On 2 September, De Nederlandsche Bank said it had shifted roughly 86 tonnes of gold from New York and Ottawa towards London. The Netherlands still owns the same 612.4 tonnes. What changed was where some of it sits. The move was not presented as a bet on the gold price or a rejection of North America. It was an exercise in crisis preparedness: London offered greater tradability, while a more balanced geographic distribution reduced concentration.
Amir Naser Hojati is a Chicago Mercantile Exchange futures trader and a fintech entrepreneur developing automated trading systems, who writes on market structure, financial risk and the impact of public policy on markets. The views that follow are his own.
That sounds like routine reserve management, but it raises a larger question. Europe has spent years examining strategic dependencies in energy, payments and critical infrastructure. Gold deserves the same scrutiny: not simply where it is stored, but whether it remains accessible and usable when normal systems stop working.
Safety is more than ownership
Gold appeals to central banks partly because it has no corporate issuer and carries no conventional credit risk. A gold bar cannot default because somebody missed an interest payment. But freedom from credit risk does not mean freedom from political, legal or operational risk.
Belgium offers an unusually revealing example. The National Bank of Belgium reported 227.4 tonnes of gold at the end of 2024, then worth 18.36 billion euros. Most was held at the Bank of England, with smaller amounts at the Bank for International Settlements and the Bank of Canada, and only a very small quantity at home.
There is nothing inherently unsafe about London. It is one of the world’s deepest physical gold markets and an established centre for official reserve custody. The problem is not foreign custody itself. The problem is that the meaning of safe changes with the crisis.
Belgium learned that lesson in 1939. After placing much of its gold abroad, it entrusted a substantial share of the remainder to the Banque de France. When Germany invaded, the gold was moved to Dakar in French West Africa. Then France fell. The political architecture around the supposedly safer location changed, and French authorities ultimately delivered the remaining Belgian gold to the German Reichsbank. In autumn 1944, the Bank of France transferred 198.4 tonnes of fine gold to the National Bank of Belgium under an agreement between the two institutions.
The gold had not failed. The vault had not failed. The political chain governing access had failed.
Czechoslovakia faced a related problem in 1939, when the BIS executed an instruction transferring gold at the Bank of England to the German Reichsbank after an order later understood to have been issued under duress.
Yet history also provides the opposite lesson. Canada became a wartime sanctuary for European central-bank gold, and Norway’s decision to move reserves abroad helped preserve sovereign control; its remaining gold was evacuated in 1940. Foreign custody protected Norway even as changing political control exposed Belgium and Czechoslovakia.
The lesson is not to bring the gold home. Full repatriation can reduce jurisdictional risk while increasing domestic concentration and reducing immediate access to global liquidity. The real challenge is balancing sovereign control, liquidity and resilience.
Diversified by address, concentrated by failure mode
There is a deeper problem: risk management can itself become a source of hidden risk. A central bank may divide reserves among London, New York and Ottawa and conclude that it has diversified. Yet those locations may still depend on overlapping political alliances, legal assumptions, communications networks, financial infrastructure and settlement systems. The reserve is diversified by address while remaining concentrated by failure mode.
Financial markets have seen this before. Long-Term Capital Management demonstrated the danger of hidden correlation in 1998: positions spread across markets and countries did not remain independent when stress intensified, and diversification weakened precisely when it was most needed. Diversification built on yesterday’s correlations can become tomorrow’s concentration.
The analogy matters because London, New York and Ottawa are geographically separate but could become more correlated in an extreme transatlantic shock involving war, sanctions escalation, cyber disruption, capital controls or settlement failure. Different vaults do not necessarily mean independent risks.
The Swiss National Bank explicitly incorporates country risk, including the possibility that a state could restrict access to assets held there. Switzerland keeps roughly 70 per cent of its gold at home, around 20 per cent in London and about 10 per cent in Canada. No allocation is optimal for every crisis. The better objective is to build a reserve system that remains functional when reasonable assumptions prove wrong, the practical value of Nassim Nicholas Taleb’s emphasis on robustness and redundancy under uncertainty.
