Paper vs Physical Gold and Silver: What the Evidence Shows

Most people talk about “the gold price” as if one market sets it. In early October 2025, a single burst of futures selling at the New York open reportedly dumped about 91 tons of paper gold in minutes. The headline price fell, while buyers in Hong Kong and Shanghai were trading physical metal on quite different terms.

That gap between futures screens and physical metal sits at the centre of the paper vs physical gold and silver debate. A year later the debate has not gone away. World Gold Council (WGC) data released this week shows physically backed gold ETF holdings at a record 4,256 tonnes in September 2026, and positioning on COMEX, the New York futures exchange run by CME Group, remains large.

The October 2025 episode serves here as a case study, not breaking news.

Read on and you will have a working way to interpret futures positioning, ETF flows and Asian physical signals. You will also be able to tell which claims in this debate rest on evidence and which rest on assertion.

What do “paper” and “physical” gold and silver actually mean?

You probably already hold a rough version of this distinction. Paper exposure is a claim on metal or its price: COMEX futures, options, over-the-counter (OTC) derivatives traded privately between banks and clients, and unallocated accounts where you are owed metal but no specific bar is yours. Physical exposure means bars, coins or allocated vaulted metal with your name on specific pieces.

Physically backed exchange-traded funds (ETFs) sit in between. You own fund units, while a custodian holds the bars and authorised participants (large dealers) create or redeem units by delivering or withdrawing metal.

The two worlds are not sealed off from each other. Futures can end in delivery, producers hedge with them, and benchmarks lean on them.

Once you see how gold futures work, the logic of the table above becomes clearer: a single contract is a leveraged claim on metal, and the central order book sets its price rather than any dealer’s bar premium.

How the two sides of the debate read the same market

CME Group describes speculators and hedgers meeting in one central order book. The Commodity Futures Trading Commission (CFTC) defines open interest as the total of open long and short positions on regulated exchanges, and stresses that big-trader activity is not automatically manipulation.

The opposing camp argues leveraged futures selling can temporarily override physical fundamentals. It cites US Department of Justice and CFTC spoofing cases against individual bank traders as precedent, cases regulators treat as exceptions rather than design features.

History offers a test. In early 2020, trouble converting 400-oz London bars into 100-oz COMEX-deliverable bars opened an unusually wide COMEX-London spread, which arbitrage and new logistics later closed. Divergence is a reason for you to investigate, not a permanent verdict.

How does a flash selloff happen, and what did October 2025 show?

The October 2025 case shows how one sudden move can carry two completely different stories.

Reported claims, not verified figures About 91 tons of paper gold and about 1,400 tons of paper silver were reportedly sold in flash moves at the COMEX open.

Commentator Andrew Maguire called the selling an officially sanctioned move that he said exceeded normal position limits. In his account it was timed for China’s Golden Week closure, which ran from 30 September to 8 October 2025, and flushed out Chinese retail longs before institutional buyers returned. He said Hong Kong buyers bought the dip once trading reopened.

He tied the move to three derivative-related events:

• October first notice and options expiry
• Bank for International Settlements quarter-end position squaring
• The US non-farm payrolls release

He set this against a hostile macro backdrop: a dollar index near 102.3, a 10-year US yield of 5.31% and a 30-year yield of 5.69%. He also cited Shanghai premiums of about $24 for gold and 13.3% for silver.

The mainstream reading is plainer. Large orders hitting a thin book can trigger stop-losses and algorithmic cascades, especially around payrolls and quarter-end. Bank strategists also point to miner and refiner hedging and to market makers widening spreads under stress.

What remains unverified matters most. No independent source confirmed those Shanghai premiums, and nobody can see the intent behind the selling.

A sharp futures drop alone does not tell you physical demand weakened. Check premiums, ETF flows and positioning before you treat it as a change in fundamentals, and remember that leveraged holders hit by margin calls often exit at the worst prices.

What can Commitments of Traders data and open interest tell you, and what can’t they?

The CFTC’s weekly Commitments of Traders (COT) report is the best free window you have into futures positioning. It splits traders into commercials (producers, merchants and dealers hedging business risk) and non-commercials (speculators), while disaggregated versions isolate “managed money”, meaning funds and trading advisers.

In October 2025, Maguire said managed money shorts rose about 7,000 contracts and gold open interest fell nearly 11,000 contracts. Comparable October 2025 figures could not be found, so no like-for-like change against today is claimed here.

Data as of 29 September 2026. A Japanese market commentary puts the speculative net long in gold at 218,632 contracts, down 7,221 on the week. Speculators remain heavily long, and still adding to the exit queue if sentiment turns.

Now the limits:

1. Scope: only regulated futures and options appear, so OTC hedges and swaps stay invisible.
2. Categorisation: positions are aggregated, so you cannot infer motive; a long might be an inflation hedge or a trend trade.
3. Timing: data reflects Tuesday positions released on Friday, a lag that goes stale in fast markets.

Maguire calls the delay deliberate and says the OTC omissions hide offsetting hedges. Regulators treat both as standard release mechanics. Either way, use COT to gauge crowding and direction, not to decide who is winning.

