Crude Oil Spread Trading Strategies

Crude oil spread trading means buying one contract and selling a related one — different months (calendar spreads), crude versus refined products (crack spreads), or WTI versus Brent (the Arb) — to profit from the difference moving rather than the outright price. Margins are a fraction of outright requirements, and the P&L depends on relationships, not direction.

Watch a professional energy desk for a week and you’ll notice something: most of the risk isn’t outright length or shorts. It’s spreads. Directional flat-price bets are what the market makes you take when you can’t find a better-defined relationship to trade. This guide works through the three spread families every crude trader should know, with real contract math for each — and it assumes the foundation laid in our complete guide to crude oil trading.

Why spreads are the professional’s default

Three reasons, in descending order of importance.

Spreads isolate a thesis. “Refining margins are too low into driving season” is a precise, testable view. Expressing it with an outright long in gasoline futures buries that view under crude price risk, macro risk, and OPEC headline risk. The crack spread expresses it directly and nothing else.

Margin efficiency is enormous. Because the legs are correlated, exchanges grant margin credits: a WTI calendar spread can require roughly a tenth of the capital of a naked futures position, and crack spreads earn meaningful offsets against outright product margins. Capital efficiency compounds — the same account can run several uncorrelated spread theses instead of one white-knuckle directional bet.

The volatility is survivable. Flat price moves dollars per day; most spreads move cents. That doesn’t make spreads safe — we’ll get to April 2020 and Metallgesellschaft — but it makes position-sizing errors less instantly fatal.

Family one: calendar spreads

A calendar spread is long one delivery month, short another in the same contract — long September WTI, short December WTI. It’s a pure bet on the shape of the futures curve: tightness near-term versus later. The theory of why curves slope — storage economics one way, scarcity the other — is covered in our piece on contango and backwardation; here’s the trade itself.

Worked example: selling the spread into a glut

It’s early autumn. OPEC+ has been adding barrels for three months, refinery maintenance season is about to cut crude demand, and Cushing stocks have built four weeks running. November WTI trades $66.50, February trades $66.90 — only a 40-cent contango that you think is far too shallow for the surplus forming. You sell the Nov/Feb spread (short November, long February) at −$0.40.

Over six weeks the builds keep printing and the contango deepens to −$1.55. You cover. The spread moved $1.15 in your favor; at 1,000 barrels per spread and $10 per cent, that’s $1,150 per spread. Margin tied up was likely a few hundred dollars rather than the several thousand an outright CL position demands — a return on margin no directional trade of similar risk would have offered. Your risk was the curve flipping tighter: had a pipeline outage cut Cushing supply and squeezed the spread to +$0.20, you’d have lost $600 per spread. Manageable, definable, and completely independent of whether flat price spent those weeks at $60 or $75.

Execution note: trade the exchange-listed spread instrument, not two separate orders. One fill, one bid-ask, no leg risk. Every serious platform supports it.

The bull/bear spread playbook

Desk shorthand: a bull spread is long the near month, short the deferred — it profits when the market tightens and the curve pushes toward backwardation. A bear spread is the reverse and pays when surplus builds. The vocabulary matters because it maps cleanly onto the data flow. Cushing draws week after week? Bull-spread pressure. Refinery maintenance season cutting crude runs while production holds? Bear-spread setup. OPEC+ surprise cut? The front reprices hardest — bull spreads move first and furthest.

What makes calendar spreads the thinking trader’s instrument is that the drivers are narrow, observable, and largely free to follow. The Wednesday EIA report — specifically the Cushing line, since Cushing is the WTI delivery point — is the single most important weekly input. Around it sit a handful of slower dials: the pace of U.S. production growth, export loadings on the Gulf Coast, and the seasonal rhythm of refinery runs, which drop in spring and autumn turnarounds and peak with summer driving demand. None of this requires satellite subscriptions. It requires showing up every Wednesday at 10:30 ET and keeping a simple running view of whether the delivery point is filling or draining.

Two warnings before treating spreads as a solved game. First, the seasonal patterns that made calendar spreads famous — buying summer months into driving season and the like — are the most arbitraged patterns in the market, and their reliability has faded as systematic money crowded in. A seasonal tendency is a tiebreaker, not a thesis. Second, calendar spreads are still directional bets — just on the curve instead of the price. The curve can trend against you for months when the underlying balance shifts, and “the spread always comes back” has the same fatal ring as “house prices only go up.” The 2026 whipsaw made the point twice in four months: spreads that blew out to crisis extremes in March round-tripped by June, punishing latecomers in both directions.

