Why Central Banks Are Hoarding Gold — and What Stackers Should Take From It

There's a special kind of signal in watching the people who print money buy the one asset they can't print. Central banks purchased a net 863 tonnes of gold in 2025, per the World Gold Council — down from the extraordinary 1,000+ tonne years of 2022–2024, but still nearly double the 473-tonne average that prevailed from 2010 to 2021. Official holdings now stand around 36,000 tonnes — territory last seen when the world was on a gold standard.
Source: World Gold Council; 2022–24 shown at the 1,000 t floor; 2026 is JPMorgan projection
And in a milestone that would have sounded absurd a decade ago, gold has overtaken US Treasuries as the largest asset in global official reserves — roughly 27% of the total. The institutions that built the paper system are, quietly and without announcements, re-anchoring themselves to metal.
Who's buying
| Buyer | 2025 activity | Context |
|---|---|---|
| Poland | +102t | Largest buyer, second consecutive year |
| Brazil | +43t (Sep–Nov) | First purchases since 2021; gold still only ~7% of reserves |
| Turkey | +27t (to Oct) | Steady accumulation through volatility |
| Czech Republic | +20t | Marching toward a stated 100t target by 2028 |
| China | Reported monthly buying streaks | 18–19 consecutive months through mid-2026; reserves ~2,300t reported |
Source: World Gold Council reporting; China excluded (reserve figures are estimates)
Beyond the league table: 22 central banks added a tonne or more in 2025, and WGC surveying found 45% of central banks planning to increase gold reserves — the highest reading since the survey began tracking it — with roughly nine in ten expecting global official gold holdings to keep rising, and three-quarters expecting the dollar's reserve share to fall within five years. Q4 2025 alone saw 230 tonnes of net buying; Q1 2026 estimates (including unreported purchases) ran around 244 tonnes.

Why they're buying
- Sanctions-proofing. The freezing of Russian reserves in 2022 taught every treasury the same lesson: foreign-held bonds can be switched off; gold in your own vault cannot. Russia itself has since restricted gold exports and stopped publishing reserve data — holding, not selling.
- De-dollarization at the margin. China has paired multi-hundred-billion-dollar Treasury reductions with steady gold accumulation. Not an abandonment of the dollar — a rebalancing away from any single counterparty.
- Debasement insurance. The same logic as any stacker: global debt compounds, currencies are policy instruments, and gold is the reserve asset with no issuer. Central banks kept buying through record 2025 prices — 53 all-time highs that year — because tonnage targets, not price targets, drive the program.

The honest counterpoints
A fair reading includes the other side. The 2025 total was a 21% slowdown from 2024 — JPMorgan projects ~755 tonnes for 2026, partly because at $4,000+ gold, fewer tonnes achieve the same allocation targets. Reported purchases can lag or understate reality (Q1 2026's reported figure was a mere 16 tonnes against ~244 estimated), and the data has genuine holes — Russia claims mine output 50% above World Gold Council estimates with no explanation. And some banks sell under duress: Turkey unloaded 58 tonnes in two weeks during a currency squeeze. The trend is powerful; it is not a straight line, and it is not a price guarantee.
The demand floor argument. What central bank buying most plausibly provides isn't a rocket — it's a floor. A price-insensitive buyer absorbing 750–1,000 tonnes a year (a fifth or more of annual mine supply) changes the downside math for everyone else holding gold. When gold corrects, as it did through mid-2026, that bid keeps showing up. Meanwhile retail behaves the same way at street level: Chinese bank gold quotas — 600kg daily allocations — have sold out in minutes on dips.

What a stacker should actually take from this
Not "gold only goes up" — 2026's drawdown disproved that within months of the record highs. The real takeaway is about behavior: the most sophisticated, best-informed reserve managers on earth accumulate steadily, on programs, through both rallies and corrections, prioritizing tonnage held over entry price. That is, almost word for word, the stacker playbook — systems over timing, ounces over dollars. If you want to mirror the trade at household scale, gold's entry point is lower than most assume: our fractional gold guide covers starting with tenth-ounces, and the ratio guide covers balancing it against silver.
Frequently asked questions
Does central bank buying help silver too?
Only indirectly — central banks hold gold, not silver. But the monetary-distrust demand that drives official gold buying tends to spill into silver at the retail level, and silver's smaller market means the spillover hits harder (2025: silver +148% vs gold +54%).
Should I copy the central banks and buy only gold?
Their constraints aren't yours — they need billion-dollar liquidity in a single asset. A household stack gets divisibility and upside from silver that reserve managers can't use. Most stackers run both; see gold vs. silver.
What would make this trend reverse?
Genuine fiscal consolidation in reserve currencies, durable high real rates, and a de-escalation of sanctions risk — the exact conditions the last decade has moved away from. Watch the WGC's annual central bank survey; it's the closest thing to the buyers announcing their intentions.
When nine out of ten central banks expect official gold holdings to rise, stackers don't need to predict anything — they just need to already own what the slowest, biggest money in the world keeps accumulating.














