Central banks are still net buyers of gold in 2026, but at a slower pace than in 2025. World Gold Council (WGC) data published in September 2026 shows about 130 tonnes of net central bank buying in the year to July 2026, against about 160 tonnes in the same period of 2025. Poland and China led the buying; Turkey and Russia were the biggest sellers.
Central bank gold buying in 2026: the headline numbers
The WGC tracks reported changes in official gold reserves each month. Its September 2026 update covers figures to the end of July 2026.
- Year to July 2026: about 130 tonnes net bought.
- Same period of 2025: about 160 tonnes net bought.
- Difference: 160 − 130 = 30 tonnes less, or about 19% lower (30 ÷ 160 ≈ 0.19).
So the buying continued, but it slowed. "Central banks are buying gold" is true. "Central banks are buying more than ever" is not what these figures show.
A related signal points the other way. A WGC survey in June 2026 found a record 45% of central banks plan to increase their gold holdings. Intentions and actual monthly purchases are different things, but the survey suggests the long-term interest has not faded.
Who is buying gold in 2026?
Here are the main buyers in the year to July 2026, with holdings where the WGC data gives them:
| Central bank | Net change, year to July 2026 | Holdings and share of reserves |
|---|---|---|
| Poland | +90 t | about 640 t, 28% of reserves; target 700 t |
| China (PBoC) | +60 t | 2,366 t, about 8% of reserves |
| Uzbekistan | +40 t | — |
| Kazakhstan | +29 t | — |
A few points stand out.
Poland is the most aggressive buyer. It has said it aims to hold 700 tonnes. At about 640 tonnes, that leaves roughly 700 − 640 = 60 tonnes still to go, if the target stays unchanged.
China buys steadily but holds a small share. At about 8% of reserves, gold is a minor part of China's reserves compared with Poland's 28%. Because China's reserves are so large, a small policy change means a lot of tonnes.
Smaller banks are joining in. The Bank of Korea bought about 2 tonnes through gold ETFs, its first gold addition in 13 years. The Bank of Namibia aims to raise gold from 1% to 3% of its reserves by March 2027. Neither moves the market alone, but both fit the pattern.
Who is selling gold in 2026?
The sellers matter as much as the buyers, because the 130-tonne figure is a net number.
| Central bank | Net change, year to July 2026 | Holdings |
|---|---|---|
| Turkey | −85 t | — |
| Russia | −50 t | 2,277 t |
Selling does not always mean a central bank has turned against gold. Some banks sell to raise local currency, support their exchange rate or meet domestic demand. Treat a sale as a data point, not a verdict.
If you add up only the six countries in the tables, the four buyers total 90 + 60 + 40 + 29 = 219 tonnes and the two sellers total 85 + 50 = 135 tonnes. That nets to 219 − 135 = 84 tonnes. The rest of the roughly 130-tonne total came from many smaller buyers and sellers.
July 2026: one month in detail
The July 2026 figures give a snapshot of the current mix:
- Buyers: China 20 t, Poland 8 t, Czech National Bank 2 t, and Kazakhstan, Malaysia and Bolivia 1 t each.
- Sellers: Russia 6 t, and Turkey, Jordan and Uzbekistan 1 t each.
Among these listed banks, July buying was 20 + 8 + 2 + 1 + 1 + 1 = 33 tonnes and selling was 6 + 1 + 1 + 1 = 9 tonnes, a net of about 24 tonnes. Note that Uzbekistan was a big buyer for the year but a small seller in July. Monthly figures are noisy, so it is better to judge trends over several months.
Why central bank buying matters for gold
Central bank demand works on a different clock from trading. Here is how to think about it.
- It is slow and price-tolerant. Reserve managers usually buy over months or years. They are not trying to catch the low of a 15-minute candle.
- It is a structural source of demand. Steady official buying soaks up supply that would otherwise have to find another home. Over long periods, that can act like a cushion under the market.
- It reflects long-term motives. Common reasons are diversification away from the US dollar, sanctions risk (Russia's reserves were frozen in 2022) and owning an asset with no counterparty.
- It is reported with a lag. The September 2026 report covered July. By then the market has moved on.
Central banks were often net sellers in the 1990s and early 2000s, then turned into steady buyers after the 2008 financial crisis. That shift is one of the reasons gold's long-term story changed. For the wider set of drivers, see our guide to what moves gold prices.
Why central bank buying did not stop the September 2026 drop
If central banks were buying, why did gold fall so hard? On 28 September 2026, gold futures settled at $4,135.40, down 3.52% on the day, and spot gold hit a seven-week low. That was roughly 26% below the late-January 2026 peak of about $5,590 (4,135 ÷ 5,590 ≈ 0.74, so the price was about 74% of the peak).
The answer is that short-term price is set by the marginal buyer and seller, and in September those were mostly rate-sensitive investors and traders. The widely cited reasons were:
- the Fed's rate hike on 16 September 2026, which lifted the target range by 0.25 percentage points to 3.75–4.00%;
- a stronger US dollar;
- US 10-year Treasury yields above 5.2% in late September, the highest since 2007.
When safe bonds pay more, gold, which pays no interest, becomes less attractive to hold. We explain that in Treasury yields above 5%, and the wider picture in why gold is falling.
About 130 tonnes over seven months is real demand, but it is spread out, while futures and ETF flows can be far larger over a few days. So central bank buying can shape the long-term backdrop without stopping a sharp short-term fall.
How a gold trader can use this information
Central bank data is context, not a trade signal:
- Read it monthly. Note whether net buying is rising or slowing, and who drives it.
- Do not trade the headline. "China bought 20 tonnes" does not tell you where gold will be next week.
- Let structure decide entries. Wait for a reaction at a higher-timeframe level and a market structure shift before acting.
- Keep risk fixed. A strong long-term story is not a reason to risk more per trade.
Our own model ignores fundamental headlines. It looks for a higher-timeframe imbalance, a rejection, an M15 market structure shift and then a limit entry. Every signal, including losses, appears on our live results page.
FAQ
Are central banks still buying gold in 2026?
Yes. WGC data published in September 2026 shows about 130 tonnes of net central bank buying in the year to July 2026. That is below the roughly 160 tonnes bought in the same period of 2025, so buying continued but slowed.
Which central bank bought the most gold in 2026?
Poland, with about 90 tonnes added in the year to July 2026. It holds about 640 tonnes, 28% of its reserves, and targets 700 tonnes. China was second with about 60 tonnes.
Why are Russia and Turkey selling gold?
Russia sold about 50 tonnes and Turkey about 85 tonnes in the year to July 2026. Central banks sell for many reasons, including raising local currency or meeting domestic needs, so a sale is not necessarily a bearish view on gold.
Does central bank buying stop gold from falling?
Not in the short term. In September 2026, gold fell to a seven-week low despite ongoing official buying, as a Fed hike, a stronger dollar and high Treasury yields dominated. Central bank demand is better seen as long-term background.
This article is educational and not financial advice. Trading gold and leveraged products carries a high risk of loss.
Trade the tested model
Live XAUUSD signals, an MT5 indicator and a fully automated MT5 EA, all from one tested rule set.
Trading involves substantial risk. This is educational content, not financial advice. See the risk disclosure.
