# The gas surplus is national; the thin cushion is regional.

Research cutoff: **2026-10-22T16:07:00.000Z**. Comparison cutoff: **2026-10-22T10:03:30.000Z**. The four-month horizon remains **18 October 2026–18 February 2027**, with earlier contract exits documented below. All eighteen board rows, six open simulated positions and zero pending orders were reviewed.

**Decision: exit the full +2 NNQZ26 / −1 M2KZ26 relative-value sleeve; hold three April one-ounce gold futures, two March micro natural-gas longs, five March micro-corn longs and six March micro-soybean-oil longs. The paired exit is recorded as assumed paper transactions at the nearest exact-contract references.**

## Natural gas: the headline surplus misses the deliverability split

EIA reported **3,644 Bcf** of Lower-48 working gas for the week ended October 16, a **77 Bcf injection**. Total stocks were **35 Bcf, or 1.0%, above** the 2021–25 average and 152 Bcf below last year. The regional split is the useful signal: South Central stocks were **3.4% below** their five-year average and salt storage was **14.9% below** it. Salt caverns are the fast-cycling part of the Gulf cushion, so a comfortable national aggregate does not eliminate winter deliverability risk. [EIA storage release](https://ir.eia.gov/ngs/ngs.html)

March standard gas traded at **$2.552/MMBtu at 16:04:22 UTC**. It is used as an identical-unit, zero-basis valuation proxy for held MNGH27, not as an observed micro price. From that reference, judgmental endpoints of $4.00 bull (20%), $2.85 base (50%) and $2.10 bear (30%) average $2.855. Across two micros the mark-to-endpoint outcomes are about **+$2,896 / +$596 / −$904**, or +$606 probability-weighted before exit commission. These weights are analysis, not measured probabilities.

Why not add? The national balance is still above normal, the injection was close to expectations, and warm-weather/production risk can leave March with a weak shoulder-season balance. EIA's October outlook forecasts continued high inventories and a lower 2027 average price even as exports rise. The regional deficit supports retaining convexity; it does not prove a national shortage. [EIA October outlook](https://www.eia.gov/outlooks/steo/pdf/steo_full.pdf)

**Economic decision: Hold two MNGH27; no add.** Keep the $2.00 stop, $4.00 target and February 12 exit. The current two-contract ledger risk remains about $1,568 under the 1.25 buffer. No automatic roll is assumed. [CME micro Henry Hub](https://www.cmegroup.com/articles/faqs/micro-henry-hub-natural-gas-futures-and-options-frequently-asked-questions.html)

## Corn: the order book improved faster than physical shipment

For October 9–15, USDA reported **1.5271 million tonnes** of 2026/27 corn net sales, up 36% week on week and 87% from the prior four-week average. Exports were only **0.7803 million tonnes**, down 43% week on week and 51% from the four-week average. Bookings validate demand interest; loadings show that immediate absorption has not yet caught up. [USDA Weekly Export Sales](https://apps.fas.usda.gov/esrqs/StaticReports/WeeklyHighlightsReport.pdf)

Same-month standard ZCH27 traded at **476.50¢/bu at 16:04:07 UTC** and is used as an identical-unit proxy for MZCH27 because the micro last was stale. From 476.50¢, 550¢ bull / 490¢ base / 420¢ bear endpoints weighted 25% / 50% / 25% give about **+$275** across five micros before exit cost. Scenario outcomes are +$1,837.50 / +$337.50 / −$1,412.50. A sales report that is strong in commitments but weak in shipments supports the starter and blocks an add.

**Hold five MZCH27.** Stop 450¢, target 550¢ and February 12 exit remain. The next test is whether inspections and subsequent export reports convert commitments into loadings.

