Copper tightened. Its trade became less attractive.

Copper tightened. Its trade became less attractive.

Research cutoff: 16 October 2026, 04:15 UTC / 00:15 Toronto. Changes are measured against 15 October 22:15 UTC. The four-month thesis horizon remains 13 October–13 February; each contract's payoff date and exit plan are shorter where necessary.

Decision in brief

Exit the one-contract March micro-copper paper starter; do not reverse. Prompt physical scarcity remains credible. But March futures have rebounded enough to consume most of the remaining modeled compensation. A small probability change now makes holding unattractive.

The best new challenger is long micro diesel against short micro WTI. Its lower entry spread means one whole package now fits the risk allowance. The previous size rejection is obsolete. Nevertheless, a modest shift toward U.S. product/crude divergence, together with higher costs, erases expected value. Wait for better New York Harbor-relative evidence or a materially lower spread, not broker-grade price precision.

No replacement trade is forced. Relative equity, smaller gold and AUD/NZD expressions were investigated alongside every existing board thesis. Current affirmative research authorizes the copper exit; other decisions below are economic waits or avoids.

What changed—and what did not

Warsh's completed discussion did not provide an October policy signal. It should no longer sit in the calendar as an upcoming catalyst, and silence is not a dovish pivot. Goolsbee's 23:48 UTC report distinguished potentially reversible supply inflation from AI-driven excess demand. Our inference: AI financing stress can support bonds and weaken the dollar even while AI demand sustains inflation. That makes a simple inflation-to-short-bonds trade unreliable. Sources: Warsh coverage, Goolsbee, Reuters, 15 October 23:48 UTC.

Reuters' 16 October 01:20 UTC currency report indicated EUR/USD 1.1210, AUD/USD 0.6921 and NZD/USD 0.5621 as U.S. financing concerns weighed on the dollar. These are news indications, not exact futures quotes. The report predates Warsh's discussion and is not evidence of a market reaction to his remarks.

The newly retrieved London copper row is dated 15 October, not an overnight physical observation. SMM's morning assessment was published 16 October 01:01 UTC. The IEA's October report, yesterday's emergency-stock news, NOPA processing figures and the latest retrieved U.S. weather outlook remain dated context. No repeated headline is counted as a new shock.

Exact contracts and observation clocks

All last-price observations below are 16 October UTC, obtained from IBKR native contract snapshots. They are intraday ticks, not settlements; quote books can differ from the last trade. Different contracts were not sampled synchronously. Analytical entry tests use the conservative retrieved ask for longs and bid for shorts; assumed paper execution uses the documented nearest-reference convention.

Units: gold USD/oz; copper USD/lb; WTI USD/barrel; ULSD USD/gallon; gas USD/MMBtu; corn/wheat cents/bushel; soybean oil cents/lb; meal USD/short ton; equity and Treasury futures index/price points; SOFR 100 minus the contract's annualized reference rate. These are the exact named standard or micro instruments—not invented quotes for equivalent micro products.

Copper: a physical thesis is not a permanent holding instruction

Westmetall's LME table records 15 October cash copper at $14,539/t and three-month at $14,362/t: a $177/t cash premium, versus $116 on 14 October. Stocks fell from 228,000 to 226,600 tonnes. That supports immediate-delivery scarcity. Cash price itself fell $212/t, so a widening premium is not equivalent to a uniform price rally.

SMM's 16 October morning report describes muted downstream buying and weaker Guangdong premiums. Increased SHFE warrant registration around delivery is not necessarily newly produced metal; Shanghai premiums also reflect rollover and cargo availability. This limits the case for upgrading distant COMEX demand assumptions.

March micro copper rose from the previous 6.6705 observation to 6.7460. March/December COMEX premium is now 10.60 cents/lb, versus 10.35 cents previously. This is an upward-sloping COMEX interval, distinct from LME prompt backwardation. Neither is free carry: financing, location, delivery and tenor differ.

Keep the February 12 distribution unchanged:

These are endpoint outcomes, not claims that a stop prevents gap losses. Remaining payoffs use 2,500 lb and $4.75 prospective exit friction; historical entry costs are sunk. Weighted mean is 6.79 and remaining expected value +$105.25, down $188.75 from the previous +$294. A five-percentage-point transfer from favorable to adverse produces −$119.75. Cutting the base endpoint five cents alone leaves +$49; combining both gives −$176.

