Here is a quirky fact about the oldest alternative investment there is. In the second quarter of 2026, people who own gold through ETFs sold it, jewelry buyers bought less of it, and the metal still had one of its strongest quarters for a very particular kind of buyer. Central banks added roughly 289 tonnes to their reserves, according to figures the World Gold Council published on July 30. Same asset, same quarter, two completely different crowds making opposite moves. Which one do you suppose knows something the other does not?
A rally that cooled on schedule
The World Gold Council’s Q2 2026 Gold Demand Trends report described a market that peaked early and then caught its breath. Louise Street, a senior markets analyst at the council, put it this way: “Gold’s early-year rally reversed in the second quarter, with prices consolidating after correcting from record highs.” Notice what that sentence does and does not say. Prices fell back from records. Demand did not collapse.
Total demand came in at 1,269 tonnes for the quarter, flat against the same period a year earlier. For the first half, the council counted 2,522 tonnes, about 2% higher than a year before, worth roughly US$380 billion. A flat quarter after a record run is not a crash. It looks more like a market digesting a very large meal.
That distinction matters for anyone who treats gold as a diversifier inside a broader portfolio. A rising price and rising demand are different things, and the second is usually the more durable story. When the price runs ahead of the buyers, corrections tend to hurt. When buyers are still arriving at lower prices, they tend to be shallower. The Q2 data suggests the buyers were still arriving, just not the ones retail investors usually watch.
Who sold, and who did not
Start with the funds, which were the quarter’s net sellers. Global gold ETFs, held by institutions as well as individual investors, lost about 45 tonnes in the second quarter, with June the heaviest month at roughly 74 tonnes of outflows. Over the first half as a whole, ETF demand was still modestly positive, at about 18 tonnes, because the first quarter had been so strong. So the funds gave back some of what they had gathered. They did not unwind the year.
Nor did they stay away. ETF buyers came back strongly in the third quarter. The council’s September 2026 ETF report shows Q3 was a record quarter for gold ETFs, with about US$31 billion of inflows, holdings up 211 tonnes and a record 4,256 tonnes in total, even as the price eased in September.
Physical retail demand told a different story. Bar and coin buyers stayed net buyers, adding about 307 tonnes in the second quarter. That was about 3% less than a year earlier, yet demand was still around 21% higher across the first half. Jewelry was the weak spot. Demand fell about 17% year over year to roughly 278 tonnes, and India, one of the largest jewelry markets, dropped about 15% to roughly 75 tonnes. The council’s report noted that consumer affinity for gold jewelry appears intact despite the volume decline, with buyers shifting toward lighter pieces. When the price is high, people do not stop wanting gold. They buy less of it by weight.
There was one small counterexample worth mentioning. Indian gold funds added about 4 tonnes during the quarter, bucking the regional trend, as local investors treated the price correction as a chance to buy. That is a modest number, and nobody should build a thesis on it. But it is a reminder that the same dip reads as a warning to one investor and an invitation to another.
The buyer with the longest time horizon
Now the central banks. Net purchases reached about 289 tonnes in the second quarter, up roughly 62% from the same quarter a year earlier. The jump from the first quarter was even sharper, since the council’s data put Q1 purchases at about 56.5 tonnes. The council’s Central Bank Gold Reserves Survey 2026 adds a layer of intent: a record 45% of respondents said they expect to increase their own gold holdings over the next 12 months.
There is a caveat, and the council states it plainly. Central banks remain on course for another strong year of net buying, but full-year 2026 purchases will likely land below 2025 levels. A slower pace is not a retreat. It is still a large, steady bid from institutions that are not trading on a quarterly scorecard.
That is not a comforting story for everyone, and it should not be told as one. A market supported by a handful of large official buyers can look sturdy right up until those buyers change their pace. Anyone leaning on the central bank bid as a reason to hold gold should keep in mind that it is a trend, not a guarantee, and that trends in official reserves shift with politics, currencies, and priorities that individual investors cannot see.
This is where the official and investor views of gold diverge in an instructive way. A central bank holding gold is not trying to beat an index. It is holding an asset that carries no counterparty, pays no coupon, and cannot be diluted by a policy decision. An investor holding a gold ETF, whether an individual or an institution, is often doing something subtly different, expressing a view on where the price goes next, and acting on it when the price moves. One approach is a structural allocation. The other can easily become a trade.
It helps to remember how the arithmetic works. Central bank buying of 289 tonnes in a single quarter is large enough to offset a meaningful amount of fund selling, which is part of why the total came in flat rather than lower. The other part was over-the-counter (OTC) demand, the council’s estimate of less visible off-exchange investment, which reached about 327 tonnes, up 91% from a year earlier. Without OTC, demand fell 14%. The council’s figures show total supply matching demand at 1,269 tonnes, with mine production of about 966 tonnes, up roughly 2%, while recycling fell about 6%. That match is built into the tables, where supply and demand balance by construction, so on its own it proves little. What the mix shows is that sellers found buyers, and the buyers with the deepest pockets and the least urgency did much of the absorbing.
What this means for a self-directed investor
None of this is a signal to buy or sell anything. It is a reason to ask a sharper question about why gold is in a portfolio at all. If the answer is that it diversifies against stress in other assets, then a quarter of ETF outflows after record highs is mostly noise. If the answer is that it will keep making new highs, then the second quarter should have been uncomfortable, and that discomfort is information worth taking seriously.
It is also worth being honest about how the holding is built. Physical metal, a bullion-backed ETF, and a fund that owns mining shares behave differently, carry different costs, and answer different needs. The Q2 numbers describe flows into and out of these vehicles. They do not tell you which vehicle suits a particular portfolio, and nothing in a demand report can.
The broader lesson applies well beyond gold. Alternatives are often sold on their independence from stocks and bonds, yet the people holding them can still behave in correlated ways, selling together when prices wobble. The institutions with the longest horizons, in this case central banks, tend to behave differently from the investors with the shortest ones. Understanding which camp you belong to, before the next drawdown rather than during it, is probably worth more than any forecast.
Gold may well revisit its highs, or it may spend a long stretch going sideways. Nobody reliably knows, including the analysts who publish price targets each autumn. What the second quarter did show is how differently the same asset can look depending on who is holding it and why. That is a useful thing to know before deciding whether, and how, to own any of it.
This article is written for educational and informational purposes only and does not constitute financial or legal advice. The views and analytical frameworks presented draw on publicly available information and reported commentary from industry participants. Readers are encouraged to consult primary sources and form their own informed views on these complex topics.
