Gold’s drop finally paid for the risk.

Gold’s drop finally paid for the risk.

Research cutoff: 2026-10-22T04:13:00.000Z. Comparison cutoff: 2026-10-21T22:13:27.000Z. The four-month horizon remains 18 October 2026–18 February 2027, with earlier contract exits where documented. All eighteen board rows, five inherited open simulated positions and zero pending orders were reviewed.

Decision: enter three April 2027 one-ounce gold futures in the simulated portfolio. Hold two Nasdaq E-nanos, one micro Russell short, six micro soybean-oil longs, five micro corn longs and two micro natural-gas longs. No add, trim, exit or reversal.

Gold: price compensation changed the decision

April one-ounce gold traded at an exact $4,176.75/oz at 03:47:41 UTC, down 0.99% on the provider day. The reference is an individually timed IBKR last, not a broker-grade fill. The new entry matters because the thesis did not improve: the price did. Gold still offers policy and geopolitical convexity, while a hawkish real-yield shock remains the central countercase.

The four-month endpoints remain $4,600 bull, $4,320 base and $3,920 bear/stop, weighted 25% / 40% / 35%. From the reference, one-ounce net outcomes after estimated exit cost are +$415.75 / +$135.75 / -$264.25. That produces +$65.75 expected value per contract, or +$197.25 for three. Moving five probability points from bull to bear and doubling future costs still leaves +$72.75 across the position. Lowering every endpoint another $50 makes the result -$77.25, so this is a small starter, not a high-conviction allocation.

Three contracts risk about $985 to the $3,920 stop after the 1.25 buffer and modeled round-trip friction, about 0.10% of NAV. A separate $3,700 gap case is about $1,810. The contract is one troy ounce with a $0.25 tick and cash settlement; the public margin table showed $232 maintenance, while the simulation reserves a conservative $1,000 per contract. The position exits February 18, ahead of the March 29 date bound, with no automatic roll. CME one-ounce gold; AMP margin table

Economic decision: Enter starter. Assumed paper execution: three 1OZJ27 longs now from the exact dated $4,176.75 reference; the accounting engine applies one adverse tick and commissions. This is simulated, not an observed exchange execution or real order.

Earnings breadth survived; the valuation bar did not fall

Nasdaq E-nano futures fell 1.30% to 29,617.5 at 04:09:05 UTC while micro Russell futures fell 0.57% to 2,778.4 at 04:08:24 UTC. The Nasdaq move was broad repricing after mixed company reactions: Meta remained higher, while Tesla and IBM were lower. It would be incorrect to attribute the aggregate move to one issuer.

Meta’s follow-up call also narrows a common dilution interpretation. Management said the “hundreds of billions” figure was a theoretical founder-control calculation, not a statement of current equity-financing need. The pro-rata Class C distribution is split-like; future issuance remains optionality, not immediate economic dilution. The harder issue is cash conversion: compute-intensive products still require large efficiency gains and revenue scale. Meta follow-up transcript; Meta earnings-call transcript

Texas Instruments reported strong industrial, automotive and data-center growth, while IBM maintained 4–5% constant-currency growth and raised its free-cash-flow growth outlook. Tesla’s operating margin of 2.6%, 56% year-over-year operating-income decline and $1.53bn free cash flow show why the market still distinguishes demand from conversion. TI prepared remarks; IBM results; Tesla deck

The simulated equity pair therefore remains two NNQZ26 longs against one M2KZ26 short. Fresh-mark marginal expected value is +$204.30, but simultaneous modest adverse probability shifts leave only +$0.80; doubling future exit costs turns that to -$10.40. Joint ordinary-stop loss is about $1,266 and a joint adverse gap/rotation case about $2,666. The pair is not beta-neutral or loss-proof. Hold, do not add. Stops remain 28,000 on NNQ and 2,900 on M2K; both exit November 20.

