Low inventories do not prove stronger demand

Low inventories do not prove stronger demand

Decision in brief

Research cutoff: October 15, 2026, 18:15 Toronto / 22:15 UTC. This review compares with 16:15 UTC. The four-month horizon remains October 13–February 13, 2027; each contract's modeled payoff date is its intended exit, not automatically the end of that horizon.

Exit one December Micro Nasdaq; hold one March micro copper; add nothing. The Nasdaq rebound has consumed most of the recovery trade's forward value. Copper retains a small positive retention case. The earlier SOFR and oil/meal exits remain closed; their historical reasons are not rewritten.

The strongest physical-market insight is that weak soybean processing can reduce both oil and meal output. Low oil inventories are therefore not independent proof of booming demand, nor sufficient evidence for buying oil against a meal short. We test a simpler micro-oil long as well as the pair below.

What changed, and what was only newly verified

Soybean processing: less output is not stronger demand

Reported September NOPA crush was 196.3 million bushels, below the Reuters survey's 203.737 million. Reported oil stocks were 1.053 billion pounds, versus 1.116 billion expected, and the lowest since November 2014. The actual figures are attributed to the distributor's report; the raw subscription-distributed NOPA release was not downloaded. ADM Investor Services, October 15 Reuters survey NOPA distribution

These results were newly retrieved for this review. The result headline existed around 16:01 UTC, before the prior cutoff; they are not presented as a new evening release or as information that motivated the earlier paper exit.

The crush shortfall was 7.437 million bushels, or 3.65%. CME's conventional conversion of 11 pounds of oil per bushel translates that processing shortfall into roughly 81.8 million pounds of oil, compared with a 63 million-pound inventory shortfall versus expectations. This is an illustrative scale check, not actual September yield, a counterfactual forecast, or a complete supply-demand balance. It demonstrates why the inventory number alone cannot identify stronger consumption. Lower processing also reduces meal output. CME processing mechanics

Petroleum: supply composition and policy response matter

The October 15 EIA report covers the week ended October 9. Crude stocks fell 8.557 million barrels, but production fell 300,000 barrels/day, imports fell 791,000, exports also fell 195,000, and refinery inputs rose 135,000. Four-week gasoline and distillate supplied remained 0.4% and 0.2% below year-earlier levels. The fuller breakdown is newly examined here, not asserted to have first appeared after noon. The draw is not evidence of an export or broad product-demand boom. EIA overview Release data

The IEA's October 15 statement accelerates roughly 100 million barrels remaining under the earlier action, including more than 20 million barrels of diesel. About 330 million had already been released; total deliveries will exceed the original 400-million pledge because some members exceeded or will exceed commitments. Do not add the diesel portion again or describe all remaining barrels as a new extra package. Faster delivery challenges extreme product-scarcity payoffs, although timing and refinery constraints remain material. This is newly verified policy context, not necessarily first published after 16:15. IEA statement

Bonds rallied despite inflation; equities did not regain clear AI leadership

Treasury's October 15 indicative end-day curve, sampled near 19:30 UTC, puts two-year yields at 4.77%, five-year at 4.91%, ten-year at 5.10%, and thirty-year at 5.43%. The ten-year fell nine basis points from October 14. Ten-year real yields fell six basis points to 2.78%. The nominal-minus-real spread narrowed three basis points; it is not pure expected inflation. These are fresh closing measurements, not proof that the entire daily move happened after our prior cutoff. Treasury nominal curve Real curve

Lower real discount rates support long-duration assets, but a safety bid is not a monetary pivot. Reuters' close reports the Dow +1.08%, S&P 500 +0.33%, and Nasdaq Composite +0.08%, with semiconductor losses continuing. This is rotation, not unambiguous restoration of AI leadership. The Composite is not the Nasdaq-100 underlying our futures position. Reuters close

Earlier TSMC results and guidance still support supplier demand. They predate this comparison window. Strong chip sales do not establish satisfactory returns on customers' debt-funded investment. Later reporting on AI valuations corroborated concerns already reported earlier; it is not assigned as a fresh causal shock. Morning PPI and gasoline-led nominal retail growth remain incorporated context from the noon review. TSMC official results

Gas: warm forecasts, but regional vulnerability

The 19:00 UTC October 15 CPC update favors exceptional warmth in parts of the Upper Midwest for October 21–25; week-two confidence is lower. This is a forecast, not measured demand. EIA national stocks of 3,567 Bcf are 45 Bcf above the five-year average, yet South Central stocks are 3.1% below normal and salt storage 15.4% below. Warmth weakens a December long; the regional deficit limits confidence in a winter short. CPC discussion EIA storage, week October 9

Current reference sheet

Exact native IBKR last ticks below were retrieved at 22:07 UTC on October 15. Bid/ask studies are separately retrieved indications without independent exchange timestamps. A last trade need not equal the current book.

