Central banks bought 61 tonnes of gold in September, WGC says
Central banks added a net 61 tonnes of gold in September, the highest monthly total of 2026, but the nine-month total of 228 tonnes is 7% below last year.
Why central banks buy, and why it is slow
Official-sector gold demand is unlike any other part of the market. A central bank is not trading a view on the next quarter; it is changing the composition of a national balance sheet, usually over years and usually without announcing a target. The motives are reserve diversification, a hedge against sanctions or frozen assets, and a desire to hold an asset that is nobody else's liability. Those motives do not respond to a 5% move in the price, which is why the WGC series is watched less for its monthly signal than for its direction.
What September showed
The 61 tonnes added in September was the largest monthly net purchase of 2026, and the buyer list was broad: Poland and China led, with Uzbekistan, Kazakhstan, Jordan, the Czech Republic and Bolivia also adding. Breadth matters more than the headline. A single buyer can be a one-off; seven buyers in a month suggests the reserve-management logic is still intact across emerging-market central banks.
| Measure | Figure |
|---|---|
| September net purchases | 61 tonnes, the 2026 monthly high |
| January to September 2026 | 228 tonnes |
| Against the same period of 2025 | Down 7% |
| Largest buyers | Poland, China |
But the nine-month total of 228 tonnes is 7% below the same period of 2025. That is the number to hold on to. Official demand is still positive, and still a structural bid under the market, but it is not accelerating. The surge in central-bank buying that followed the freezing of Russian reserves has settled into a slower, steadier pattern.
The best argument against
Official buying is a stock, not a flow that sets the price. The marginal buyer of gold is the futures, ETF and over-the-counter market, and those investors respond to real rates, the dollar and Fed expectations. A central bank adding 61 tonnes in a month is small against daily turnover, and the 7% year-on-year decline shows the official bid is fading, not building. If the Fed keeps tightening, as Standard Chartered expects, the opportunity cost of holding a non-yielding asset rises and official purchases will not prevent a lower price.
What happens next
October and November WGC data will show whether September was a peak or a pause. The more useful read is China's own monthly reserve disclosure, which converts the WGC estimate into an official number and shows whether Beijing is accelerating or merely maintaining its slow accumulation. A fourth quarter that closes the 7% gap would restore the structural-bid story; another month below 40 tonnes would suggest the official sector has already done most of the diversification it intends to do at these prices.
Source: World Gold Council data, reported by Guan Dian, 19 November 2026.
Original source: Guan Dian, citing World Gold Council · 19 Nov 2026
