📅 2026-10-04 · 紐約WTI原油 · 單品種深度報告

數據截止: 2026-10-02 · 結算口徑: 交易所最終日結算價

WTI Crude (CL=F) Deep Dive | 2026-10-04 — Bullish Above 88.28, Backwardation vs Soft Tape | 2026-10-04

Call: Bullish WTI, invalidation on a settle below the S1 pivot 88.28. The prior session settle was 91.11 (2026-10-02), down 1.9% on the day and 1.41% over five sessions, but still +3.5% over 20 sessions and only 32.9% up the 20-day channel of 85.92–101.69 — a pullback inside an uptrend, not a trend break. The curve is in backwardation of 1.68 (1.88%) M1-M2 with 22.5% annualized roll yield, the cleanest evidence of prompt tightness; gasoline stocks sit 5.8% and distillates 12.3% below their five-year same-week averages while crude is only 2.1% above. The 3:2:1 crack at 64.65 sits in the 92nd percentile of the past year and 97th of three years, so refining pull is strong. Managed-money net length fell 22,236 lots in the week to 2026-09-29 to 79,592, but crowding is only the 18th percentile of three years — light, not crowded. The offsetting risk is the seasonal window (median -4.49% over the next 20 sessions) and a 5.28% 10-year yield. Buy dips toward 90.89, stop 88.28, target 96.34.

1. Bottom Line & Directional Bias

Call: Bullish WTI Crude (CLX26.NYM, the November contract referenced off the CL=F continuous series). Invalidation: a daily settle below the S1 pivot at 88.28, which would also put price back inside the lower half of the 20-day channel and break the sequence of higher lows since the September base.

Three reasons. First, the term structure: M1-M2 backwardation of 1.68 (1.88%) with a 22.5% annualized roll yield says prompt barrels are scarce — that is a physical signal, not a positioning one. Second, product tightness: gasoline stocks are 5.8% and distillates 12.3% below their five-year same-week averages, with refinery utilization at 94% and the 3:2:1 crack at 64.65 (92nd percentile of one year, 97th of three). Third, positioning is light: managed-money net length is only the 18th percentile of three years after a 22,236-lot weekly cut, so the sell-off has already flushed speculative length rather than being driven by it.

The counterweight is the calendar: the next 20 sessions have historically been the weakest seasonal window for crude (median -4.49%, up only 5 of 15 years), and the 10-year yield at 5.28% is a headwind for the whole commodity complex. That argues for buying weakness, not strength.

2. Price Action & Technical Analysis

Settle 91.11 (2026-10-02, CME Group final daily settlement), -1.9% on the day, -1.41% over five sessions, +3.5% over 20 sessions. The 20-day channel runs 85.92–101.69, putting the settle at the 32.9% position — lower third, but well off the floor. The 52-week range is 54.98–119.48, so the market is mid-range on a yearly view and has given back a large part of the late-September push toward 101.69.

Volatility is elevated and, importantly, realized is running hot relative to the option market's own history: ATR14 is 4.63, i.e. 5.08% of price as a full daily range, and RV20 is 42.9%. That is a market where a 2% intraday swing is routine, which is why the stop must sit beyond a structural level rather than a round number.

Pivots from the settle: P 90.89, R1 93.73, S1 88.28, R2 96.34, S2 85.44. The settle at 91.11 is marginally above the pivot, so the immediate bias is neutral-to-constructive inside the pivot band; a reclaim of R1 93.73 opens R2 96.34, while losing S1 88.28 exposes S2 85.44 and the 20-day low at 85.92 — note those two sit almost on top of each other, making 85.44–85.92 the real line in the sand.

In early Asian trade the tape is quoted around the prior settle with no decisive extension; treat any move on the 2026-10-04 bar as unfinished Globex/Asia price action, not a settlement.

The last completed weekly bar (2026-09-21 to 2026-09-25) opened 96.75, high 97.22, low 88.67, closed 92.41, -3.82% w/w — a wide-range down week that nonetheless held above the 88 area. The current week (from 2026-09-28, five sessions) is not closed and shows 91.11, -1.41%; no weekly-close conclusion can be drawn from it. The constructive read: the completed weekly low at 88.67 sits just above S1 88.28, so the two levels reinforce each other as the pivot of the whole setup.

View: constructive while above 88.28; the first real test is 93.73, and a settle above it would confirm the pullback is over.

3. Supply-Demand Balance & Fundamental Drivers

Crude: EIA stocks 426,398 kb as of 2026-09-18, +2,969 kb w/w, four-week cumulative -2,512 kb, and +2.1% versus the five-year same-week average. That is a modest surplus, not a glut — and it is the only one of the three main buckets that is above normal.

