The world’s physically backed gold ETFs finished October with more metal and less money under management. According to the World Gold Council, their holdings rose by 19 metric tonnes to a month-end record of 4,275 tonnes. Net inflows reached $2.7 billion. Yet dollar assets under management fell 6% to $542 billion because the lower gold price outweighed the inflows. A falling asset total, read on its own, would have missed the addition of physical gold. World Gold Council’s October ETF report
The sell-off subsequently set a lower mark. Reuters reported that spot gold traded at $3,769.68 per troy ounce at 10:50 GMT on November 18, after touching $3,757 earlier that session, its lowest level since September 2025. That is an intraday price. The October ETF figures cover a month, and official-sector purchases are measured over another window again. Putting those measures on the same footing would obscure how money and metal moved. Reuters spot-market report
What moved inside the funds
An investor who buys an existing fund share from another investor changes who owns that share. A creation expands the number of shares outstanding and can bring metal into the fund. The SPDR Gold Trust’s filing with the U.S. Securities and Exchange Commission offers a specific example. Authorized participants can create or redeem shares in baskets of 100,000 in exchange for gold and any cash represented by the baskets. Secondary trading can be heavy while physical inventory barely moves because the metal transfers belong to transactions at the trust itself. The number of shares changing hands is therefore a poor measure of how much gold went into custody. SPDR Gold Trust quarterly filing
Holdings tell holders how much gold the vehicles had at the reporting date. Assets under management reflect what that inventory was worth, and the fund-flow figure estimates the net money coming in or going out. Even a global net inflow can conceal selling in a large market. In October the council counted inflows of $1.8 billion in Europe and $1.3 billion in Asia, while North America recorded outflows of $512 million, driven by U.S.-domiciled funds. The geographical split limits any reading of the total as a uniform vote on gold. It also leaves open whether allocations changed again as November prices fell. October ETF flows by region
The bid for gold meets a higher real yield
Bullion pays no interest. When the real yield available on inflation-protected Treasuries rises, the yield forgone by holding gold becomes more significant, all else equal. The U.S. Treasury’s ten-year real par yield rose from 2.62% on November 2 to 2.92% on November 16. Yet it had already reached 2.91% on November 13 and eased to 2.89% on November 17. Those readings support the case for greater competition from bonds over the first half of the month. They leave plenty of room for other explanations of individual trading sessions. U.S. Treasury real yield data
The dollar adds another channel. A stronger dollar raises the local-currency cost of a given dollar gold price for a buyer whose currency has weakened against it. The Federal Reserve’s broad dollar index rose from 121.66 on November 9 to 122.65 on November 13. Reuters subsequently reported a firm dollar during the November 18 sell-off. The effects of currency and yields can overlap, and the data here cannot assign a share of the price move to either one. A dollar-quoted gold loss can also translate into a different result for a holder who measures returns in an unhedged local currency. Federal Reserve broad dollar index · Reuters November 18 report
The official tally has a reporting lag
The World Gold Council estimates that central banks and other official institutions bought a net 321 tonnes in the third quarter, bringing the January-to-September tally to 675 tonnes. The directly reported share of the quarter’s net buying was 120 tonnes. About 63% of the estimated total was unreported in the underlying statistics, which the council says can reflect delays as well as nondisclosure. Reserve purchases may absorb meaningful supply over time. The quarterly estimate cannot identify the buyers on November 18 or establish a level at which official buyers will step in. World Gold Council’s Q3 official-sector estimates
A useful comparison will come when ETF holdings and flows cover the same November sell-off. If tonnage falls alongside broad regional outflows, the explanation would include a withdrawal of capital from the tracked funds as well as a change in valuation. If holdings continue to rise while dollar assets fall, the October distinction would remain useful. Either result could be reconsidered once more official purchases are disclosed or the estimates revised. For now, a lower November spot price and a larger October ETF inventory are both part of the record. Each describes a different transaction or measurement. ETF methodology and flow data · Official purchase reporting