.

Reserve diversification away from Treasuries accelerated sharply after 2022, and the data show the shift is not evenly distributed: emerging-market central banks have driven the bulk of net purchases, reflecting a different calculus on dollar-system exposure than developed-market peers typically apply.

Two ways to read the same data

The same numbers support two legitimate interpretations.

The mainstream institutions, the IMF, the Federal Reserve, and the European Central Bank, read this as modest diversification within a stable fiat framework. The IMF explicitly recommends treating gold as a high-risk diversifying reserve asset, not a new monetary anchor. On this view, the buying is proactive risk management, nothing more.

A heterodox camp reads the same figures as institutional preparation for a more fundamental monetary transition, an early positioning for a world where fiat reserves lose primacy.

You do not need to resolve this tension to use it. If the driver is structural geopolitical fragmentation, the trend persists. If it is a temporary reaction to a normalising world, it reverses. Which camp is correct is precisely what the volume data cannot yet tell you, and holding both possibilities is the honest analytical position.

Why gold has no default risk (and why that concept matters more now than it did in 2010)

Start with the plain definition. A government bond is a promise: someone owes you money and agrees to pay it back. A bank deposit is the same, a claim on a counterparty who could default, freeze, or restructure.

Gold is different. It is not a claim on anyone. It is the asset itself, which means there is no counterparty who can fail to pay you.

Why does this matter more now than it did in 2010? Because the value of a no-counterparty asset rises mechanically as the default risk of the alternatives rises. In a high-debt, elevated-yield environment, sovereign bonds, the traditional reserve backbone, carry more perceived risk, and gold’s freedom from that risk becomes relatively more valuable.

Here is the risk profile, side by side:

• Default risk: Gold has none; sovereign bonds carry the issuer’s default risk
• Counterparty risk: Gold has none; deposits and bonds depend on a counterparty performing
• Inflation sensitivity: Gold tends to hold real value; nominal bonds erode in real terms during inflation
• Liquidity: Both are highly liquid, though bond liquidity depends on functioning markets and issuer confidence

The cost of misreading this property is not hypothetical. The UK Treasury sold roughly 395 tonnes of gold between July 1999 and March 2002, at an average price of around $275 per ounce.

The UK sold at approximately $275 per ounce. As of mid-September 2026, spot gold trades above $4,300 per ounce, more than fifteen times the sale price.

Sit with that calculation for a moment. It is a live illustration of the opportunity cost of underweighting a no-default-risk asset precisely as sovereign stress began to build.

The logic even extends to adjacent markets. The original source notes that some sovereigns are quietly accumulating physical silver, with Saudi Arabia reportedly purchasing shares in the SLV silver ETF, a signal that the counterparty-risk thinking is not confined to gold alone.

For you, the takeaway is structural. Gold exposure is not only a return trade; it is a decision tied directly to the creditworthiness of the sovereign bonds you might hold instead.

Physical versus paper gold exposure introduces a counterparty layer that many investors overlook: a gold ETF or unallocated account restores exactly the kind of third-party claim that physical gold eliminates, which matters if the counterparty-risk logic driving central bank buying is also the logic shaping your own allocation decision.

Fort Knox, the 1953 audit, and what the encumbrance question actually means

The last formal, independent audit of US gold reserves was conducted in 1953. A recent informal walkthrough by a US senator, whatever its optics, does not meet professional audit standards.

This matters because different forms of oversight confirm different things:

1. Physical walkthrough: Confirms that gold is visibly present in a room. It does not verify quantity, purity, or ownership status.
2. Financial-statement audit: Confirms the books are presented fairly under accounting standards. It does not assay the metal or disclose whether it is pledged elsewhere.
3. Physical assay with encumbrance disclosure: Confirms the metal’s quantity and purity, and reveals whether it has been leased, swapped, or hypothecated.

The gap between the second and third form is the entire question, and an accounting rule is what makes the gap possible.

Under IMF accounting rules dating to 1999, central banks may report leased or swapped gold as part of official reserves without separately disclosing that the metal is encumbered. That means a clean balance sheet cannot, on its own, confirm the gold is free of third-party claims.

Current oversight relies on financial-statement audits. Treasury Inspector General reports OIG-25-002, OIG-25-003, and OIG-26-004, covering fiscal years 2023 through 2025, confirm the custodial deep-storage gold schedules are presented fairly under US GAAP with no material weaknesses. That is a real assurance, but it is assurance about the accounting, not the encumbrance.