Products are the tight part of the barrel. Gasoline stocks 206,046 kb, -1,686 kb w/w, four-week cumulative -796 kb, -5.8% versus the five-year same-week average. Distillates 107,431 kb, -428 kb w/w, four-week cumulative +4,040 kb, but still -12.3% versus the five-year same-week average. Distillate inventories 12.3% below normal heading into the Northern Hemisphere heating season is the single most bullish line in this dataset, and it is corroborated by refinery utilization at 94% — refiners are already running hard and still cannot rebuild the product pool.

Upstream response remains slow. Baker Hughes US oil rigs at 455 (2026-09-25), +3 w/w and +31 y/y. A 31-rig year-on-year increase is real but modest against a 94% utilization rate and falling product stocks; it is not the kind of supply response that breaks a backwardated curve within a quarter.

The curve confirms the physical read: M1-M2 backwardation 1.68 (1.88%), roll yield 22.54%, slope -1.27. Backwardation of this magnitude is a roll cost for longs, not a cap on price — it is the market paying prompt holders to deliver now, which is what tightness looks like.

Macro transmits through two channels. The 10-year at 5.28% (+0.76%) raises the cost of holding inventory and is a headwind for demand expectations; DXY at 101.93 (-0.17%) is a mild offset, since a softer dollar mechanically supports dollar-denominated crude. Neither is decisive against a 22.5% roll yield.

View: fundamentally bullish on products and curve shape; the crude surplus of +2.1% versus the five-year average is the bear's only real foothold and it is small.

4. Positioning & Fund Flows

CFTC managed money, week to 2026-09-29: open interest 1,878,576, longs 209,028, shorts 129,436, net 79,592, change -22,236. The prior three weeks: net 101,828 (-4,451), 106,279 (-5,452), 111,731 (+17,450). So net length has been cut in three of the last four weeks, with the largest cut in the most recent week — and that week coincided with price weakness.

Crucially, this is de-risking, not capitulation into a crowded short. Net as a percentage of open interest is 4.24%, and the three-year crowding percentile is 18.43 — low. CTA trend proxy reads 98, i.e. trend followers are still positioned long, and hedge pressure is 48.56%, down from 51.68% four weeks ago. The combination — falling net length, low crowding percentile, still-long trend proxies — says speculative froth has been removed while the systematic bid has not flipped.

Volatility pricing: ^OVX (WTI implied vol) 51, -0.69 points on the day, 1-year percentile 49%. Against RV20 of 42.9%, IV minus RV is +8.1 vol points (IV/RV 1.19). Options are paying up modestly for event risk — FOMC minutes and the EIA prints this week justify some of that — but at the 49th percentile, implied vol is not stretched. For a directional long, that argues for expressing the view in futures or a defined-risk structure rather than paying up for outright calls.

View: positioning is a tailwind, not a headwind — the light crowding percentile is what gives the long room to run if the physical tightness asserts itself.

5. Cross-Asset Relative Value

WTI-Brent spread: -11.14 USD/bbl, 1-year percentile 4.37%, 3-year percentile 1.46%. WTI is historically cheap to Brent on both windows — near the extreme of the past three years. For a WTI-specific long, that is a supportive relative-value backdrop: either WTI catches up or the spread is already discounting a US-specific surplus that the +2.1% crude stock excess only partly justifies.

3:2:1 crack: 64.65 USD/bbl, 1-year percentile 92.46%, 3-year percentile 97.49%. Refining margins are near multi-year highs, which is the demand-side confirmation of the product inventory draw. High cracks pull crude through the refinery system; they also invite higher runs, which is why the 94% utilization figure matters.

Oil-gold ratio: 0.0219, 1-year percentile 90.08%, 3-year percentile 46.3%. Crude has been strong versus gold over the past year but is only mid-range over three years — consistent with a market that has recovered from a deep low rather than one that is stretched.

View: the cross-asset set is uniformly supportive of crude — cheap versus Brent, strong cracks, mid-range versus gold. None of these is a timing signal, but all three argue against a sustained breakdown.

6. Historical & Seasonal Patterns

Seasonality for the same calendar start, next 20 sessions, last 15 years: mean -1.38%, median -4.49%, up in 5 of 15 years. Best case 2011 +21.83%, worst 2018 -14.31%. Small sample, context only.

The honest read is that this is a negative seasonal window: the median outcome is a 4.5% decline and the hit rate is only one in three. That is a genuine argument against chasing strength here, and it is why the strategy below is a buy-limit into support rather than a breakout entry. It also means the burden of proof sits with the bulls: the physical data (backwardation, product draws, cracks) has to overcome a seasonal headwind that has historically produced a mid-single-digit drawdown in this exact window.

