Gold's Record Buyer Is a Quarter of One Day's London Turnover — and the Metal Still Fell 26% From Its January Peak insight cover
Markets / EventNEM · B · FNV•22 min read

Gold's Record Buyer Is a Quarter of One Day's London Turnover — and the Metal Still Fell 26% From Its January Peak

Central banks and other official institutions bought a record 321 tonnes of gold in the third quarter of 2026, 42% more than a year earlier, and gold still ended November 10 at $4,021.36 an ounce — 25.6% below its January 29 record of $5,405. The official bid cannot set the price: 321 tonnes is roughly $44bn of metal, about a quarter of the $191bn London trades in an average day. The seller is the price-elastic physical buyer, with jewellery consumption down 17% and bar-and-coin demand down 19%, while recycled supply fell 8% — so owners of old metal are not capitulating. For equities, the read-through into 2027 is a margin squeeze on rising costs, landing first on Newmont and Barrick Mining, then on jewellers such as Chow Tai Fook whose margin is a metal-price subsidy.

Published Nov 10, 2026Updated Nov 10, 2026

Gold, spot

$4,021.36/oz

Nov 10, 2026, 11:33 a.m. ET; down 2.8% on the session

Below the January record

-25.6%

Versus the $5,405/oz spot peak of Jan 29, 2026

Official buying, Q3 2026

321t

+42% y/y; the highest quarterly total since Q4 2024

In London turnover terms

~23%

Of one average trading day at $191bn a day in Q3 2026

Spot gold fell 2.8% to $4,021.36 an ounce at 11:33 a.m. ET on November 10, 2026, after US consumer prices rose 0.4% in October against a 0.3% forecast, lifting the dollar and Treasury yields. A 25-basis-point Fed hike in December is now fully priced, with the probability of a 50-basis-point move at 10.5%. The metal sits 25.6% below the $5,405 record it set on January 29, 2026 — and it did so in a quarter when official buyers set an all-time record for tonnage.

Gold, spot

$4,021.36/oz

Nov 10, 2026, 11:33 a.m. ET; down 2.8% on the session

Below the January record

-25.6%

Versus the $5,405/oz spot peak of Jan 29, 2026

Official buying, Q3 2026

321t

+42% y/y; the highest quarterly total since Q4 2024

In London turnover terms

~23%

Of one average trading day at $191bn a day in Q3 2026

Jewellery consumption, Q3

313t

-17% y/y; the weakest third quarter in the data series

Recycled gold, Q3

314t

-8% y/y; owners are sitting on metal, not selling it

The arithmetic of the bid

The record buyer is a quarter of one day's London turnover

Central banks and other official institutions took 321.2 tonnes in the third quarter, up 42% year on year and the strongest quarter since late 2024, taking nine-month buying to 675 tonnes. Converted at the quarter's average price of $4,262 an ounce, that is roughly $44bn of gold. The LBMA reports average daily London gold turnover of 44.9 million ounces worth $191.4bn. The record official buyer absorbs about a quarter of a single average trading day — which is why a record bid failed to stop a 26% drawdown.

Official buying is a shock absorber, not a price-setter. At roughly $44bn a quarter against $191bn of London turnover a day, the official sector can cushion a demand shock but cannot manufacture a bid strong enough to lift the marginal price.

Q3 2026 gold demand by buyer

Official buyers took the largest single block of demand in the quarter and the price still fell.

Unit: tonnes

Central banks and official institutions

+42% y/y

321

Jewellery consumption

-17% y/y

313

Bar and coin

-19% y/y

268

Gold ETFs

-6% y/y

211

OTC and other

+43% y/y

144

Technology

+3% y/y

85

World Gold Council, Gold Demand Trends Q3 2026: demand by buyer against supply by source.
SegmentQ3 2026 (tonnes)y/yRead for the price
Central banks and official institutions321.2+42%Record tonnage, but only about $44bn — a quarter of one London trading day
Jewellery consumption312.9-17%Weakest third quarter in the data series; value held near $43bn
Bar and coin267.8-19%Retail investment destroyed by the January price spike
Gold ETFs211.2-6%Back to inflows after the Q2 outflow; holdings a record 4,256t
OTC and other143.5+43%The residual bucket where the imbalance gets absorbed
Technology84.8+3%Highest quarterly total since Q4 2021; electronics +5%
Mine production1,070.9+3%A record third quarter — supply is not this story
Recycled gold314.5-8%Scrap supply fell even as the price fell: no capitulation

