No action after the weekly review

ATG Sunday evening futures research — September 13, 2026

Decision: no action. Retain $1,000,000 cash and all existing board statuses; take no paper action. The simulated paper portfolio remains unchanged: 0 positions, 0 orders, $0 P&L, $0 costs. No futures-price change is booked as a portfolio return. No verified post-baseline development meets every thesis, executable-market, evidence, cost, lifecycle, margin and portfolio gate.

Comparison baseline: complete weekly review 2026-09-13T23:15:00Z (September 13, 19:15 Toronto/EDT). Consolidated research cutoff: 2026-09-14T00:05:46Z (September 13, 20:05:46 Toronto), a 50m46s delta window. Commodity coverage ends 00:04:07Z; macro/FX ends 00:05:45Z; regime coverage ends 00:05:46Z. These are separate evidence cutoffs, not synchronized market timestamps. Thesis horizon remains 2026-09-13 through 2027-01-13; this monitor neither rolls that horizon forward nor resets dated scenario probabilities, model dates or prior full-review judgments. Inherited ER and priced-in scores below remain dated baseline outputs, not new forecasts.

What actually changed

• Timestamp-qualified, delayed exact-contract observations refreshed CLG27, HGH27, 6JZ26, 6EZ26 and both SOFR legs. None establishes the required entry condition or a new invalidation. Commodity mark differences partly span time before the review cutoff because the stored comparison observations predate it. A refreshed quote is not proof that its entire difference occurred after baseline.

• The only newly timestamped global-regime publication verified in the supplied research was Reuters' 2026-09-13T23:25:08Z report that Genta Koja won Okinawa's gubernatorial election with 59%. It adds strategic East Asia context, not a demonstrated broad-futures regime break or actionable contract signal. Publication time is not necessarily event time. Reuters

• No timestamp-qualified new commodity physical-flow, inventory, certification, central-bank announcement or positioning release was established after baseline. This is the result of the bounded research, not proof that no news occurred. No material new opportunity or rank/eligibility promotion was verified.

Overlap-only news: relevant, already known, not counted twice

The September 10 ECB hike, September 12 Lagarde interview, September 11 French fiscal report and September 8 CFTC holdings are inherited context. The Sunday yen-positioning article has no verified exact publication time in this evidence set, so its novelty is unclassified, not post-baseline. Primary CFTC data corroborate the holdings but not new Sunday flows. ECB decision, ECB interview, French fiscal report, CFTC

Monitoring marks and unavailable-data treatment

All observations below are non-executable. Last, settlement, source-observation time, retrieval time and review cutoff remain distinct. USD/JPY-unit yen futures are not spot USD/JPY and not the vendor's 100-times-scaled display. Do not infer fills, current bid/ask, depth or current OI from these fields.

The SOFR signed traded-price expression P(H27)−P(Z26)=−0.220, versus stored −0.225, corresponds to R(H27)−R(Z26)=+22.0bp, versus +22.5bp. Apparent 0.5bp flattening is indicative only: observation separation worsened to 405 seconds from 79.511 seconds. Do not reverse the spread sign, treat this as a quoted package, or book a hypothetical gain.

At retrieval, yen/euro observations were approximately 10m03s / 10m13s old; H27/Z26 were approximately 10m43s / 17m27s old. CL was about 48 minutes old and HG about 23 minutes old. Minute-bar endpoint markers can have zero volume and are not fresh traded-size assertions. The unobserved interval after each last is not evidence that a level did not trade.

Clarification added September 14, 2026 — September 13 Instagram carousel: The +5.14% estimate at $86.05/bbl and +9.0% estimate at a hypothetical $83/bbl entry refer to February 2027 WTI futures (CLG27). Each is modeled net scenario P&L divided by the corresponding futures notional, after the stated modeled costs. They are conditional four-month scenario estimates through January 13, 2027, not returns on posted margin or the simulated portfolio. No trade was executed; the simulated portfolio remained $1,000,000 in cash. This clarification defines the published percentages and does not change the dated scenarios or investment conclusion.

One decision per board idea

All entry/target/stop levels are research conditions, not orders. Scores use the inherited board scale out of 5. ER values preserve the baseline model convention; they are not returns on posted margin or cash-book P&L. No supplied dated probability vector is recalibrated or assigned a new as-of date by this monitor.

Constant-net-terminal-payoff sensitivities would mechanically re-mark yen to about +0.315% and euro to +2.217%, but these are not adopted as refreshed ERs. They omit new scenario validation and full carry/roll/cost re-estimation. The direction of an unentered short's prospective ER is the opposite of the gain that an already-held short might have experienced; this book owns neither.

