Futures vs physical, across exchanges, over time

Web-verified 2026-06-08. Structured + edges + sources: macro-futures-vs-physical.json. Source discipline: much commodity commentary is perma-bull/manipulation-themed (SchiffGold, Sprott, DiscoveryAlert, 24/7) — used only for market-structure facts checkable against exchange/CFTC data. Documented enforcement is fact; blanket "suppression" is graded contested. Overlay, not used in the proofs.

The idea: price is made in paper, the good is delivered in physical

The headline commodity price is set in leveraged futures on one venue; the deliverable good trades on others (LBMA, LME, SGE/SHFE, DGCX, physical spot) at different prices. The divergence tells you where paper ≠ physical:

• Backwardation (spot > futures) — you pay up for metal now vs a promise later.
• Cross-exchange premia — the "same" asset priced differently in NY / London / Shanghai; arbitrage that won't close means metal can't move.
• Registered vs open interest — how many paper ounces claim each deliverable ounce.
• Lease rates — the cost to borrow physical.

Gold & silver (2025–26) — structure facts strong; suppression contested

• Silver backwardation — reported first sustained instance since 1980 (Dec-2025 ≈ $47.62 vs spot ≈ $47.67).
• LBMA–COMEX premium blew out to >$2.50 (usually cents); London interbank physical seized, lease rates spiked.
• COMEX futures-to-registered ≈ 4.2 paper claims per deliverable oz; much "eligible" metal is allocated (ETFs/industrial/banks), not freely deliverable.
• Shanghai (SGE) ≈ 12–13% physical premium over Western futures (Q4 2025+) — the East pulls physical.
• Earlier 2025: a London→COMEX gold airlift on US tariff/positioning fears.
• Documented manipulation (fact): JPMorgan paid ~$920M (Sept 2020, CFTC/DOJ) for spoofing precious-metals + Treasury futures. This is the factual anchor — manipulation occurred — but it was short-horizon spoofing, not proof of a decades-long price cap.

Copper (2025) — strong (exchange + policy record)

• July 30 2025: a US 50% tariff on copper semis (cathode/concentrate exempt). Before the carve-out was known, the COMEX–LME spread widened to >28%, driving cross-exchange arbitrage and US pre-positioning.
• On the announcement, COMEX copper fell >18% in a day — a record — as the premium collapsed.
• Venue split: CME in contango (tariff distortion) while LME ran backwardation in nearby months; SHFE arbitrage transmits into LME–SHFE spreads.
• Lesson: one policy action blew a >28% wedge between two exchanges pricing the same metal — the "price" is venue- and policy-contingent, not a single physical truth.

Oil (2025–26) — strong (ICE/CME curve data)

Three benchmarks price "oil" differently: WTI (US light-sweet), Brent (waterborne, slight premium to WTI), Dubai (Asia sour) — physical differentials (sulfur/density/freight) mean the futures benchmark isn't the barrel a refiner buys.

• Brent in pronounced backwardation: front ~$94.22 (Jun 5 2026) → ~$85.79 (Dec 2026) → ~$77.68 (Dec 2027); WTI also backwardated amid a supply crunch.
• But the structure whipsawed: Brent's 6-month calendar spread collapsed to 41¢ backwardation in Nov 2025 (from >$2 on Oct 24) and briefly went contango — the lowest in ~2 years; May 2025 curves flattened into contango.
• Steep backwardation makes storage uneconomic — so the curve shape is the live tightness signal, and in 2025–26 it flipped between scarcity and glut within weeks.
• Dubai / sour (Asia marker): the Brent–Dubai EFS narrowed sharply in 2025 and flipped negative (−11¢/bbl, Aug 27 2025), inverting Brent's usual premium and pulling Atlantic crude to Asia; it's expected to widen again in 2026 as OPEC+ sour supply dominates and sour refining lags. The price is a curve and a set of quality/region spreads — not a number — and both whipsawed on OPEC+ flows and refining capacity.

Zero-trust conclusion

Treat the futures print like a BLS headline — the most tradable, not necessarily the most physical, number — and cross-check it against backwardation, cross-exchange premia, registered/eligible stocks, lease rates, and the Eastern physical bid. This complements the gold lens (macro-gold-silver-reprice): not only are nominal gains debasement, the metal price itself is a paper construct that physical periodically overrides.

← Research index · structured data: macro-futures-vs-physical.json · macro-futures-vs-physical.md