Held commodities

Soybean oil — Hold six, no add

MZLH27 is valued from the same-month standard ZLH27 proxy at 66.28¢/lb at 04:05 UTC. Standard and micro contracts share settlement economics, but this is explicitly not an observed micro print. Target/base/stop weighting gives about +$334 forward expected value after estimated exit commission; a basic adverse shift leaves about +$120. Strong crush can rebuild oil stocks, so energy support without fresh product disappearance is not enough to average down. Stop 64.56, target 70.50, February 12 exit. CME micro agriculture FAQ

Corn — Hold five, no add

MZCH27 is valued from same-month standard ZCH27 at 478.00¢/bu at 03:59:41 UTC. The weekly EIA data showed ethanol output rising to 1.094m b/d from 1.065m, but stocks also rose to 24.6m barrels from 23.9m. Production is not proof of final consumption. Current target/base/stop weighting gives about +$358 using the more conservative endpoint set and latest public proxy; today’s export-sales release is the cleaner absorption test. Stop 450, target 550, February 12 exit. EIA weekly overview

Natural gas — Hold two, no add

MNGH27 retains the carried same-month NGH27 proxy at $2.574/MMBtu from 22:01:09 UTC on October 21. The latest actual storage was 3,567 Bcf, 45 Bcf above the five-year average; South Central and salt inventories were below their five-year averages. That regional flexibility supports a small winter option, not a national shortage claim. NOAA’s current forecast keeps above-normal temperatures likely across much of the eastern United States into October 29–November 4. The severe endpoint sensitivity is negative, reinforcing no add before today’s storage report. Stop $2.00, target $4.00, February 12 exit. EIA storage; NOAA current discussion; EIA gas outlook

Copper and oil: large headlines, poor payoff

March micro copper fell to an exact $6.5955/lb at 04:05:38 UTC. LME stocks rose 11,475 tonnes on October 21 and the cash/three-month backwardation narrowed. At revised 7.45 / 6.55 / 5.80 endpoints and 25% / 45% / 30% weights, one micro has about -$123 net expected value. Highland Valley strike notice is a genuine 72-hour catalyst, but no verified outage existed at the cutoff. One contract fits; negative economics still mean Avoid. Reuters copper inventory report; Teck release

Brent rose above $110 and WTI above $98 in Asian trade as markets continued to price Saudi attack and shipping risk. The diplomatic-departure report itself predated the comparison cutoff; the new information is session pricing, not a newly announced evacuation. U.S. crude and gasoline inventories fell, but distillates rose and four-week distillate consumption was 2.4% below a year earlier. An emergency heating-oil loan provides limited regional relief. Both oil tails remain wide, so neither a long nor short offers robust stressed value at current references. Avoid a new oil trade. Reuters October 22 oil update; EIA petroleum summary

Rates, FX and remaining challengers

• Treasuries: December 10-year notes rallied to 105.484375, but a long produces negative expected value after costs under the current inflation/supply endpoints. Avoid.

• SOFR: March SR3H27 at 95.485 implies 4.515% average quarterly SOFR. The December-to-March curve slopes toward higher implied rates, but the short’s modeled expected value is only about +$6 and fails a small adverse shift. Avoid.

• FX: no newly verified relative-policy shock supports a fresh CAD, sterling or euro position. Avoid.

• Wheat and soybeans: March recovery studies fail their basic expected-value tests; held corn remains the better small absorption expression.

• Silver and platinum: both weakened, but neither offered a more robust smaller-contract payoff than the one-ounce gold starter.

The broad Asia handoff was risk-off: the regional equity index slipped, S&P futures were lower, Treasury yields fell and oil rose on Gulf risk. This mix supports small convex gold exposure, not a wholesale growth-break bet. Reuters global-markets report

Next 72 hours

Simulated portfolio

PORTFOLIO_SUMMARY

Method and limitations

Scenario endpoints and weights are analyst judgments, not measured probabilities or proprietary scores. Whole-contract risk uses 1.25 times stop-distance loss plus modeled costs. Dated lasts and proxies are not settlements or observed executions. The simulated gold entry uses the nearest exact dated reference and the engine’s standard adverse tick; it is not a real order. Standard soybean-oil, corn and gas references are clearly labeled valuation proxies for their micro contracts.

This is research and a simulated portfolio, not personalized investment advice. Futures can gap, require additional margin and lose more than a modeled stop.