Buffered risk to 6.35 is $1,242.25; the 5.80 scenario is $2,961. Neither breaches the thesis cap. This is an economic exit, not a forced stop, invented intraday touch or assertion that the physical thesis collapsed. A new entry at ask 6.747 has only +$98 EV after full costs. Re-entry requires freshly justified value—roughly a lower price around 6.63 under this distribution or a defensible increase in endpoints—not an automatic standing order.

Paper implementation: close one MHGH27 at an assumed 6.7455, based on exact-contract last 6.7460 observed 04:02:40 UTC, with one adverse 0.0005 tick. Execution is assumed at the actual decision/completion time, not backdated to the reference. The completed trade's net result is recorded below. No real order is placed.

Diesel versus crude: a better expression, but not yet robust

The IEA report published 14 October shows Middle Eastern crude exports recovering faster than refined products. Its outlook keeps diesel inventories impaired into 2027 despite a prospective first-quarter surplus. This supports product-relative exposure more directly than outright oil. Counterevidence is demand destruction, refinery recovery and accelerated emergency supply; global tightness need not translate into New York Harbor outperformance.

Use March HO ask 4.2627 and CL bid 90.52 as explicitly compatible references for one 4,200-gallon MHO long / one 100-barrel MCL short. Equal-month basis is assumed; these are not observed micro quotes. The crack—diesel's barrel-equivalent price minus crude—is 42 × 4.2627 − 90.52 = $88.5134/barrel.

One package risks $1,154.18 to crack 79.60 after the 1.25 buffer and $40 cost reserve. A divergent 49.60 scenario risks $4,904.18, inside the approximately $5,017 current thesis allowance. Size now fits. The $40 allowance includes $16.84 ordinary two-leg commissions/tick friction plus an explicit extra $23.16 reserve; no spread-margin credit is assumed.

The last state includes product-specific easing or export restrictions alongside firm crude. Ten percent is a scenario judgment, not an estimated probability of an export ban, and no ban implementation is claimed. The older four-state model gives +$794.66 EV; including this previously unweighted location risk gives +$396.66.

Five points shifted from severe scarcity to divergence reduce EV to +$39.66; doubling the $40 reserve gives −$0.34. A further 20-cent reduction in the diesel base endpoint makes the combined test −$254.34. This is not robust enough to open now. The combined stressed mean crack is 86.37; after doubled costs, breakeven entry is 85.57. A level near 85.5 merits reassessment, not a promise to buy.

CME's MHO rules specify 4,200 gallons and cash settlement linked to the corresponding standard contract. Native contract calendars show March MCL last trade February 19 and MHO February 25; a February 10 exit avoids both. A prospective $12,000 combined paper reserve would be a conservative assumption, not an exchange initial-margin claim. No accounting or quote precision issue is being used to veto this trade.

Re-underwriting the rest of the board

Expected values below are scenario judgments after stated round-trip costs, not forecasts with statistically estimated confidence. One-unit risk is checked before scaling. Unless specified otherwise, outcomes are evaluated by February 12; December equities/Treasuries use November 20 and December oil/gas November 12.

SOFR March/December imply a 26-basis-point difference across distinct accrual quarters, not an exact number of policy moves. WTI December/March last difference is $5.29/barrel; gas December/March $0.798/MMBtu. Seasonal and location differences prevent treating these as guaranteed roll profits.

Gas retains the 15 October EIA storage evidence: national inventory 45 Bcf above the five-year average, but regional South Central/salt deficits. The latest retrieved CPC outlook, issued 15 October 19:00 UTC, favors warmth. No fresh weather revision was verified. September NOPA crush of 196.3 million bushels and oil stocks of 1.053 billion pounds are attributed to ADM Investor Services' October 15 report, not a newly downloaded subscription release. Weak processing simultaneously constrains oil and meal supply; the previously closed pair is not reopened just because oil inventories are low.