Interactive Brokers native contract data. No value is described as an observed execution.

For March WTI, ULSD and agricultural contracts, the provider returned vendor-close values without a date or clock. Those are not promoted to October 15 settlements. Current study indications retrieved 22:07:28–40 UTC were: CLH27 bid/ask 90.62/90.71; HOH27 4.2813/4.2970; ZCH27 491.50/491.50 cents; ZWH27 685.75/685.75 cents; ZLH27 67.55/67.55 cents/lb; ZMH27 355.80/355.80 dollars/short ton. Locked books and missing independent clocks reduce confidence; they do not automatically prohibit a disclosed assumed-price study. No trade is authorized in these markets on today's economics.

Micro studies use identical-unit, same-month standard references with an explicit equal-basis assumption; these are not observed micro prices. The soybean-oil input is conservatively rounded up to the valid 67.56 micro tick.

Position decisions at current opportunity cost

Scenario probabilities are subjective, not calibrated frequencies. Outcomes are unmanaged terminal comparisons at the intended exit date, not stopped backtests. Retention subtracts remaining exit friction; new-entry comparisons subtract full round-trip friction. Buffered risks apply 1.25 to the adverse price move, then add costs. Margin reserves are separate.

December Micro Nasdaq — Exit one, no reverse

Keep the November 20 endpoints and weights 25/35/25/15 unchanged:

Multiplier is $2/point and remaining friction $4. The rebound from the prior 29,994.75 mark consumed $249.50 of expected remaining value. A fresh long at the retrieved 30,120.25 ask offers +$126.50 after $8 full friction.

Five probability points transferred from favorable to gap produce −$233 retention EV. Lowering favorable and base endpoints 250 points gives −$168; combining consistently gives −$508. Positive central EV is too fragile against unresolved customer-financing risk. Lower real yields and supplier strength are real counterarguments, but do not make that compensation robust.

Ordinary forward buffered risk to 29,600 is $1,306.75; the 28,000 stress is $5,306.75. The latter describes current opportunity-cost stress, not a claim that the ledger's formal entry/stop limit was breached. Historical entry-based risk differs. We do not move the tail or widen the stop simply to retain one indivisible unit. Exit economically; no unobserved stop touch is asserted. A reverse short has negative value under the same distribution.

March micro copper — Hold one, add zero

Latest retrievable LME physical context remains October 14: stocks 228,000 tonnes, cash $14,751 versus three-month $14,635, a $116/tonne premium. Prompt physical tightness supports retention, but LME availability, U.S. stock concentration and COMEX deferred pricing are not interchangeable. No October 15 physical row is invented. Earlier mine developments remain incorporated. Dated physical table

Keep February 12 endpoints 7.60/7.00/5.80, weighted 25/45/30. At 6.6705, a 2,500-lb contract and $4.75 exit friction produce +$2,319 / +$819 / −$2,181, weighted +$294. Ordinary remaining risk to 6.35 is $1,006.31.

One additional unit at 6.676 ask would have +$275.50 EV after $9.50 full costs, with ordinary risk $1,028.25 and adverse-endpoint risk $2,747. But a five-point favorable-to-adverse shift plus five cents lower base gives −$5.75. Two units would expose roughly $5,494 to that adverse endpoint, above the approximately $5,017 current thesis allowance. Retain one; do not add.

Keep 6.35 invalidation, 7.00 review, 7.60 target, exit February 12. March's 10.35-cent premium to December is not LME backwardation and not free carry. Lower refinery/industrial demand or easing accessible-metal scarcity can invalidate the thesis before the price stop.

Re-underwriting closed ideas and fresh challengers

Oil/meal pair, then a simpler oil-only expression

Reduce favorable probability from 40% to 35%, increasing adverse from 25% to 30%, with base unchanged at 35%. Weak processing does not independently validate abundant meal output. Keep February 12 joint oil/meal endpoints 70.50/336, 68/360, 64.56/382; micro pair values are $870/$480/$53.60.

At conservative indicative inputs 67.56/355.80, current pair value is 60×67.56−10×355.80=$495.60. Net outcomes after $20.40 full pair friction are +$354/−$36/−$462.40; weighted EV −$27.42 per pair, or −$82.26 for three. One pair's $572.90 buffered adverse risk fits; value does not. Stay closed.

A simpler one March micro soybean-oil long, 6,000 lb, removes the meal-short thesis. At 67.56 with oil endpoints 70.50/68/64.56 and 40/35/25 weights, EV is +$25.40 after $9.40 friction, ordinary risk $234.40. Five favorable-to-adverse points leave only +$7.58; doubling friction as well gives −$1.82. It fits as a tiny unit but lacks robust economic compensation. Low stock headlines alone do not justify scaling it to reach a customary starter budget.