Which months to trade is its own decision. The front pair (e.g., the first and second listed months) is the most liquid and the most responsive to weekly inventory data; spreads further out — six or twelve months apart — express slower, structural views on the balance and move less per week. Beginners belong in the front pair, where the tuition is cheaper and the exit is always open.

Family two: crack spreads

Refineries buy crude and sell products. The crack spread — product price minus crude price — is the market’s live estimate of the gross refining margin, and it trades as an instrument in its own right. When you trade a crack, you are effectively running a paper refinery.

The 3-2-1 math, worked completely

The U.S. benchmark is the 3-2-1: three barrels of WTI yield roughly two barrels of gasoline and one of distillate. Product futures price in dollars per gallon, so convert at 42 gallons per barrel. Say WTI (CL) is $72.00, RBOB gasoline (RB) is $2.40/gal and ULSD diesel (HO) is $2.55/gal:

• Gasoline per barrel: $2.40 × 42 = $100.80
• Diesel per barrel: $2.55 × 42 = $107.10
• 3-2-1 crack = (2 × $100.80 + 1 × $107.10 − 3 × $72.00) ÷ 3 = ($201.60 + $107.10 − $216.00) ÷ 3 = $30.90 per barrel

That $30.90 is the theoretical gross margin per barrel refined. The EIA maintains a clean primer on the mechanics in its introduction to crack spreads, and CME publishes contract-level detail including the ratio conventions.

To buy the full 3-2-1 in futures you’d buy two RBOB and one ULSD contract and sell three CL — both product contracts are 42,000 gallons (1,000 barrels), so the ratio is barrel-for-barrel. Most traders simplify to a 1:1 single-product crack: long one RBOB, short one CL is the gasoline crack; long one HO, short one CL is the diesel (heating oil) crack.

Worked example: long the gasoline crack into driving season

It’s late March. Refinery turnarounds have crimped gasoline output, stocks are below the five-year average, and the switch to summer-spec fuel is approaching — the seasonal setup that has historically firmed gasoline cracks into May. The June gasoline crack (June RB versus June CL) trades at $18.00/bbl. You buy it: long one June RBOB at $2.20/gal, short one June CL at $74.40.

Five weeks later a Gulf Coast refinery outage tightens supply further and the crack reaches $26.50. You unwind both legs. You made $8.50 per barrel on 1,000 barrels: $8,500 per spread set, regardless of the fact that crude itself rallied $3 over the period — your short crude leg lost, your gasoline leg gained more. The tick math: RBOB moves $4.20 per $0.0001/gal, CL moves $10 per $0.01/bbl; quoted as a crack in $/bbl, each $0.01 is $10, same as crude.

What can go wrong is just as instructive: a demand scare (recession headlines, a soft driving season) can crush cracks even while crude rallies — the 3-2-1 fell hard in past cycles when product demand cracked before crude supply did. Crack spreads are economic-cycle trades as much as energy trades. The 2022 diesel episode showed the other tail: post-Ukraine distillate scarcity drove diesel cracks to records, well past $50/bbl — ruinous for anyone short the crack on “mean reversion” logic without a stop.

Family three: the Brent–WTI spread (the Arb)

The third family trades geography instead of time or refining: long one global benchmark, short the other. Brent prices seaborne Atlantic-basin crude; WTI prices barrels delivered inland at Cushing, Oklahoma. The differences in quality, delivery mechanics, and exposure run deeper than most traders realize — our comparison of WTI vs Brent covers the full anatomy — but for the spread trader the essence is: Brent carries the world’s seaborne and geopolitical risk; WTI carries Cushing’s plumbing. The differential between them, quoted on some platforms simply as “the Arb,” is the market’s live price on that distinction.

Brent has spent most of the shale era at a premium of a few dollars, roughly the cost of moving a U.S. barrel to tidewater. The spread earns its keep when that relationship breaks. A Middle East supply scare lifts Brent faster than WTI. A U.S. pipeline outage that strands barrels at Cushing crushes WTI against Brent. An export-infrastructure boom that debottlenecks the Gulf Coast pulls the two grades back together, which is broadly the story of the last decade.

2026 delivered the sharpest lesson in years. When the Strait of Hormuz was effectively closed from late February, the market repriced seaborne supply violently — Brent broke $100 within days and touched roughly $126 in March — while WTI, sitting behind U.S. borders with its own domestic supply, lagged the move. Traders long Brent against short WTI were positioned in exactly the instrument designed for that event: a bet on seaborne risk, hedged against the general level of oil prices. The reversal was just as instructive. As the ceasefire took hold into June and Brent unwound toward the low $70s, the arb compressed — and anyone still long “geopolitical risk” via the spread gave profits back even while headlines stayed scary. The spread prices flows, not fear.

