ATGGLOBAL MARKETS
RESEARCH
Research archive

Oil drew, but demand did not accelerate.

U.S. crude and distillate inventories fell, yet gasoline stocks rose and four-week gasoline and diesel demand remained below last year. The crude draw was amplified by a large export swing, so chasing the post-report rally offers weak asymmetry.

Dated research, not live quotes. Exact IBKR contract books were retrieved around 16:00 UTC. The Bank of Canada and EIA releases are primary evidence. Reuters supplies attributed market reaction and the post-decision Canadian-dollar reference. Earlier positioning inputs retain their dates. Research views and scenario estimates are separate from simulated positions and performance.

Oil drew, but demand did not accelerate

Market-data cutoff: 28 October 2026, 16:04 UTC / 12:04 Toronto. This ordinary daily edition preserves the 25 October 2026–25 February 2027 horizon and re-underwrites all eighteen board rows, the single open simulated position and the liquid challenger set after the Bank of Canada and EIA releases.

Decision: hold three April 2027 one-ounce gold futures, with no add. Keep the completed euro exit closed. Add no simulated exposure. The strongest fresh finding is that the EIA crude draw was amplified by a large export swing rather than broad domestic-demand acceleration: gasoline stocks rose, while four-week gasoline and diesel demand remained below last year. Physical tightness is real, but chasing December Micro WTI near its session high offers weak probability sensitivity.

What changed

MarketDated referenceEconomic decision
December / March Micro WTIExact midpoints $103.35 / $97.14 around 16:01 UTCWait. The $6.21 backwardation and post-EIA rally already pay for much of the prompt scarcity.
Canadian dollarUSD/CAD spot about 1.4050 at 14:50:50 UTC after the decisionWait. The expected hike arrived without clean currency follow-through.
April one-ounce gold 1OZJ27$4,120.50 bid / $4,121.50 ask around 16:02 UTCHold three; no add. Positive weighted payoff survives, but the Fed, high yields, contango and thin April trading cap size.
December 10-year Treasury ZNZ26Exact midpoint 105.7734375 at 16:01 UTCWait. The short has negative weighted payoff and worse sensitivity before the Fed.
March Micro Nasdaq MNQH27Exact last 30,189 at 15:59:41 UTCWait. Earnings upside exists, but Fed and megacap event gaps dominate.
March Micro Copper MHGH27Exact midpoint $6.75225/lbWait. Asian scarcity remains real; the U.S. expression is less clean.
March corn ZCH27Standard-contract midpoint 464.125¢Wait. Harvest pressure remains, but a micro short still fails sensitivity.

Oil: the draw came with a demand caveat

The EIA reported that commercial crude inventories fell 1.9 million barrels to 410.6 million in the week ended 23 October. Distillates fell 1.5 million to 102.8 million, while gasoline rose 3.9 million to 209.2 million. Crude and distillate stocks remain below their five-year seasonal averages, so the physical system is tight. EIA weekly overview

The composition matters. Crude exports rose 1.412 million barrels per day to 4.854 million, while refinery inputs slipped 123,000 b/d. Net crude imports fell 796,000 b/d. Four-week total products supplied was only 0.2% above last year; gasoline was 0.7% below and distillate 0.6% below. The crude draw therefore says more about export timing and supply scarcity than a broad acceleration in U.S. consumption. EIA weekly estimates are noisy, and Reuters reported that ship-tracking data put exports roughly 1 million b/d below the EIA estimate, an unusually large discrepancy. EIA weekly overview Reuters oil coverage

December Micro WTI displayed $103.34/$103.36 around 16:01 UTC after trading from $99.74 to $103.47. March displayed $97.09/$97.19. The resulting $6.21/bbl of indicative backwardation is about 6% of the prompt price and already rewards near-term scarcity while pricing later normalization.

For a one-contract December Micro WTI long from $103.35, the four-month study uses $125/$106/$82 endpoints with 25%/45%/30% weights. It produces only +$12.50 after $7.50 modeled costs. Moving ten probability points from the bull case to the bear case produces −$417.50. One contract fits mechanically: a $21.35 move to the $82 study invalidation is $2,135, or about $2,677 with the 1.25 buffer and costs, below the 0.50% thesis cap. The economic distribution still fails. Wait. Do not chase the rally, and do not short a physically tight prompt market merely because the draw had an export component.

