Central Banks Bought a Record 289 Tonnes of Gold Into a Falling Price. Somebody Is Wrong.
- Central banks bought a net 289 tonnes of gold in Q2 2026, up 62% year over year.
- They did so during gold's steepest quarterly decline in roughly a decade.
- Gold traded around $4,460 an ounce in early September 2026.
- The pullback tracked rising Fed rate-hike expectations and higher US Treasury yields, which raise the opportunity cost of holding a non-yielding asset.
- Central bank demand is driven by reserve policy and diversification away from currency risk, not by price momentum.
In the second quarter of 2026, central banks bought a net 289 tonnes of gold — a record quarter, up 62% from a year earlier.
They did it during gold's steepest quarterly decline in about a decade.
That is not a paradox. It is two entirely different buyers, operating on different time horizons, meeting at the same price.
Buyer one: the market
Gold traded near $4,460 an ounce in early September, after retreating from its highs. The reason is unglamorous and mechanical.
Gold pays nothing. No coupon, no dividend, no earnings. Its entire return is whatever the price does. That means its main competitor is the real yield on government bonds — the yield after expected inflation.
Through the summer, expectations for a Federal Reserve rate increase rose sharply, and the US 10-year Treasury yield climbed to around 4.83%. Every increase in that yield raises the cost of holding an asset that pays nothing. Traders and momentum funds respond by selling.
This is why gold can fall during a war, during an inflation shock, and during exactly the conditions people assume gold is supposed to protect against. In the short run, gold is a rates trade far more than it is a fear trade.
Buyer two: the sovereigns
Central banks are not running that calculation.
A central bank holding gold is not seeking return. It is holding a reserve asset that has no counterparty and no issuing government — one that cannot be frozen by a foreign authority, defaulted on, or inflated away by another country's policy choices. That property became considerably more valuable to a number of governments after the sanctions of the early 2020s demonstrated what can happen to reserves held in someone else's currency.
Those decisions run on multi-year reserve allocation policy. A falling price does not deter a buyer working toward a target percentage of reserves — it makes the target cheaper to reach.
So the record buying and the falling price are not in conflict. One buyer is trading a rates signal; the other is executing a decade-long reallocation.
What this means for an individual investor
Three things worth being clear about.
Gold is not a growth asset. Over long periods its real return has been roughly zero — it has preserved purchasing power rather than compounded it. Equities compound because companies retain and reinvest earnings. Gold sits in a vault. Anyone owning gold expecting stock-like long-run returns has misunderstood the instrument.
It is insurance, and insurance has a premium. The case for a small gold allocation is that it behaves differently from stocks and bonds in specific crises — currency debasement, sovereign stress, geopolitical rupture. The cost of that diversification is the return you give up in the many years when nothing goes wrong. That is a defensible trade, provided it is made deliberately and sized small.
Gold bullion and gold miners are different assets. A gold ETF backed by physical metal tracks the metal. Gold mining companies are leveraged operating businesses with labour costs, energy costs, capital budgets, jurisdictional risk and management quality — they can rise and fall far more than the metal, in either direction, and they can underperform a rising gold price entirely. Owning miners because you wanted gold exposure is one of the more common mismatches in retail portfolios.
The structural story underneath
Set aside the quarter. The multi-year pattern is that a growing group of central banks has been shifting reserves out of foreign currency holdings and into an asset nobody else controls.
That is a slow, deliberate, price-insensitive source of demand — and it is genuinely new relative to the 1990s and 2000s, when central banks were net sellers of gold.
It does not tell you where the price goes next quarter. Nothing does. What it tells you is that the buyer base for gold has changed composition in a way that is unlikely to reverse quickly, because the reason for it — a desire to hold reserves outside anyone else's jurisdiction — is not a market view. It is a policy.
Practical takeaway
If you own a broad global index fund, you already own gold miners in proportion to their market weight, and you own no bullion at all. Those are two different exposures and a lot of investors assume the first covers the second.
If you want the diversification argument, the honest version is: hold a small, fixed percentage, rebalance to it mechanically, and expect it to disappoint you most years. That is what insurance looks like when nothing is on fire.
Frequently asked questions
Why does gold fall when interest rates rise?
Gold pays no interest or dividend. Its entire return is price change. When government bonds offer a higher real yield — the yield after expected inflation — the opportunity cost of holding gold instead rises, and investors who want a safe store of value have a paying alternative. This is why gold often falls on hawkish central bank news even when the underlying reasons people own it have not changed.
Why are central banks buying gold if the price is falling?
Because they are not trading it. Central banks hold gold as a reserve asset that carries no counterparty and no issuing government, which makes it useful as insurance against currency and geopolitical risk. Their buying decisions run on multi-year reserve policy rather than quarterly price moves, and a lower price simply means the same policy allocation costs less to fill.
How much gold should an individual investor hold?
There is no single correct answer, and gold's long-run real return has historically been close to zero — it preserves purchasing power rather than compounding it. Investors who hold it typically do so as a small diversifying allocation, often in the low single digits to around 10% of a portfolio, on the reasoning that it behaves differently from stocks and bonds in certain crises. Anyone holding it should be clear it is insurance, not a growth engine.
Primary sources
Data & disclaimer: This article is for educational purposes only and is not financial or investment advice. Figures reflect data available as of Sep 10, 2026, and conditions change — always confirm current pricing, rates and rules with the provider before you act. Written by Elizabeta Dimoska. See our editorial standards and disclosure.
Comments