Who is actually selling

The seller is the marginal buyer who stopped bidding, not the owner of old metal

Strip out the official sector and the balance of the market is small and shrinking. Price-elastic physical demand — jewellery fabrication of 340.1 tonnes plus bar-and-coin demand of 267.8 tonnes — fell by roughly 143 tonnes year on year, about $19bn of metal at the quarter's average price. The official sector's extra 95 tonnes replaced barely two-thirds of that withdrawn bid. The rest was picked up by the over-the-counter bucket, which grew 43% to 143.5 tonnes and is the least transparent line in the dataset.

  • The price-elastic household in China: wholesale demand of 234 tonnes in Q3 was down 21% year on year and 44% below the ten-year average of 418 tonnes, even as the People's Bank of China bought 63 tonnes in the same quarter.
  • The western ETF holder: North America was the only region with net outflows in October, at $512mn, even as Europe took in $1.8bn and Asia $1.3bn — the marginal buyer migrated, it did not disappear.
  • Parts of the official sector itself: Russia's gold holdings fell to their lowest level since January 2020, Kazakhstan has been selling, and Poland's central bank chief has proposed raising up to 48 billion zloty by selling gold.
  • Not the recyclers: recycled supply fell 8% to 314.5 tonnes and slipped again from the prior quarter, so high-price owners are holding rather than liquidating.
  • Not the miners: mine production hit a third-quarter record of 1,070.9 tonnes, up 3% year on year and 13% sequentially, with strong output gains in Canada, Indonesia and Chile.

October 2026 gold ETF flows by region

Four straight months of global inflows were carried by Europe and Asia while North America kept selling.

Unit: US$ million

Europe

18th consecutive weekly inflow

1,800

Asia

Fourth consecutive month of inflows

1,300

Other regions

62

North America

The only region with outflows

-512

Here is the part that the headline numbers hide. Total demand of 1,368.6 tonnes was only 1% lower than a year earlier and total supply fell 1% as well — the tonnage balance was almost unchanged while the price dropped about 11% during the quarter. A near-flat balance cannot produce an 11% price fall on its own. What changed is who is willing to hold metal at the margin: the price-insensitive buyers kept buying, and the buyers who set the clearing price in the futures and exchange-traded complex stopped adding.

October made that concrete. Global gold exchange-traded funds took in $2.7bn and holdings rose 19 tonnes to a record 4,275t, yet assets under management fell 6% to $542bn because the metal got cheaper. Record tonnes at a shrinking dollar value is the signature of a market where new capital has stopped arriving from the price-sensitive side, and where the price is being set in a thinner futures market rather than in the physical trade.

Gold's bear market is not a demand shortage. It is a change in who will hold metal at the margin, and the official sector is now large enough to be the buyer of last resort rather than the buyer that sets the price.

Producer read-through

Costs are rising into a falling price, and guidance still assumes $4,500 gold

The producer complex shows the squeeze in a single comparison. Barrick Mining reported third-quarter revenue of $5.07bn, up 22% from $4.15bn a year earlier, while all-in sustaining costs rose 24% to $1,911 an ounce from $1,538. Revenue growth and cost growth are converging on each other, and the convergence happens at a price that has since fallen 5% below the quarter's realized level of $4,243. Cost inflation is arriving just as the revenue line stops widening — the worst possible order for operating leverage.

Margin arithmetic at the November 10 spot price of $4,021.36 an ounce.
ProducerQ3 2026 realized priceQ3 2026 all-in sustaining costMargin per ounce at $4,021 spot2026 guidance price assumption
Barrick Mining$4,243/oz$1,911/oz$2,110/oz, down from $2,332$4,500/oz; AISC guided $1,760-$1,950
Newmont$4,317/oz$1,454/oz$2,567/oz, down from $2,863$4,500/oz; AISC guided $1,520/oz

Newmont is the clearest case because it guides to its own cost step-up. Fourth-quarter all-in sustaining costs are guided to $1,900 an ounce against a full-year guide of $1,520, and the full-year plan assumes $4,500 gold. At the November 10 spot price, that implies a margin near $2,121 an ounce — 26% below the $2,863 it earned in the third quarter — and a fourth quarter in which record production volumes and record cash generation arrive with a materially thinner spread.