Bank of Canada: the hike was expected enough to mute CAD

The Bank of Canada raised the overnight rate 25 basis points to 2.50%, its first increase since July 2023. It said inflation is now expected to remain around 3% through much of 2027 and return close to 2% around mid-2028. Higher energy prices and increased pass-through make inflation more persistent, while growth has improved modestly. The Bank left the door open to further tightening but said it would proceed carefully. Bank of Canada decision and MPR Bank of Canada October MPR

USD/CAD was about 1.4050 at 14:50:50 UTC after a two-way reaction, little changed from the pre-decision area. That lack of follow-through is evidence that the 25-basis-point move was substantially priced. It does not prove CAD must fall: higher rates and expensive oil support the currency, while trade uncertainty and a still-large leveraged-fund short position pull the other way. CFTC data for 20 October showed leveraged funds net short 91,113 Canadian-dollar contracts; that is a dated positioning input, not a live flow reading. CFTC financial-futures positioning

Economic decision: Wait. No paper intention. The post-decision response rejects both a reflex long and a reflex short. Re-underwrite after sustained spot trade below 1.3970 or above 1.4125 CAD per U.S. dollar, rather than assume an intraday break occurred.

Gold: hold the small hedge, do not add before the Fed

April one-ounce gold displayed $4,120.50 bid / $4,121.50 ask around 16:02 UTC. The last trade was $4,124 at 14:27:43 UTC and was stale, so the edition uses the conservative displayed bid. April volume was 75 and open interest 302. December was much more liquid at roughly $4,049.25/$4,050.00. The approximately $71.38/oz April premium, or 1.76%, is an indicative carry measure from non-synchronous books, not an executable calendar spread.

The fundamental mechanism remains balanced. Energy scarcity and geopolitical uncertainty support defensive demand, while high real yields, a firm dollar and the possibility of a hawkish Federal Reserve path pressure gold. September gold-ETF inflows and strategic central-bank demand remain supportive older context; the upcoming Fed decision and Thursday's U.S. GDP/PCE combination are the near-term tests. World Gold Council ETF flows Federal Reserve calendar

Four-month gold caseWeightApril endpointP&L on three contracts from $4,120.50
Bull: yields ease and strategic demand persists30%$4,600+$1,438.50
Base: partial recovery45%$4,200+$238.50
Bear: real-yield pressure and liquidation25%$3,700−$1,261.50
Probability-weighted100%$4,195+$223.50 gross / +$205.50 after future costs

A ten-point shift from bull to bear reduces expected payoff to −$64.50. Preserve the $3,920 stop, $4,600 objective and 18 February exit. From the current bid, raw price-to-stop loss is $601.50; with the 1.25 gap buffer and $18 future-cost reserve, planned remaining risk is about $770, or 0.077% of NAV. Current price-to-target versus price-to-stop reward/risk is about 2.39x. Gaps can exceed the modeled loss.

The simulated lot entered at an assumed $4,177 per ounce using the 22 October exact-contract reference and adverse tick. At this edition's mark, gross open P&L is −$169.50; after $10.50 of entry commission, net open performance is −$180.00, or −6.00% on the original $3,000 modeled entry-margin reserve. The reserve is collateral, not maximum possible loss. Hold; do not add.

Rates, equities, copper and corn

December 10-year Treasury: The exact midpoint was 105.7734375. One hypothetical short at 104.50/105.75/108.00 endpoints and 30%/45%/25% weights gives −$171 after $7 costs, falling to −$521 under sensitivity. The Bank of Canada outcome reinforces global inflation risk, but the Fed can still reverse duration quickly. Wait.

March Micro Nasdaq: The exact last was 30,189 and displayed midpoint 30,193.75. One hypothetical long at 33,000/31,000/27,000 endpoints and 30%/45%/25% weights gives +$806 after $7 costs, but −$395 after a ten-point bull-to-bear shift. Alphabet and Microsoft report after the close, so a tight pre-event stop is not reliable. Wait.