For readers wanting to go further on positioning, our full explainer on gold futures speculator positioning shows what happens to gold’s volatility when the usual speculative players go quiet.

Why is gold decoupling from yields and the dollar, and where do ETFs fit?

Here is the puzzle. Gold usually falls when yields and the dollar rise, because it pays no income and is priced in dollars, yet in late 2025 all three held firm together.

Maguire offered two comparisons. In 2002, with yields at similar levels, gold sat near $300-311 and the dollar index above 120. In 2022, a dollar peak near 114 preceded gold’s break above $2,000. These illustrate his argument; they do not prove it.

Analysts cite four explanations: steady central-bank buying tied to diversification away from the dollar, sovereign-debt and systemic-risk worries, ETF flows that persist when prices weaken, and an open debate over whether the shift is lasting.

On flows, Maguire said managed money cut net longs by about 27,000 contracts in September 2025 while ETFs added about 1.7 million oz, and that 46 of 121 tons of vanished open interest moved into funds. Treat his 3,145-ton ETF figure from October 2025 carefully: it is defined differently from the WGC series below and cannot be compared directly.

September’s mix of record tonnage and falling value tells you investors were adding metal into weakness. Watch tonnage separately from dollar value, because the two can send opposite signals.

Structural regime or cyclical quirk?

The structural camp says gold now trades mainly as a neutral reserve asset, so yields and the dollar matter less. The cyclical camp sees temporary forces, such as positioning extremes and policy uncertainty, and expects the old correlations to return.

Continued official buying through a period of easing crisis would favour the structural view. A renewed tight inverse link to real yields would favour the cyclical one.

Is Asia’s physical infrastructure really changing price discovery?

This is where the thesis gets most ambitious, and where the evidence gets thinnest.

Single-source caution The claims in this section rest on Andrew Maguire’s account alone. Independent searches found no corroboration.

Maguire argues the Shanghai Gold Exchange (SGE) is drawing interest for physically settled protocols. He links this to People’s Bank of China storage corridors and a Hong Kong gold-backed renminbi initiative he expected to open broadly in Q1 2026. He says the London Bullion Market Association (LBMA) has about $1 billion in resources against a possible $4 billion lawsuit, that only 3-5 tons are physically deliverable against 600-800 tons of daily cash settlement, and that about 400 bars have flowed to Hong Kong and the SGE.

The concept itself is sound. A price set against deliverable metal behaves differently from one set against cash-settled claims, because a seller must hand over bars. One verifiable change points that way: Basel 3 net stable funding rules made LBMA FX gold fixes physically deliverable from January 2023.

A deliverable-metal price behaves differently from a cash-settled one, which is why the East-West pricing divide between COMEX, the LBMA and the Shanghai Gold Exchange draws so much scrutiny.

On silver, Maguire cited a Goldman Sachs report estimating physical-demand price impact 8-10 times larger than paper estimates, and floated $140 silver by Q4 2025. That date has passed, the note was not accessible, and no evidence was found of a 2025 London silver squeeze or tariff-driven US stockpiling. These statements are speculative; past forecasts do not guarantee future results.

Evidence to watch:

• Published SGE and Hong Kong premiums over London spot
• Disclosed LBMA and Asian vault tonnages
• Formal SGE or LBMA settlement rule changes or filed litigation

Until those appear, treat this as a hypothesis to monitor, not a proven driver of your holdings’ price.

Putting the framework to work: how to read paper-versus-physical signals

When the next sharp move hits, run it through four checks:

1. Timing and liquidity: did it land at an open, a holiday or a data release, when books are thin?
2. COT and open interest: did speculators cut longs or add shorts, and is positioning crowded?
3. ETF tonnage versus AUM: are funds adding metal even as value falls?
4. Asian physical indicators: are published premiums rising or fading?

Know the risks attached to how you hold exposure:

• Leverage: margin calls in a flash move can force sales at poor prices, and exchanges can raise margins under stress.
• ETF structure: custodians, vaults and authorised participants all sit between you and the metal, and fees create small gaps from spot.
• Counterparty and operational risk: clearing reduces credit risk, but settlement disputes or rare defaults can still bite.

Gold ETF risks sit mostly in the chain between you and the bars, from custodians and vault arrangements to authorised participants, so it pays to know who holds the metal behind your units.

Keep the strongest counter-argument in view. Futures are binding contracts inside the same market, and the 2020 spread showed that arbitrage tends to close gaps.

A signal confirmed by several independent indicators deserves more weight than any single dramatic headline. Size your reaction accordingly.

What the evidence supports, and what remains open

Some of this debate is settled by data. WGC figures confirm record ETF tonnage of 4,256 tonnes, and CFTC figures confirm heavy speculative COMEX positioning. The China, LBMA and Goldman claims remain single-source.

The value of this framework lies in cross-checking, not in picking a side. Three releases give you the next read: the weekly COT report each Friday, the WGC’s monthly ETF figures, and any published SGE or LBMA settlement changes. If physical and paper signals line up across them, take notice; if only one source says so, wait.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Past performance does not guarantee future results. Financial projections are subject to market conditions and various risk factors.