Both major producers are running 2026 plans off a $4,500 gold assumption that is 12% above spot. The fourth-quarter results and the 2027 guidance that follows them are where those assumptions have to be re-based, and that is the next scheduled repricing of the entire gold equity complex.

Royalty read-through

Royalties are the cleanest print on the marginal price — and one of them lands tomorrow

A royalty has no cost inflation, so every dollar of price change lands straight on the revenue line. Wheaton Precious Metals reported third-quarter revenue of $772mn, up 62% year on year, with the increase driven primarily by a 36% jump in the average realized gold-equivalent price; operating cash flow was $645mn and net earnings $410mn. The same purity runs through Franco-Nevada, which recognised $580.9mn of second-quarter revenue on 132,405 gold-equivalent ounces sold, or $4,388 per ounce. That is the clearest published number for what a marginal gold ounce is worth to a business with no cost base.

  • Franco-Nevada publishes third-quarter results on November 11, 2026 — the first clean print of a lower gold price on a royalty income statement, and the cleanest single read on the new marginal price.
  • Hold the same volumes constant and $4,021 gold implies roughly $532mn of quarterly revenue at Franco-Nevada, about 8% below the second quarter, with no offsetting cost relief.
  • Wheaton Precious Metals sold 192,270 gold-equivalent ounces against 230,965 produced, so inventory covered the difference — a buffer that reverses if production rises while the price falls.
  • Both companies left 2026 volume guidance intact, Wheaton Precious Metals at 860,000-940,000 gold-equivalent ounces and Franco-Nevada on its existing plan, which means the price reset shows up in 2027 revenue rather than in 2026 volumes.

That timing matters for portfolios. Royalty and streaming vehicles have been the defensive expression of gold because they carry no operating cost inflation, and that defence is intact at the corporate level. It is not intact at the revenue level: a royalty on a falling gold price is simply a smaller royalty, and the buffer that absorbed Wheaton Precious Metals's third-quarter gap — inventory drawn down against 230,965 ounces produced — is a finite resource, not a permanent one.

Downstream read-through

Jewellery margins are a metal-price subsidy that reverses before volumes recover

The clearest evidence that a metal-price subsidy exists at scale sits in the jewellers' accounts. Chow Tai Fook reported fiscal 2026 revenue of HK$94,398mn, up 5.3%, with gross profit margin up 280 basis points to 31.5% and a record 20.1% operating margin on operating profit of HK$18,850mn. Luk Fook reported gross margin up 3.6 percentage points to a record 36.7% and profit attributable to shareholders up 86.0% to HK$2.05bn on revenue of HK$17.21bn. Neither company sold more gold jewellery than it did when gold was $3,000; they earned more per unit of volume because the metal got more expensive.

What a reversion in metal prices costs the jewellers, holding revenue constant.
JewellerLatest gross marginHistorical reference pointGross profit if the margin revertsCash flow signal
Chow Tai Fook31.5% (FY2026)20.5% in FY2024About HK$19.4bn, roughly HK$11.9bn below the FY2026 levelOperating cash flow collapsed to HK$1.02bn from HK$13.8bn two years earlier
Luk Fook36.7% (FY2026, record)27.0% in FY2023About HK$4.6bn, roughly HK$1.7bn below the FY2026 levelNet profit margin of 11.7% against 8.0% in FY2025

The cash flow tells the harder story. Chow Tai Fook put HK$14.8bn into inventory during fiscal 2026 and funded it with HK$8.7bn of long-term borrowing while paying HK$5.3bn of dividends; operating cash flow fell to HK$1.02bn and free cash flow to HK$425mn. A jeweller that has raised debt to hold metal bought at peak prices is exposed to both a lower selling price and a slower inventory turn — the hedge that protected the margin unwinds into the write-down.

The jewellers' record profits and record margins are a function of the metal price, not of the customer. If gold settles near $4,000 or below, roughly HK$13bn of combined annual gross profit at the two Hong Kong retailers disappears even before any volume benefit shows up.