View: seasonality argues for patience and for buying weakness toward 88–89 rather than paying up above 93; it does not by itself overturn the bullish physical setup.

7. Scenario Analysis (Base / Bull / Bear)

Base case — 50% — grind higher within the pivot band. Trigger: price holds above P 90.89 and the EIA prints on 2026-10-07 show continued product draws. Path: 91.11 → R1 93.73, with a settle above 93.73 opening R2 96.34. Action: hold the long from the 90.89 area, trail the stop up as R1 is approached. This agrees with the section 1 call.

Bull case — 30% — breakout on a product squeeze. Trigger: a settle above R1 93.73, ideally accompanied by a distillate draw that pushes the -12.3% versus five-year gap wider, plus a soft ISM Services print on 2026-10-05 that weakens the dollar and rates. Path: 93.73 → 96.34 → the 101.69 20-day high. Action: add on the settle above 93.73, move the stop to breakeven-plus, target 96.34 first and let a runner work toward 101.69.

Bear case — 20% — seasonal drawdown breaks the base. Trigger: a daily settle below S1 88.28, which would also breach the completed weekly low at 88.67 and put the 20-day low 85.92 / S2 85.44 in play. Catalysts: a hawkish FOMC minutes read on 2026-10-08 with the 10-year already at 5.28%, or an EIA crude build that widens the +2.1% versus five-year surplus. Path: 88.28 → 85.92 → 85.44. Action: exit the long on the settle below 88.28; do not attempt to average down, because a break there invalidates the higher-low structure that the entire bullish case rests on.

Probabilities sum to 100%. The base case is the section 1 call; the bull and bear cases are the weighted paths around it, not competing conclusions.

8. Trading Strategies & Risk Management

Strategy 1 — Buy the pivot retest (primary). Direction LONG. Entry 90.89 (the P pivot, working a limit into the current 91.11 area). Stop 88.28 (S1, also just below the completed weekly low 88.67). Target 96.34 (R2). Horizon 1-5 days. Conviction 7/10. Size: half of normal risk budget, because ATR14 of 4.63 (5.08% of price) means the stop is roughly 0.6 ATR away and the seasonal window is negative — the position should be small enough to survive a 4.63-point adverse day without forcing a decision.

Strategy 2 — Add on confirmation (secondary). Direction LONG. Entry on a daily settle above 93.73 (R1). Stop 90.89 (P). Target 101.69 (20-day high). Horizon 3-10 days. Conviction 6/10. Size: the second half of the risk budget, only triggered if Strategy 1 is working. This converts the trade from a mean-reversion buy into a trend continuation if the market proves the pullback is finished.

Risk management: the single invalidation for both is a settle below 88.28. If that occurs, both positions are flat and the bias flips to neutral pending a new base. Do not add on the way down; the 85.44–85.92 zone is the next support but it is a level to reassess at, not to pre-position for.

9. This Week's Data Calendar

BJT 10-05 22:00 | ET 10-05 10:00 — ISM Services PMI SEP, forecast 54 vs previous 55.4, surprise if outside 54±1.4 (USD/HIGH; transmits to DXY and thus crude).
BJT 10-07 04:30 | ET 10-06 16:30 — API Crude Oil Stock Change, OCT/02 (USD/MEDIUM; CL, BZ).
BJT 10-07 22:30 | ET 10-07 10:30 — EIA Crude Oil Stocks Change, OCT/02 (USD/MEDIUM; CL, BZ).
BJT 10-07 22:30 | ET 10-07 10:30 — EIA Gasoline Stocks Change, OCT/02 (USD/MEDIUM; CL, BZ).
BJT 10-08 02:00 | ET 10-07 14:00 — FOMC Minutes (USD/HIGH; GC, SI, DXY).

The two events that matter most for this call are the EIA product prints on 2026-10-07 and the FOMC minutes on 2026-10-08; the former tests the distillate tightness thesis, the latter tests the rates headwind.

This report is generated automatically from public quantitative and macro data for research and market tracking only. It does not constitute investment advice or a recommendation to trade.

單品種深度報告深度分析

COMEX黃金GC=F 2026-10-04

Gold (GC=F) Deep Dive | 2026-10-04 — Bearish, 4143.1 20-day floor in focus | 2026-10-04

WTI Crude (CL=F) Deep Dive | 2026-10-04 — Bullish Above 88.28, Backwardation vs Soft Tape | 2026-10-04

今日風險地圖

單品種深度報告 深度分析

往期研報封存

⟠ 本月策略回測表現

單品種深度報告深度分析