The World Gold Council published its Gold Demand Trends report for the third quarter on November 3, and the quarter split the gold market cleanly in two. Institutions bought. Households did not.
Central banks added a net 321 tonnes to reserves, up 42% year on year and the heaviest quarter since Q4 2024. Gold ETFs reversed a second-quarter outflow with 211 tonnes of inflows, lifting holdings to a record 4,256 tonnes. Over the same three months, U.S. retail demand for bars and coins came to 8.2 tonnes — down 40% from 13.5 tonnes a year earlier, and the weakest quarter for American physical buying since the fourth quarter of 2019.
The official sector bought roughly 39 times as much gold as U.S. retail investors did.
The reason given was price direction
Global bar and coin demand fell 19% to 268 tonnes, a four-year low. The Council's explanation is specific: "Rangebound prices, expectations that US interest rates would remain higher for longer, and rallying equity markets discouraged fresh buying."
Louise Street, Senior Markets Analyst at the World Gold Council, framed the divergence this way: "Gold ETF inflows came roaring back, with geopolitical risk and concerns around fiscal discipline driving institutional interest, in contrast to retail investors who held back amid uncertainty over the gold price outlook."
Each of those three stated reasons is a forecast about what happens next. None of them is a statement about how much gold a portfolio should hold.
The pullback is from a record, not a collapse
Context matters here. Bar and coin demand across the first nine months of 2026 reached 1,053 tonnes — the strongest such stretch since 2013 — worth $156 billion, nearly matching the entire full-year 2025 total of $157 billion. Retail buyers were not absent in 2026. They bought early and heavily, then stopped once the price stopped climbing.
The quarter they sat out was not a bargain, but it was not an outlier either: gold averaged $4,262 an ounce in Q3, up 23% year on year and the third-strongest quarterly average on record.
The behavior was not universal. European bar and coin demand fell 31% to 20.5 tonnes and India's fell 40% to 55 tonnes — India alone accounted for roughly 60% of the global year-on-year decline. China's rose 7% to 78.9 tonnes, with year-to-date demand at a record 393 tonnes. Jewellery demand fell 17% to 313 tonnes, the weakest third quarter in the Council's data series.
Practical takeaways
- Separate the allocation decision from the entry decision. Whether metals belong in your retirement portfolio is a question about diversification and your other holdings. Whether gold is about to rise is a different question, and only the first one has a defensible answer.
- A rebalancing band replaces the forecast. Pick a target weight and a drift tolerance, then buy or trim when the band breaks. That mechanism would have had you selling into the 2026 run-up and adding during a flat quarter — the opposite of what "waiting for a clearer trend" produces.
- Central bank demand is not a buy signal. The Council attributes official-sector buying to portfolio diversification and risk management, not to a return forecast. Reserve managers are solving a different problem than you are.
- Round trips are costly in physical metal. Acting on sentiment swings means paying a dealer's bid-ask spread on the way in and again on the way out. Metal held in an IRA adds custodian and storage fees to that bill.
- Check which window a headline uses. The same report supports "retail demand at a four-year low" and "strongest nine months since 2013." A quarterly change and a year-to-date total can point in opposite directions without either being wrong.
Sources: World Gold Council, Gold Demand Trends Q3 2026 and accompanying press release, "Central bank buying and ETF inflows support gold demand in Q3" (November 3, 2026); World Gold Council, Gold Demand Trends Q3 2026, Investment chapter; GoldSilver, "Q3 Central Bank Gold Buying +42%, Bar-and-Coin Demand -19%" (November 2026).