Every market has a marginal buyer — the participant whose demand sets the price at the edges. For decades in gold that was jewellery in India and China, then investment funds in the West. Since 2022 it has increasingly been central banks, and that change explains a good deal about how gold has behaved.
What central banks are doing
Central banks hold reserves — assets they can sell to defend their currency or pay for imports in a crisis. For seventy years those reserves were overwhelmingly US dollars and dollar bonds. From 2022 onward, a number of banks, led by China, Poland, Turkey, India and several Middle Eastern and Central Asian institutions, began buying gold at a pace roughly double the average of the previous decade — around a thousand tonnes a year, which is close to a quarter of annual mine supply.
The reasons are political as much as financial. The freezing of Russian dollar reserves in 2022 showed every reserve manager that dollar assets can be switched off. Gold held in your own vault cannot. Add the long-run wish to reduce dependence on a single currency, and you have a buyer with a strategic reason to keep buying regardless of price.
Price-insensitive demand
That last phrase is the important one. An investment fund buys gold when it thinks gold will go up and sells when it thinks it will go down; its demand rises and falls with the price. A central bank filling a reserve target buys a set amount every month whether the price is $2,000 or $4,000. It does not chase rallies and it does not sell dips.
On a chart that behaviour shows up as a floor. Sell-offs that would once have run for months find buyers within days. The dips get shallower and shorter. Traders notice, start buying the dips themselves in anticipation, and the floor rises.
This is the mechanism that let gold rise through 2024 and 2025 while real yields — the usual driver — were not especially supportive. The yield signal was pointing down; the flow was pointing up; the flow won, for a while.
What could change it
Floors built on flows last as long as the flow. Three things would weaken it:
- Reserve targets being reached — a bank that wanted 10% of reserves in gold and now has it stops buying.
- A dollar shortage in emerging markets that forces reserve sales — the same banks selling gold to raise dollars in a crisis.
- A sharp rise in real yields large enough that even strategic buyers slow down.
None of these is predictable from a chart. What a trader can do is watch the quarterly reserve data published by the World Gold Council and the IMF, which show whether the buying continues. As long as it does, the floor stands.
Retail and jewellery: the other side
Meanwhile the traditional buyers have become price-sensitive in the opposite direction. High prices reduce jewellery demand in India and China — the wedding-season buyer waits. Retail investors in the West tend to buy after a rally has been in the news for months and sell after a fall. Neither group sets the price any more; they respond to it. That is a new state of affairs and a reason the old seasonal patterns in gold (strong in autumn for Indian festival demand, for example) have weakened.
How much is a thousand tonnes?
Scale helps here. Global mine production is about 3,500 tonnes a year. Central banks buying roughly 1,000 tonnes annually are absorbing close to 30% of new supply before it reaches anyone else. Jewellery has historically taken around half of supply; investment and industry the rest. A new buyer of that size does not just support the price — it changes who sets it. For comparison, in the 1990s central banks were net sellers, offloading reserves at what turned out to be the bottom of a twenty-year bear market. The switch from seller to largest buyer is the single biggest structural change in gold in a generation.
What it means on the platform
For a copier the central-bank story has two practical readings.
First, gold masters with a long-only bias have had the wind at their backs for three years. A track record that starts in 2023 has never seen a real gold bear market. Look for a master whose record includes 2022, when gold fell for most of the year, and see how they handled it.
Second, the floor makes dip-buying strategies look better than they might in a different regime. A master who buys every 3% pull-back and shows a high win rate is riding the flow. That is a legitimate strategy; it is also one whose drawdown figure is understated if the flow stops. When the quarterly reserve data show the buying slowing, that is the moment to re-read the master's Max drawdown and ask whether you would be comfortable with twice it.
A buyer who does not care about the price puts a floor under it. Central banks have been that buyer since 2022. Floors last as long as the buying does — watch the reserve data, not the chart.
This article is education, not investment advice. Trading and copy trading in leveraged instruments carry a high risk of losing the funds you allocate. Read the Risk Disclosure.