What transparency legislation would and would not resolve

The legislative response has arrived, though it has not become law. Representative Thomas Massie introduced H.R. 3795, the “Gold Reserve Transparency Act of 2025,” on 6 June 2025. Senator Mike Lee introduced a Senate companion on 19 November 2025. As of mid-September 2026, neither has been enacted and no physical re-assay has begun.

The bills mandate a full physical assay and inventory, recurring every five years, plus disclosure of all gold transactions over the past 50 years.

The Gold Reserve Transparency Act introduced by Representative Massie reflects a broader legislative recognition that financial-statement audits and physical assays answer fundamentally different questions, a distinction that reserve managers and commodity investors increasingly apply when assessing the credibility of declared sovereign holdings.

The physical assay is not the part that answers the encumbrance question. The 50-year transactional disclosure is. Only a record of leases, swaps, and hypothecation would reveal whether the metal, even if physically present, is already owed to someone else.

And there is a separate usability issue. A significant portion of the stored US gold is reportedly in coin melt form from the 1930s, which does not meet modern good delivery standards. Confirming the metal is present is not the same as confirming it is deliverable.

For a gold or energy investor, this is a genuine market variable. Full independent verification would either confirm the reserve or reveal partial encumbrance, and either outcome would reprice the metal.

The revaluation arithmetic and why the numbers do not add up the way gold bugs suggest

Give the transparency advocates everything they ask for, and then do the arithmetic. The result is the honest answer, and it cuts against the popular thesis.

The US holds 8,133 tonnes, currently carried at the statutory $42.2222 per ounce. Proposals to revalue the metal for a monetary reset range from $10,000 to $20,000 per ounce.

Even at $20,000 per ounce, the reserve produces roughly $1.2 trillion against a $40.11 trillion debt load. Gold cannot solve the problem. The math simply does not reach.

The structural conditions the arithmetic exposes are these:

• The debt-to-gold ratio is so wide that no plausible revaluation closes it
• Backing the currency with gold would require dramatic revaluation relative to the existing money stock
• That revaluation would risk severe deflation or volatile supply constraints, a point mainstream economists have made via the Cato Institute

Here is the part the gold-backed-dollar thesis misses, and it is the more interesting part. The gap between $1.2 trillion and $40.11 trillion is not an argument against holding gold.

Gold’s price appreciation in this environment is driven by the absence of a clean solution, not the presence of one. With spot gold ranging between roughly $4,326 and $4,428 per ounce through September 2026, that dynamic is already visible in the price.

Analysts add a demand-side pressure point: a shift of just 1% to 3% of the global population into precious metals would be enough to move prices dramatically. If you have heard the gold-backed dollar pitch, understand its arithmetic ceiling before you build a position around it. But understand too that the same conditions feeding that pitch are structurally supportive of the metal, regardless of whether the reset ever happens.

What the evidence actually supports, and what it leaves open

Three evidence streams converge here. The sovereign debt trajectory is documented and escalating. The central bank accumulation motivation, particularly post-2022 sanctions logic, is a counterparty-risk response rather than a simple inflation trade. And the audit and encumbrance questions remain genuinely unresolved.

The honest synthesis is that two things are true at once. The mainstream view, that this is diversification, holds. And the structural view, that the conditions making diversification rational are themselves intensifying, also holds. You do not have to collapse them into a single verdict.

Two variables would actually move the analytical needle:

• A credible independent US audit with encumbrance disclosure. This would either confirm the reserve as free and available or reveal partial encumbrance. Either outcome reprices gold, which is why the stalled status of H.R. 3795 and its Senate companion matters.
• A formal announcement of China’s true gold holdings. Analyst Alasdair Macleod estimates China’s actual reserves may exceed 20,000 tonnes, far above the officially declared figure. A public reconciliation would reset global pricing expectations.

Until one of those events lands, the demand structure supporting gold rests on conditions that are not resolving: elevated sovereign debt, geopolitical fragmentation evidenced by 39 of 46 central banks citing sanctions, and persistent reserve opacity. With spot gold between roughly $4,253 and $4,428 per ounce across September 2026 reference dates, that structural bid is priced in, not speculative.

For a mining and energy investor, that is the point. The institutional logic driving central bank purchases is the same logic that shapes the valuation context for gold mining equities. Read the price movements against that backdrop, and separate the structurally grounded narratives from the speculative ones.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions.

Past performance does not guarantee future results. Financial projections and revaluation scenarios discussed here are speculative, subject to market conditions and various risk factors, and may change based on policy and market developments.