Benchmark and legal

The benchmark that settles physical contracts is still being litigated

The LBMA's third-quarter report shows how much of this market now runs on the London auction: most London gold prices settled between $4,000 and $4,400 an ounce during the quarter, London vault holdings rose 3.78% to 9,821 tonnes worth $1.3 trillion, and average daily turnover was 44.9 million ounces worth $191.4bn. Those are the contracts that miners, royalty companies and refiners settle against. The benchmark's legal stability is therefore an input into every physical gold revenue line in the world.

The benchmark in numbers

LBMA Gold Price

Electronic auction run by ICE Benchmark Administration

Replaced the telephone-based London Gold Fix in March 2015

Q3 2026 auction range

$4,000-$4,400/oz

Majority of London auction prices settled inside that band

London vault holdings

9,821t, $1.3tr

Up 3.78% during Q3 2026

Average daily turnover

44.9m oz, $191.4bn

LBMA reported gold volumes, Q3 2026

Predecessor benchmark

London Gold Fix

Class-action settlement approval hearing held October 1, 2026

The LBMA Gold Price is itself administered by ICE Benchmark Administration through an electronic, tradeable, auditable auction... in line with the IOSCO Principles for Financial Benchmarks.

CFTC submission on precious-metals benchmark administration, September 8, 2026
  • The litigation that still matters is legacy: the class action over the 2004-2014 London Gold Fix names Bank of Nova Scotia, Barclays, Deutsche Bank, HSBC, UBS and Société Générale, and its approval hearing was scheduled for October 1, 2026, on top of announced settlements that bring the total to $152mn.
  • Those cases do not touch the current electronic auction, but they keep the benchmark question alive in front of regulators and in the contracts that reference a London settlement price, and the CFTC's own rules keep the LBMA Gold Price as the reference point for physical metal.
  • Intercontinental Exchange now runs the LBMA Platinum and Palladium benchmarks as well, from July 1, 2026, widening the administrator's product set and its regulatory surface.
  • For producers and royalty companies, the margin consequence is direct: forward sales, streaming deliveries and lease financing settle against a London auction price, so any dislocation between that auction and futures shows up in realized prices rather than in accounting footnotes.

Horizons

Near term the trigger is a December meeting; over 2027 the floor becomes a policy decision

  • The December Fed decision is the near-term hinge: 25 basis points is fully priced, and a 50-basis-point hike carries 10.5% odds after October's hot inflation print.
  • Watch the dollar and real yields rather than the headline rate — the pullback has been driven by higher real yields and a stronger dollar, and the Fed has already raised 25 basis points in September to about 3.9%.
  • Franco-Nevada reports third-quarter results on November 11, 2026, and fourth-quarter prints from Barrick Mining and Newmont will carry the 2027 guidance reset.
  • Monthly ETF flows are the tell for the marginal buyer: if North American outflows persist while Europe and Asia keep absorbing, the official sector becomes the buyer of last resort at ever-lower prices.
  • The structural supports are intact for 2027: 89% of reserve managers expect global central bank gold holdings to rise over the next twelve months, and a record 45% expect their own to rise, with official buying averaging 1,000 tonnes a year over the past four years.

The one-to-three-year conclusion is uncomfortable for the bull case and constructive for the volume case. The official bid is large, persistent and indifferent to price, but it is capped: at roughly $44bn a quarter it can absorb a demand shock without setting a price. That makes gold's floor a policy decision rather than a market clearing, and it makes volume growth more valuable than price growth — which is why Zijin Mining, which lifted mined gold 13% to 1,502 thousand ounces in the first half of 2026 with basic earnings per share up 68% to RMB 1.473, is on a different footing from the producers whose earnings are a function of the metal's price.

One more Chinese data point closes the loop. Zhongjin Gold, the refiner that mints and supplies metal into the official and retail channels, grew nine-month revenue 23.4% to CNY 66.6bn and net income 54.6% to CNY 5.69bn — yet it reported negative operating cash flow of CNY 293mn in the third quarter as inventory built, and net debt to EBITDA jumped from 0.84x at June 30 to 2.79x at September 30. The refiner that monetises the record official bid is funding it from its own balance sheet, and its market value still fell from CNY 129.5bn at March 31 to CNY 106.2bn at September 30. The equity market has already decided that a lower gold price is the base case.