March Micro Copper: The exact midpoint was $6.75225/lb. Asian premia and low regional stocks support the long mechanism, while the deferred COMEX curve and large tariff-distorted U.S. inventory make the U.S. future an imperfect expression. One micro long at $7.50/$6.90/$5.90 endpoints and 30%/45%/25% weights gives +$187 after costs, but −$213 after sensitivity. Wait.

March corn: The standard-contract midpoint was 464.125¢ and serves as a same-month proxy for a smaller Micro Corn study. Harvest supply and fund liquidation are bearish, while the latest ethanol-production improvement offsets part of the demand concern. Two hypothetical micro shorts at 420/470/510 endpoints and 35%/40%/25% weights give +$18.50 after costs and −$161.50 after sensitivity. Wait for a stronger rebound or export deterioration.

Remaining board: Soybean oil, the soybean-oil/meal spread, SOFR, micro S&P, distillates, natural gas, wheat/corn, Russell and retired expressions received no verified post-cutoff evidence strong enough to change their decisions. Earlier observations retain their dates. The euro short remains closed; no trigger, stop or target is assumed to have occurred.

Portfolio and performance

The economic decision produces no new transaction. The paper engine records the gold review and exact $4,120.50 mark at the current decision time. Edition-matched simulated NAV is $1,002,319.68. Realized net P&L remains $2,499.68 and open net P&L is −$180.00. Current modeled margin reserve is $3,000 and open planned risk remains within portfolio limits.

The combined margin-weighted trade return is approximately 1.42%: total net P&L divided by allocated original entry margins across open and closed simulated deployments. The aggregate denominator mixes sourced initial margins and disclosed modeled reserves, and reused margin counts for each deployment. It is not annualized, compounded or the account return. Account return on $1,000,000 starting capital is approximately 0.23%.

At the recorded valuation of 2026-10-28T16:05:58.563Z, simulated NAV was $1,002,319.68, net P&L since inception was $2,319.68, and open gross P&L was -$169.50 before remaining entry commissions. The book contained 1 open positions and 20 fully closed positions. Results include modeled costs and dated marks; they are simulated, not live investment performance.

There is one open position and no pending order. No real order was placed.

Next 72 hours

Toronto timeCatalystDecision question
Oct 28, 14:00 / 14:30FOMC decision / press conferenceDoes guidance keep a December hike live for gold, the dollar and duration?
Oct 28, after closeAlphabet and Microsoft earningsDo cloud returns and cash flow justify AI capital spending?
Oct 29, 08:30U.S. GDP, income and PCECorrelated growth and inflation risk immediately after the Fed.
Oct 29, 09:15 / 09:45ECB decision / press conferenceRelative U.S.–euro policy after the completed euro exit.
Oct 30Bank of Japan decision and OutlookIntervention and global-duration spillover; release time is not fixed.

Official schedules: Federal Reserve, BEA, ECB, Bank of Japan.

Decision sheet

1. Hold three 1OZJ27; no add. Retain the $3,920 stop, $4,600 objective and 18 February exit.

2. Keep Micro WTI flat after EIA. The physical market is tight, but the rally, export noise and demand mix make a new long fragile; a short fights prompt scarcity.

3. Keep Canadian-dollar futures flat after the hike. The expected decision produced no clean follow-through.

4. Remain flat Treasury, Nasdaq, copper and corn. Each fits mechanically or has a smaller expression, but each fails probability sensitivity or immediate event compensation.

5. Keep the euro short closed. The prior assumed exit is final; no automatic re-entry.

6. No simulated fill this edition. The paper book is re-marked and reviewed.

Scenario weights are analyst judgments, not measured probabilities. Displayed midpoints may not be executable, delayed references are labeled, stops can gap, thin contracts can show stale books, and future costs and margins can change. This is research and simulated-portfolio management, not personalized investment advice.

Return on entry margin

Margin-weighted trade return: +1.42%, from $2,319.67 net P&L divided by $163,062.50 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: 2026-10-28T16:05:58.563Z. 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: +0.23%.

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

ATG / GLOBAL MARKETS

Follow the evidence.

New research, useful explanations, and the decisions that follow.

Follow on Instagram Subscribe by RSS