During my university days, I worked part-time at a large foreign exchange shop in Washington, D.C. The city was full of diplomats and staff from the IMF and World Bank, so the counters never sat quiet. Beside me was the precious metals booth, run by a man about 15 years my senior. He had lived through Vietnam, then Nixon taking the US dollar off gold, then the Hunt brothers’ silver saga, then the US dollar’s swings around the Plaza and later the Louvre Accord.
Although G7 central banks worked to keep the reserve currency stable, he still didn’t trust the US dollar. So, he sold gold Krugerrands with conviction and told customers that gold bought you sleep. While I never bought any, his line stayed with me.
Decades later, that sentiment has re-emerged, only now through central banks themselves. Today, two signals frame the present. Recent World Gold Council (WGC) tallies show official sector buying near 1,000 tonnes a year. At the same time, IMF data show a slow drift in reserve composition, with the US dollar’s share edging down from about 70 per cent to the mid-50s. One speaks in tonnes, the other in mix. Together they show how central bank balance sheets change when the world feels less stable than the models implied.
The geographic centre of global reserves has visibly shifted. By IMF measures, more than half of global reserves now sit across Asia, the Middle East and emerging markets. The institutions holding most precautionary assets are also those most exposed to trade invoicing in US dollars and to funding squeezes where trust itself has become a sought-after asset. Not surprisingly, therefore, their priority is resilience.
China is the scale case. Total reserves sit around $US3,500 billion. About seven per cent is in gold, or about $US300 billion. Put that in perspective: China’s gold holdings alone are nearly three times Australia’s total reserves and would rank among the world’s total top 10. That’s scale.
So why build that position? Their motive isn’t carry. It’s portfolio resilience for a more uncertain world. Many emerging-market reserve managers have liabilities and trade flows anchored to the US dollar. Diversifying the asset side limits future currency mismatch and lowers political dependence. Geopolitical signalling and balance-sheet diversification often overlap. Gold helps because it is nobody’s debt to pay. It clears broadly, sits outside sanction channels, and is recognised across legal systems. Those traits travel well when regimes shift.
Better to read the central bank buying as institutional housekeeping, not a momentum trade. When the reserve currency still dominates but feels less permanent, central banks seek optionality. Such an adjustment quietly shows up in composition, not headlines. A few tonnes a month, another small allocation next quarter. It looks unexciting by design, which is exactly the point. The price moves fast. Reserve policy moves slow. The latter is my subject.
That same slow adaptation runs through the system’s plumbing. Payment experiments are advancing in local-currency settlement, forming an alternative to SWIFT, and in central-bank digital currency pilots. The goal is reduced friction, more routing choices and fewer single points of failure when correspondent chains get politicised. This is about keeping trade and finance moving when one corridor shuts, rather than forming a rival bloc.
Reserve portfolios mirror that instinct. More assets that travel across regimes, fewer that rely on one gate. Gold sits outside Basel’s liquidity framework, which classifies Treasuries as high-quality collateral. That exclusion helps explain why its bid is official, not financial.
Correlation risk
From a risk lens, bullion covers two exposures. First, access risk: sanctions, custody rules or settlement protocols can change overnight. Second, correlation risk: the same shock that hits risk assets can also tighten dollar funding. In that moment, an asset that is not anyone’s debt steadies the balance sheet by stepping out of the correlation chain.
Markets often treat this as a call on price, yet I highly doubt that. The question is more functional. Does the asset store purchasing power through cycles? Does it clear across borders without permission? Does it hold value when politics touches money? Gold is one of the few assets that can answer ‘yes’ to all three.
A brief aside on cryptocurrency. It too captures much of the scepticism around fiat currency, yet its liquidity leans on stable coins, which themselves are largely backed by US Treasury bills. Few central banks hold crypto in reserves. Gold already sits inside the system. That contrast between cryptocurrency and gold is enough.
Step back to the frame. The post-1980s order relied on coordination. G7 central banks would act together when exchange rates overshot. Trade lengthened supply chains on the assumption that policy would stay predictable. Those assumptions now feel weaker. Settlement systems keep testing new workarounds rather than rely on old models. In such an environment, safe assets are defined by what they do, not their passports. Dollar share remains structurally high, but the world is visibly testing alternatives.
None of this projects what tomorrow’s gold price will be. Real rates and growth obviously still matter. The point, however, is narrower. Official portfolios are now being rebuilt for function in this new era of disorder and political uncertainty. WGC tonnes on one side, evolving currency shares on the other. The scale helps anchor that story.
More than half of global reserves are now in Asia and the emerging world. Each step nudges the mix towards assets that travel, clear, and hold value when the script changes. I don’t know where gold is headed tomorrow, but I’m more certain that political disruption will keep rising. The only thing harder to predict than gold today is Trump.
As reserves diversify, auction mechanics inevitably change. A lighter official bid raises the bar for US Treasury issuance and makes ad-hoc US dollar stabilisation harder to organise. The result is a currency that trades more on domestic policy credibility and less on coordinated support; a costly adjustment for any nation accustomed to a G7-coordinated anchor.
I think back to the metals desk in Washington. Then, coordination bought stability. Today, increased diversification buys resilience. If you manage risk for a living, the lesson is simple enough. Design for the day when policy uncertainty moves faster than trust can follow. That’s what the central banks are signalling with their balance sheets. Gold buys you sleep; only now it’s central bankers who are saying this.