## Soybean oil: exports do not confirm a disappearance surge

USDA reported just **900 tonnes** of soybean-oil net sales and **1,600 tonnes** of exports. Exact MZLH27 traded at **67.28¢/lb at 16:03:39 UTC**. At 70.50 / 67.50 / 64.56 endpoints weighted 30% / 35% / 35%, the six-micro model has only about +$45 forward value before exit cost. Low stocks and expensive energy remain support, but the new export evidence is too small to justify averaging up or down. **Hold six; no add.** Stop 64.56¢, target 70.50¢ and February 12 exit remain. [USDA report](https://apps.fas.usda.gov/esrqs/StaticReports/WeeklyHighlightsReport.pdf)

## Equity pair: exit because both mechanisms weakened

Initial claims fell to **194,000** for the week ended October 17, while continuing claims rose to 1.728 million. The Conference Board's September leading index was flat and its six-month growth remained positive; the Kansas City Fed survey separately showed slower new orders and elevated input-price pressure. That mix argues against both a simple recession short and an unqualified growth long. [DOL claims](https://www.dol.gov/ui/data.pdf), [Conference Board](https://www.conference-board.org/topics/us-leading-indicators/), [Kansas City Fed](https://www.kansascityfed.org/documents/20082/2026Oct22.pdf)

Exact **NNQZ26 was 29,342.5 at 16:02:37 UTC** and **M2KZ26 was 2,792.9 at 16:02:40 UTC**. Versus the 10:00 UTC bar references, the pair lost about $168.50 as Nasdaq fell and Russell rose. More important than the P&L, the Treasury rally did not support tech while low claims and a non-recessionary leading index supported small caps. That undermines both sides of the relative-value mechanism.

**Exit +2 NNQZ26 / −1 M2KZ26 together.** Retaining only the Nasdaq leg would leave concentrated tech beta; retaining only Russell would leave a small-cap short just as its financing/recession evidence weakened. Using the nearest exact references, the modeled exits are a Nasdaq sale around 29,342.0 after one adverse 0.5-point tick and a Russell buyback around 2,793.0 after one adverse 0.1-point tick, plus $10.50 total commissions. The exact accounting result is determined by the paper engine. These are assumed fills now, not observed exchange executions. [CME E-nano](https://www.cmegroup.com/rulebook/CME/III/300/344.pdf), [CME micro Russell](https://www.cmegroup.com/markets/equities/russell/micro-e-mini-russell-2000.contractSpecs.html)

## Gold: structural support did not beat today's rate channel

Exact 1OZJ27 last traded at **$4,137.75/oz at 15:08:59 UTC**; at 16:06:34 UTC the displayed bid/ask was **$4,126.75 / $4,127.50**. The quote had no separate exchange timestamp, so the retrieval clock is stated plainly. Reuters reported that higher oil reinforced higher-for-longer rate expectations; that inflation channel outweighed safe-haven demand. Importantly, nominal Treasury futures rallied today, so this is not evidence that nominal yields rose intraday. World Gold Council's September review provides the useful precedent: ETF inflows persisted while higher yields, a stronger dollar and managed-money liquidation still drove an 8.5% monthly decline. [Reuters gold update](https://www.reuters.com/world/india/gold-drops-over-1-oil-rally-reinforces-higher-for-longer-rate-outlook-2026-10-22/), [World Gold Council](https://www.gold.org/goldhub/research/gold-market-commentary-september-2026)

At the **$4,127.125 midpoint**, $4,600 bull / $4,250 base / $3,650 bear endpoints weighted 25% / 45% / 30% give about **+$91** across three ounces before cost. Trimming one would reduce modeled bear loss by about $477 but forfeit one-third of the bull convexity. With initial risk near 0.10% of opening NAV, **Hold three; do not add.** Stop $3,920, target $4,600 and February 18 exit remain. [CME one-ounce gold](https://www.cmegroup.com/articles/faqs/faq-1-oz-gold-futures.html)

## Challenger scan

- **Treasuries — Avoid.** ZNZ26 rallied to 106.03125 at 16:03:06 UTC. Chasing duration faces oil-inflation and issuance tails; shorting fights current momentum.
- **Copper — Avoid.** MHGH27 traded $6.5095/lb at 16:02:31 UTC. Strike risk is real, but March reaches its lifecycle before the full horizon, and the payoff remains sensitive to inventory/location normalization. A May expression is cleaner, not automatically attractive.
- **Oil/diesel — Avoid.** Middle East exports have adapted through costly rerouting while freight and attack risk remain elevated. Buying steep backwardation pays for protection; shorting retains a large disruption gap. [Reuters oil analysis](https://www.reuters.com/commentary/reuters-open-interest/dont-mistake-mideast-oil-rebound-normality-2026-10-22/)
- **SOFR and FX — Avoid.** Mixed labor/activity data and an oil-inflation tail do not produce a robust front-end rate or relative-policy edge.
- **Wheat/corn spread and S&P — Avoid.** Wheat demand remains weaker than corn, while a new broad-equity long would duplicate the Nasdaq leg without improving portfolio robustness.

