Correction, October 31, 2026: earlier versions said central banks bought 1,237 tonnes ("1,200 tonnes") in 2025; the World Gold Council figure is about 863 tonnes, and we also removed an out-of-date claim that China had stopped reporting purchases.

Key Takeaways

  • Central banks bought about 863 tonnes of gold in 2025, down from over 1,000 tonnes in 2024.
  • The World Gold Council forecasts about 850 tonnes for 2026, so demand is steady rather than surging.
  • Reported net buying was 170 tonnes through August. August alone was 39 tonnes, and July 23 tonnes.
  • Poland (98t) and China (80t) lead 2026 buying. Turkey (−82t) and Russia (−56t) are net sellers.
  • Buying did not prevent a fall of about 31% from January's record. It is a demand flow, not a level that protects the price.
~863t Bought in 2025
170t Reported YTD through Aug
~850t WGC 2026 forecast
~31% Gold below record, Oct 30

Central banks are still net buyers of gold, at a slower pace than the over-1,000-tonne years. That buying continued through a sharp decline: gold was about $3,880 as of October 30, 2026, about 31% below its January record. So official demand is real and durable, but it is not a reason to expect the price to hold at any particular level.

The Numbers: 2024 to 2026

PeriodNet purchasesNote
2024Over 1,000tSources range from about 1,045t to 1,092t
2025About 863tWorld Gold Council
Q2 2026289tIncludes estimated unreported buying; a record Q2
July 202623tReported net
August 202639tReported net
January–August 2026170tReported only
2026 forecastAbout 850tWorld Gold Council

The figures above measure two different things. The 170 tonnes counts purchases that central banks have reported, while the quarterly 289 tonnes also includes the World Gold Council's estimate of buying that is not reported. That is how a single quarter can exceed the year-to-date reported total, and why the 850-tonne forecast is much larger than the reported number so far. Do not compare the two directly.

Who Is Buying and Selling in 2026

Country2026 through AugustAugustNote
Poland+98t+8tHoldings 648t, against a 700t target
China+80t+20t22nd straight month of buying; about 2,387t, 9% of reserves
UzbekistanAbout +50t+8t
KazakhstanAbout +36t+7t
Czech Republic+14tNot listed
Turkey−82t+3tNet seller for the year despite August's purchase
Russia−56t−6tNet seller

Two points stand out. Poland is about 52 tonnes from its target, so that buyer has a visible end point unless the target is raised. And the sellers are large: Turkey and Russia together sold 138 tonnes, which is equal to most of the 178 tonnes bought by Poland and China. Central banks are not a one-way flow.

Why Buying Did Not Stop the Fall

This page used to describe sovereign demand as a support zone near $4,500. That was wrong. Gold ended June near $4,060 and was about $3,880 as of October 30, 2026, while central banks kept buying.

The reason is scale and marginal pricing. Central-bank demand is one flow among many, spread over months. Gold's price at any moment is set by the marginal buyer and seller, and in 2026 that has been investors reacting to higher yields, a firm dollar and a possible Fed hike. On October 30, forced selling after a tech-stock rout pushed gold down alongside equities. A steady monthly purchase cannot absorb that kind of move.

The accurate claim is narrower. Central-bank buying adds a long-term layer of demand, and may affect how the market behaves over years. It does not set a minimum price, and it does not make deep corrections unlikely. The 2026 decline is the evidence.

Why Central Banks Buy Gold

  • Sanctions risk. The freezing of Russian reserves in 2022 showed that foreign-held reserves can be blocked. Gold held at home cannot be frozen in the same way.
  • Reserve diversification. Gold lets reserve managers reduce dollar exposure without concentrating in another currency.
  • Allocation policy. Some buyers work to stated targets, such as Poland's 700 tonnes.
  • Domestic supply. Producers such as Uzbekistan and Kazakhstan can buy from local mines.

These motives are long term, which is why buying has continued through the decline. They also do not rule out selling. Turkey and Russia show that central banks can sell gold when they have other needs.

What to Watch

  • The WGC's Q3 2026 Gold Demand Trends, expected in early November and not published as of this review. It will show how much estimated buying continued as prices fell.
  • Whether 2026 reaches about 850 tonnes. That would match 2025 and confirm a stable, lower pace.
  • China's streak. August was the 22nd straight month of buying.
  • Turkey and Russia. Further sales would offset more of the buying.
  • Poland's 700-tonne target. Watch whether it is met, raised, or left in place.

What This Means for Investors

  1. Treat central-bank demand as a long-term factor. It says little about where gold goes over the next few months.
  2. Separate reported and estimated figures. Headlines mix them, and the gap is large.
  3. Check rates first. In 2026, the Fed, yields and the dollar have mattered more for price than official buying. See gold vs the dollar.

Sources: World Gold Council, USAGOLD, TradingKey.

Disclaimer: This analysis uses publicly available World Gold Council data and is for informational purposes only. It does not constitute investment advice. Central bank data is reported with a lag, and estimates of unreported buying involve significant uncertainty. Market figures are as of October 30, 2026; central-bank data runs through August 2026 as published October 6, 2026. Consult a qualified financial advisor before making investment decisions.