The bear market is real, the record official buyer is too small to end it, and the sellers are the marginal holders rather than the old owners. Translate that into equities and the 2027 winners are volume stories; the 2027 losers are the businesses whose margins were built out of the metal's price.

Where the bear market actually lands

NNewmontNEM--
--Vol --
-
Mixed
  • Q4 2026E all-in sustaining costs of $1,900/oz against a $1,520/oz full-year guide mean the margin falls to about $2,121/oz at $4,021 spot, 26% below the Q3 level.
  • A realized price of $4,317/oz and AISC of $1,454/oz still leave the lowest-cost major producer with the most room to absorb a lower price.
  • 2027 guidance is built on a $4,500/oz assumption, so any reset to spot resets earnings, the $0.26 quarterly dividend and buybacks together.
BBarrick MiningB--
--Vol --
-
Bearish
  • AISC rose 24% to $1,911/oz while revenue rose 22%, so costs are outrunning the realized price and the gap widens as gold falls.
  • Guidance assumes $4,500/oz; at $4,021 the margin compresses to about $2,110/oz from $2,332/oz in Q3.
  • Fourth-quarter results and 2027 guidance are where the cost base gets re-based, with buybacks and the dividend absorbing the first shock.
FFranco-NevadaFNV--
--Vol --
-
Mixed
  • Q2 revenue of $580.9mn on 132,405 gold-equivalent ounces equals $4,388 per ounce, so flat volumes at $4,021 gold are worth about $532mn, down 8%.
  • The Q3 2026 release on November 11, 2026 is the first clean print of a lower gold price on a royalty income statement.
  • No cost inflation and a net cash balance sheet make royalties the defensive expression, but volumes follow the same mine plans.
WWheaton Precious MetalsWPM--
--Vol --
-
Mixed
  • Q3 revenue rose 62% to $772mn on a 36% higher realized gold-equivalent price, so the royalty income statement is a direct lever on spot.
  • 192,270 ounces sold against 230,965 produced means inventory covered the gap, a buffer that reverses if production rises while the price falls.
  • 2026 guidance of 860,000-940,000 gold-equivalent ounces is unchanged, so the price reset shows up in 2027 revenue rather than 2026 volumes.
2Zijin Mining2899.HK--
--Vol --
-
Bullish
  • Mined gold rose 13% to 1,502 thousand ounces in H1 2026 with basic earnings per share up 68% to RMB 1.473, so growth comes from volume as well as price.
  • Net debt to EBITDA of 0.70x and a 19.6% return on equity fund the expansion plan without leaning on the gold price.
  • The risk to the bull case is the copper and lithium build-out; a lower gold price removes part of the funding cushion.
6Zhongjin Gold600489.SS--
--Vol --
-
Mixed
  • Nine-month net income rose 55% to CNY 5.69bn, yet third-quarter operating cash flow was negative CNY 293mn as inventory built.
  • Net debt to EBITDA jumped from 0.84x at June 30 to 2.79x at September 30, so the refiner funds the official bid from its own balance sheet.
  • Market value fell from CNY 129.5bn at March 31 to CNY 106.2bn at September 30 even as earnings rose, so the equity already prices a lower gold price.
1Chow Tai Fook Jewellery1929.HK--
--Vol --
-
Bearish
  • FY2026 gross margin of 31.5% against 20.5% in FY2024 is a metal-price subsidy; reversion on flat revenue costs about HK$11.9bn of gross profit.
  • Operating cash flow collapsed to HK$1.02bn from HK$13.8bn two years earlier as HK$14.8bn went into inventory bought at peak prices.
  • A lower gold price relieves the cash squeeze but unwinds the margin that funded a HK$18.85bn operating profit, and the hedge runs both ways.
0Luk Fook Holdings0590.HK--
--Vol --
-
Bearish
  • Gross margin hit a record 36.7% in FY2026 from 27.0% in FY2023, and a reversion on the same revenue costs about HK$1.7bn of gross profit.
  • Profit attributable to shareholders rose 86.0% to HK$2.05bn on a 29% revenue gain, a price-led earnings base with no volume offset.
  • Slower gold prices help volume only gradually, so the margin reversal lands first and any volume recovery arrives later, if at all.