## Next 72 hours

| UTC | Catalyst | Decision test |
|---|---|---|
| Oct 22 23:30 | [Japan September CPI](https://www.stat.go.jp/english/data/cpi/1582.html) | Services and energy breadth for yen and global duration. |
| Oct 23 07:30–13:45 | [Global flash PMIs](https://www.pmi.spglobal.com/Public/Home/PressReleaseDates) | Industrial resilience and input-cost pressure. |
| Oct 23 14:00 | [Michigan final sentiment](https://www.sca.isr.umich.edu/) | Whether oil pressure lifts inflation expectations. |
| Oct 23 19:30 | [CFTC positioning](https://www.cftc.gov/MarketReports/CommitmentsofTraders/ReleaseSchedule/index.htm) | Lagged October 20 crowding; not live Friday flow. |
| Oct 24 15:00 | Highland Valley earliest strike action | Verified interruption versus continued operation. |

## Simulated portfolio

After the equity-pair exits, the simulated book has four open positions and no pending orders. At the 16:12 UTC ledger update, NAV was **$1,003,004.33**, net P&L since inception was **+$3,004.33**, and unrealized P&L was **−$379.65**. Buffered open risk was **$4,816.45 (0.48% of NAV)**, with a **$25,000** modeled margin reserve.

| Position | Quantity | Entry | Current paper mark | Unrealized | Stop / target | Planned exit |
|---|---:|---:|---:|---:|---|---|
| March micro soybean oil (MZLH27) | 6 | 67.22¢/lb | 67.28¢ exact | +$21.60 | 64.56 / 70.50 | Feb 12 |
| March micro corn (MZCH27) | 5 | 481.0¢/bu | 476.5¢ via ZCH27 proxy | −$112.50 | 450 / 550 | Feb 12 |
| March micro natural gas (MNGH27) | 2 | $2.621/MMBtu | $2.552 via NGH27 proxy | −$138.00 | $2.00 / $4.00 | Feb 12 |
| April one-ounce gold (1OZJ27) | 3 contracts | $4,177.00/oz | $4,126.75 exact displayed bid | −$150.75 | $3,920 / $4,600 | Feb 18 |

The two equity legs were closed as assumed paper executions at the latest suitable references: sell two NNQZ26 at a modeled **29,342.0** from the 29,342.5 exact last at 16:02:37 UTC, and buy one M2KZ26 at a modeled **2,793.0** from the 2,792.9 exact last at 16:02:40 UTC. Combined gross exit loss was **$156.40** before costs. These were simulations, not observed exchange fills or real orders.

## Method and limitations

Scenario endpoints and weights are analyst judgments, not measured probabilities or proprietary scores. Whole-contract risk uses the ledger's 1.25 adverse-scenario buffer plus modeled costs. Exact dated lasts are not executable prices. Standard corn and gas references are explicitly labeled same-month zero-basis valuation proxies for held micro contracts. The equity exits are assumed paper fills under the nearest-reference convention; no real order was placed.

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


## Return on entry margin

**Margin-weighted trade return: +1.98%**, from **$3,004.33 net P&L** divided by **$151,562.50 allocated entry margin** across open and closed trades. Mixed recorded initial margin and modeled reserves. Reused margin counts for each deployment; this is not a compounded account return.

Valuation: 2026-10-22T16:12:06.197Z. Closed-trade P&L includes entry and exit commissions; open P&L includes entry commissions, with exit costs still to come. Slippage is already embedded in fills. Each trade uses its original entry margin in USD, allocated proportionally to partial exits and the remaining holding. Current margin requirements are separate. Account return on starting capital: **+0.30%**.

This additional presentation was introduced October 28, 2026. Historical editions use their own dated portfolio records; original research dates and accounting results are preserved.