Gold can be yours and still be unusable
Europe’s sanctions architecture offers a modern reminder that ownership and access are different concepts. The EU has immobilised around 210 billion euros in Central Bank of Russia assets. At the end of June 2026, Euroclear Bank said 202 billion euros of its 241 billion euro balance sheet related to sanctioned Russian assets.
Russia’s circumstances are plainly different from those of Belgium, the Netherlands or another EU member state. The narrower lesson is institutional: an asset can remain legally yours while the jurisdiction hosting the relevant financial infrastructure determines whether you can move or use it.
Reserve safety should therefore be judged through three tests. Asset safety: does the reserve retain value? Access safety: can the sovereign retain effective control over it? And system safety: can the asset be converted into usable purchasing power when markets, transport, communications or payment systems are under stress?
Gold can pass the first test while failing the other two. Bullion may remain physically intact in a foreign vault, with legal title undisputed, yet still become temporarily ineffective if settlement fails, transport is interrupted or legal restrictions prevent mobilisation. A July 2026 IMF note recommends assessing gold’s effective liquidity using risk-based haircuts rather than headline market value alone. A reserve that survives a crisis but cannot be used during it has failed one of its central purposes.
Stress the reserve, not the forecast
Europe therefore needs more than geographic diversification. It needs functional redundancy. A resilient architecture could combine a sovereign-control tranche of physical gold at home with a liquidity tranche in a major bullion centre such as London, alongside holdings under genuinely different legal and operational arrangements.
Resilience should also extend beyond bullion. Foreign currencies, highly liquid securities, central-bank swap arrangements, alternative settlement channels and tested emergency procedures provide other routes to liquidity. The aim is not to eliminate risk. It is to ensure that several supposedly independent safeguards do not fail for the same underlying reason.
That principle should become part of a European reserve-geography stress test. Central banks already stress-test banks, liquidity and financial institutions. Reserve management should apply the same discipline to custody, access and infrastructure. Instead of asking only what happens if one vault becomes unavailable, policymakers should test common-mode failures: what if London and North America are impaired together? What if ownership survives but settlement fails? What if the vault is accessible but the payment channel is not?
Reverse stress testing would go further. Start with the outcome policymakers most want to avoid, the reserve system becoming unusable when the country needs it most, and work backwards. Which failures produce that outcome? Which assets and settlement channels survive? Which supposedly independent safeguards fail together? This shifts reserve management from searching for a single safest location towards preserving options: domestic gold for sovereign optionality; London for liquidity; multiple jurisdictions for legal optionality; and foreign-exchange reserves, swap lines and alternative settlement channels for monetary and operational optionality.
Make liquidity less dependent on location
There is also a longer-term opportunity: some domestically held gold could remain under sovereign custody while being independently verified and eligible for temporary liquidity under pre-agreed standards. Geography would still matter for enforcement and risk, but it would no longer determine usability entirely. The broader aim is simple: reserve liquidity should have more than one route into the financial system.
Gold still has an address
The Netherlands has moved towards a more balanced geographic structure, with no single location now holding even one-third of its gold. That is sensible. But geography should be the beginning of reserve-risk analysis, not the end. Belgium, whose central bank’s 2024 corporate report states that most official gold was held in London, has an equally legitimate reason to ask whether its balance between liquidity, sovereign access and jurisdictional diversification remains optimal. The answer may well be yes. What matters is that the assumptions behind that answer are tested before a crisis rather than during one.
Reserve strength should be measured not only by market value but by whether gold can be accessed and converted into liquidity under stress. Diversification should be judged by the independence of custody, legal authority, settlement and liquidity channels, not by the number of jurisdictions on a map. Different addresses and different custodians do not necessarily mean independent systems. Gold may have no issuer. But it still has an address.
The post Gold Reserves: Diversified By Address, Concentrated By Risk appeared first on The Fintech Times.