Trend, curve, carry, basis, positioning and liquidity audit

Verified changes are narrow. Yen's delayed December rise extends its existing upward trend; euro's small bounce does not overturn its existing short-term decline or certify rejection. CL/HG price refreshes are too small and incompletely interval-matched to establish a trend break. No new commodity curve, carry, basis, official positioning/crowding or physical-inventory change was verified. Unknown is not favorable.

• Coffee: stale last and certification plans do not establish trend/physical confirmation. No refreshed basis, curve, carry, inventory or crowding evidence removes squeeze risk. Its 6.44K volume and 37,400 OI display are stale/context, not current executable liquidity. KCH27

• Crude/copper/gold: displayed session volumes 3.75K / 60 / 119, respectively; OI 58,706 / 45,407 / 33,505, with current OI as-of not independently established. These partial-session counts lack time normalization and spread/depth. Particularly thin deferred copper/gold observations cannot support sizing or a slippage assumption. No synchronized term structure or COMEX/LME/physical-basis update was verified. CLG27, HGH27, GCG27

• Yen/euro: vendor session volumes 16,310 / 5,199, without executable bid/ask/depth or certified current exact-contract OI. Exact March context quotes were yen 0.006601 at 22:56:35Z, volume 1, and euro 1.16735 at 23:00:27Z, volume 54, retrieved 00:03:53.091Z / 00:03:53.267Z. Both are pre-baseline and asynchronous. They are consistent with upward forward curves, potentially costly roll for a long yen and beneficial roll for a short euro, but establish no current tradable spread or quantified net carry. Individual endpoint URLs for these contextual observations were unavailable, so they are not used as actionable marks.

• Positioning context, not new flow: CFTC Sep 8 all-maturity legacy non-commercial yen 178,791 long / 167,995 short = +10,796 net, versus prior −92,227; euro 198,509 / 241,125 = −42,616 net, weekly net change −17,691. Yen squeeze history raises chase/reversal risk; euro bearish crowd raises squeeze risk. These do not locate 6JZ26/6EZ26 holders or prove post-baseline crowding change. SOFR all-tenor directional net is −3,103,337, with 2,607,702 spreading, not H27/Z26-specific crowding. Primary CFTC

• SOFR: H/Z vendor volumes 11,706 / 17,275 are not package liquidity. One contract per leg has contractual $25 per rate bp, but matching that unit does not remove legging, meeting-specific or reference-period risk. Basis, convexity, year-end funding and risk-premium effects remain unmodeled gates. CME rulebook

• Grains/nickel/sugar/gas/OAT/VIX: no usable new synchronized curve, basis, positioning or liquidity evidence changes eligibility. Grain legs remain stale; nickel exact mark and FX conversion remain unavailable; sugar remains suspended; gas remains asynchronous; OAT/Bund lacks contract construction; VIX has only Sep 11 settlements.

Correlation judgment is unchanged and qualitative. Copper and nickel shorts share China/manufacturing and supply-policy risk. Oil and gold longs are not reliable offsets when energy inflation drives higher rates. Coffee and grains share weather/freight/physical-inventory channels without a certified hedge ratio. Long yen plus short euro can partly offset USD notionals only if explicitly sized, while concentrating EUR/JPY and policy-divergence risk. SOFR legs need not remain event-neutral; OAT/Bund remains exposed to shared duration. No fresh synchronized data support a new measured correlation, diversification credit or larger risk budget.

Lifecycle and portfolio gates

No new commodity lifecycle event was verified; inherited stops/roll controls are retained without marking unverified dates as freshly checked. Before any new position, obtain official exact-contract notice, last-trade, delivery and broker liquidation dates; verify a horizon-consistent exit/roll, fees, spread, slippage and carry; and establish current initial/maintenance margin, contract accessibility, stress variation-margin needs and aggregate risk. The reviewed evidence does not provide a complete current margin/execution package. $1m cash is not itself proof that a leveraged trade is safe or eligible.