Cross-asset comparison

CME micro agricultural specifications Micro WTI Micro gas

The diesel change transfers five points from favorable to adverse because accelerated emergency supplies reduce the extreme-scarcity case. The 25/40/30/5 joint diesel/crude endpoints for a February 10 crack exit are 5.50/110,4.60/94,3.80/80,3.00/65. Matching 4,200 gallons to100 barrels gives current crack 42×4.297−90.62=$89.854/barrel. Weighted terminal crack96.86 yields +$660.60 after$40 costs. Ordinary risk to79.6 is$1,321.75. But divergent diesel3.80/crude110 produces49.6 and $5,071.75 buffered loss, still above the current thesis allowance. Reject this package too: barrel matching does not eliminate basis risk.

For a fresh December Treasury comparison, not merely the retired short's invalid stop, use an explicitly illustrative November20 distribution 107.75/105.75/104/102, weighted25/35/25/15. Mean105.25 balances safety buying with inflation/term premium. A short at105.578125 bid offers +$296.13 after$32, but shifting five points from102 to107.75 leaves +$8.63. Buffered risk to107.75 is $2,746.84; a long at105.59375 ask loses$375.75 expected. Neither wins. Smaller yield futures could reduce unit risk but introduce a different payoff and do not repair this fragile policy edge.

Euro short: Reuters' article published20:04 UTC reports spot about1.12, without a separate quote clock and only two-decimal precision. Using1.1200 solely as a model representation—not an observed four-decimal tick—and zero December basis, three €12,500 micro equivalents with25/35/25/15 November20 endpoints1.07/1.11/1.135/1.16 produce +$168.38 after$66 costs. Ordinary risk to1.135 is$769.13. A five-point favorable-to-gap shift gives−$0.38; an adverse0.005 basis/rounded-reference adjustment alone gives−$19.13; combined−$187.88. Do not infer a verified0.0010 decline from the earlier1.1210 report. Wait for Warsh/Friday relative-policy reassessment: the edge is not robust to economically material policy and basis uncertainty. Reuters via MarketScreener

Curve, positioning and the four-month view

Retain the noon regime map 42/19/18/18/3: persistent inflation/policy divergence, resilient normalization, renewed supply shock, growth break, AI/liquidity disappointment. Do not translate these qualitative regime judgments into contract probabilities mechanically.

March/December SOFR prices imply funding rates4.515%/4.255%, a26-basis-point difference across distinct quarters. March/December copper and gas curves reflect their own storage, location and seasonal economics. Indicative March oil/product books are not synchronized enough to claim an observed curve change; the crack above is an explicitly assumed comparison, not an executed spread.

October6 CFTC positioning remains dated weekly context. Friday's report will describe October13 positions, before CPI and PPI, so it cannot establish current post-release flows. No fresh liquidation or crowding totals are fabricated. CFTC schedule

The earlier CONAB production forecast remains incorporated: more Brazilian corn is not automatically bearish within February because domestic use rises faster and much second-crop supply arrives later. Legacy September equity and INR policy ideas remain context-only, not orders. Retired gas-calendar and OAT/Bund expressions have no new compelling seasonal spread or economically validated DV01/cheapest-to-deliver construction. This ordinary review retains the full October 4 monthly scan and latest weekly review as background without claiming a new worldwide monthly census.

Next 72 hours and lifecycle

Federal Reserve calendar BLS schedule USDA holiday release notice

Proceed with the one copper unit through these events; no automatic blackout or additional risk. Nasdaq is exited now, reducing actual equity risk. Copper's intended February12 exit precedes its February24 last-trade date. No automatic roll is assumed; the four-month thesis does not authorize carrying into expiry. Closed positions have no current lifecycle exposure.

Paper execution and results

The economic decision is to close Nasdaq and retain copper. The accounting below is an assumed paper execution, not a witnessed exchange fill. The exact MNQZ26 reference is30,119.50 observed October15 22:07:19 UTC; one adverse0.25-point tick implies a sale at 30,119.25. From the recorded29,838.75 entry, gross profit is$561 and completed-trade net after both commissions is $554. The remaining copper mark is6.6705, observed22:04:25 UTC, versus6.6665 entry.

The assumed exit was recorded at 22:15:13 UTC on October 15. Simulated NAV is $1,003,402.48, cash $1,003,392.48, cumulative completed-trade net profit $3,395.98, and open gross copper profit $10.00. One position and no pending orders remain. Ledger open risk is $997.31, about 0.099% NAV; this entry-based ledger measure differs from forward scenario comparisons. No capital reset, historical rewrite, invented trigger path or real order is used. The remaining copper reserve is a disclosed $6,000 paper margin assumption, above twice the previously sourced generic maintenance amount, not an exact exchange initial-margin quotation.

*Research and simulation only; not investment advice. Futures may lose more than margin, and stops do not cap gaps or limit moves.*