Fresh alternatives, not recycled waits

Long broad equities / short Nasdaq. A November 20 joint study uses ES/NQ endpoints 8100/31650 (30%), 7900/30000 (35%), 7350/28000 (20%), 7650/29600 (15%). These represent renewed AI leadership, rotation, financing stress and restrictive inflation. Long 3 MES/short 2 MNQ is approximately dollar-notional matched—not beta-neutral. At 7801.75/30191.50 its net outcomes after $44.50 are −$1,404.75/+$2,195.25/+$1,945.25/+$45.25, EV +$742.75. But ES 8000 / NQ 32500 divergence risks $7,869.81, above the thesis cap.

The smaller 1 MES/1 MNQ package has +$349.25 EV and $4,549.69 divergence risk, yet it is materially net-short equity. Five points shifted from financing shock to AI leadership plus 250 higher Nasdaq points in leadership/rotation make EV −$153.25. Wait: resizing changes the thesis and leaves fragile compensation.

AUD/NZD short. Newly examined—not newly released—ABS 15 October employment data showed 39,000 fewer jobs and 4.7% unemployment. Hours and trend employment each rose 0.1%; ABS warns of residual survey-method effects. This is mixed evidence for a relative-policy trade, not proof of an Australian recession.

The RBA's 4.60% policy rate exceeds the RBNZ's 2.75%. At indicated cross 1.23128, an intentionally favorable challenger distribution 50/25/25 at 1.19/1.15/1.28 offers about 2.34% spot expectation before roughly 0.6% four-month negative carry and fees. Conditional New Zealand recovery and noisy Australian data are counterarguments.

CME's micro suite supports Australian-dollar micros but no equivalent NZD micro. Short 8 M6A/long 1 standard 6N is a near-dollar-matched whole-contract study. Holding NZD/USD 0.5621 constant for sizing, cross 1.28 risks $3,161.75 including buffer, $89 costs and approximately $334 carry; 1.33 risks $5,972.25. Both relevant allowances are exceeded. Wait; outright AUD/USD cannot inherit the cross's distribution.

Smaller gold. Seven April 1-ounce gold contracts, short at a conservative 4,274.00 same-month proxy with $52.50 friction, give +$115.50 EV. Risks $2,030 to 4500 and $2,905 to 4600 fit. But the same five-point gold sensitivity yields −$122.50. Wait for economic value, not because standard contracts are too big.

Positioning, catalysts and lifecycle

No new verified CFTC positioning release occurred within this window. Price, volume or delivery-warrant changes are not substitutes for an identified investor-position series. Friday's report describes Tuesday 13 October positions and cannot measure subsequent inflation-driven flows.

Asia is trading; European cash markets have not opened at the cutoff. U.S. futures trade overnight, while the latest U.S. cash session is 15 October. The 72-hour window ends 04:15 UTC Monday; later Monday releases are outside it.

All present positions are closed after implementation. Prospective March copper would exit February 12 before February 24 last trade; March diesel/WTI exits February 10 before February 19/25 bounds. December equity/rates studies exit November 20; energy studies leave earlier as above. Agricultural and April gold studies exit February 12 before applicable delivery/expiry restrictions. A four-month economic thesis does not authorize holding an expiring contract or an automatic roll.

Simulated book and limitations

The assumed copper exit completed at 16 October 04:17:06.967 UTC. (Exit 6.7455 − entry 6.6665) × 2,500 lb yields $197.50 gross, less $7 combined commissions = +$190.50 net. Slippage is already embedded in the prices, not deducted twice.

The actual ledger now shows NAV and cash $1,003,586.475, cumulative closed-trade net +$3,586.475, twelve completed positions, zero open positions and zero pending orders. Valuation time remains 04:17:06.967 UTC; it is not relabeled as the research cutoff. No capital reset or historical return rewrite.

Risk conventions remain unchanged: whole contracts, 1.25 adverse-scenario buffer, normal starters 0.10–0.25% NAV, 0.50% total thesis cap, 4% aggregate and 2% common-theme limits. Margin is reserved, not spent; no assumed spread credit. Closing copper genuinely removes exposure rather than merely changing a label. Its physical-demand watch continues without a position.

Forecast endpoints, weights, proxy basis and conservative cost/margin reserves are analytical assumptions. They are not exchange facts or calibrated probabilities. The price table preserves actual observation dates; assumed fills are simulated decisions now. Research is educational, not a real-order instruction.