Specific unresolved horizon issues:

• 6JZ26/6EZ26: physically delivered; rule-implied last trade Dec 14, delivery Dec 16, subject to final exchange/payment-holiday confirmation. They cannot be carried unchanged to Jan 13. Require a separately approved and priced roll or earlier exit; the stale March context quote is not an approved replacement. Yen rulebook, euro rulebook

• SOFR: Z26 reference quarter is Dec 16, 2026–Mar 17, 2027; H27 is Mar 17–Jun 16, 2027. At Jan 13, some Z26 rate observations have realized while H27 remains forward. Both final settlements occur after the horizon: model horizon mark-to-market, not assumed terminal convergence. CME reference-period explanation

• NGF27: vendor calendar lists Dec 29 expiry / Dec 30 first notice; the exact official CME January row was not readable in this pass, so retain that discrepancy/official-validation gate. January gas expires before Jan 13; no unpriced automatic roll. Vendor calendar

• VIX: Oct 21 / Dec 16 expiries precede horizon, reinforcing why stale official marks cannot revive an unchanged four-month position. Cboe

• Fixed grain legs: standard equal 5,000-bushel contracts imply signed hypothetical P&L 50 × [2(W₁−W₀)−(S₁−S₀)] USD when both legs are in cents/bushel. The diagnostic ratio is neither dollar neutrality nor a complete terminal payout. Do not size, assign spread margin or claim reward/risk from ratio targets alone.

Regime and new-opportunity screen

The inherited regime remains energy-supply/inflation stress with central-bank/event risk, not a newly proven post-baseline break. Delayed equity weakness is consistent with that context; it is not an authorization to introduce equity shorts or revive VIX. Growth-risk demand can support bonds while inflation hurts duration, so do not infer a new universal bond selloff from oil alone.

ES/NQ rechecks were 2026-09-14T00:04:45.373–00:04:45.900Z; NKD failed exact-contract retrievals were 00:05:14.273–00:05:14.853Z. ES/NQ pages identify December and September 13 rollover: do not compare to September/continuous prices without adjustment. No superior, fully evidenced new opportunity was established.

Upcoming catalysts and event risk

Hong Kong/mainland cash discovery and a lower-tier NBS release are upcoming market information, not major policy decisions. The market, evidence and margin gaps remain unresolved, so no new risk is justified. Relevant economic releases, their market responses and fresh exact-contract evidence can change that assessment. Unscheduled Gulf or intervention headlines remain additional sources of risk.

The 72-hour window ends 2026-09-17T00:05:46Z. Scheduled events below were known or verified as schedule context, not newly announced surprises. All UTC conversions preserve source-zone caveats.

Beyond this precise window: BOE Sep 17 and the BOJ Sep 17–18 meeting with decision on Sep 18; do not move the BOJ decision to Sep 17 from an aggregator timezone. BOJ Jan 21–22, 2027 also lies beyond the fixed Jan 13 thesis horizon. BOJ official schedule

Unscheduled: verified Saudi pipeline restart/flow changes, shipping incidents, a confirmed replacement Oman meeting, FX intervention, and dated ICE certified-stock releases. The postponed meeting has no confirmed replacement time. Distinguish publication, event, observation and retrieval timestamps for every escalation.

Urgency queue and what changes the cash judgment

1. Immediate — remain cash; no orders or new positions. No stop, target, entry or thesis invalidation is proven newly triggered within the covered observations; stale/missing intervals remain unverified. Preserve the dated scenario assumptions. Do not silently mark carried rows as fresh.

2. Next usable market evidence — repair execution gates, not rankings by proxy. Seek fresh exact-contract bid/ask/depth, genuinely synchronized spread legs, certified current volume/OI context and official lifecycle/margin information. Prioritize the stale CL/GCG27/coffee/grain observations and unavailable ni2702/NKDZ26 evidence; refresh ES/NQ with authenticated observation dates/timezones during Asian discovery. No proxy backfill.

3. Physical/news checks — promote only facts that change terminal scenarios or satisfy entry conditions. Coffee needs actual certification plus failed retest; crude needs verified operational/export developments plus value entry; copper/nickel need China/physical evidence plus failed retests; gold needs a defended zone and rate/flow support; yen/euro need their exact entry behavior plus policy-response and roll evidence. A headline alone does not pass pricing/execution gates.

4. Major event uncertainty remains unresolved. China and FOMC outcomes and subsequent market responses can change the research judgment. Outcomes are not assumed before release, and previously reported implied probabilities are not a fresh forecast.

5. Only after every condition passes — reconsider risk. Require a fresh executable entry, evidence-backed horizon scenarios, positive net USD ER after costs/carry/roll, explicit stop/invalidation and gap stress, verified access/margins, synchronized spread construction, lifecycle fit and correlated portfolio stress. Reassess all eligible ideas together against cash. No row currently meets this test; no risk allocation or paper trade is justified by this review.