Worked example: fading a stretched arb

Suppose Brent trades $6.50 over WTI after a seaborne scare, against a recent norm near $4. You judge the panic overdone: no barrels are actually failing to load, and U.S. exports are ramping to capture the differential — the physical response that historically drags the spread back toward transport economics. You sell Brent, buy WTI at +$6.50. Three weeks later loadings normalize and the arb settles at $4.25. Both legs are 1,000-barrel contracts, so each $0.01 of spread movement is $10: a $2.25 compression pays $2,250 per spread. The risk was event risk, and it is real: had the scare escalated into an actual supply interruption, the arb could have doubled against you overnight. Size for the tail, not the average week — this spread’s quiet months are deceptive.

Execution options: ICE lists a Brent/WTI futures spread as a single instrument, and CME offers the equivalent via its WTI–Brent products; either fills both legs simultaneously at your differential. One mechanical trap: ICE Brent is cash-settled against the Brent Index while NYMEX WTI is physically delivered at Cushing, and their expiry calendars differ. Hold a spread into the front month’s final days and you are no longer trading a differential — you are managing two different settlement processes. Roll early.

What the margin math actually looks like

Spread margins are set by the exchanges’ portfolio risk models, which recognize that correlated legs hedge each other. The numbers move with volatility, so treat these as representative magnitudes rather than quotes — current requirements are published by CME Group:

The efficiency is the point — and the danger. Because a calendar spread margins at perhaps a tenth of an outright, the temptation is to trade ten spreads where you would have traded one future. Do that and you have rebuilt the risk you thought you removed, concentrated in a single curve view, with worse liquidity in the back legs. Margin relief is the exchange telling you the position is usually quieter. It is not a license to multiply size until it isn’t.

Execution mechanics: how professionals actually enter spreads

Three rules cover most of what separates clean spread execution from expensive improvisation.

Trade the spread instrument, never the legs. Both CME and ICE list calendar spreads, the Brent–WTI differential, and crack combinations as single order-book instruments. You bid or offer the differential itself — say, −$0.42 for a Nov/Feb WTI spread — and the matching engine fills both legs atomically. Legging in with two outright orders exposes you to the market moving between fills, and in a market that can reprice 50 cents in the seconds after a headline, “between fills” is where accounts get hurt. The spread books on the first few WTI calendar months are among the deepest in commodities; there is no execution-quality excuse for legging.

Know your quote convention before you press the button. Calendar spreads are conventionally quoted front minus back, so contango prints negative; but platforms differ, and the Brent–WTI arb flips sign depending on which leg your venue lists first. Every desk has a story about a junior who bought a spread believing they were selling it. Check the sign of the market against the curve before sizing anything.

Mind the calendar. Spreads involving the front month inherit the front month’s expiry mechanics — WTI stops trading around the 20th of the preceding month, and the days into expiry are when delivery-driven distortions are most violent. Unless your thesis is specifically about expiry dynamics, roll or exit with a week to spare.

For traders who want defined-risk expressions of the same views, the options market trades on these structures too: CME lists calendar spread options (CSOs) on WTI, which pay off on the spread itself rather than the flat price, and standard options on futures can be combined into synthetic spread views. Options carry their own pricing dimension — premium, volatility, time decay — which we cover in our guide to crude oil options trading strategies; the practical takeaway is that CSOs let you buy a curve opinion with strictly limited downside, at the cost of paying premium for a market that spends most of its time going nowhere.

When spreads kill: two case studies worth memorizing

Everything above makes spreads sound civilized. Two episodes show what the uncivilized days look like, and every spread trader should know both cold.

April 20, 2020: the day the front month stopped being a spread leg

When the expiring May 2020 WTI contract (CLK20) settled at −$37.63, the flat-price story got the headlines. The spread story was arguably more instructive. The May/June spread — normally a few dimes wide — blew out to nearly $60 intraday, because the May leg had detached from the futures market and become a question about physical storage at a delivery point that had none to offer. Anyone holding a “cheap” long May/short June bull spread on the theory that spreads mean-revert learned that a spread is only as sane as its craziest leg. In the super-contango of that spring, storage economics — the anchor that normally caps how far contango can stretch — simply failed, because there was no marginal storage left to price. The mechanics of why curve extremes happen are covered in our piece on contango and backwardation; the risk lesson here is narrower: never assume a bounded range for a spread whose front leg is approaching physical delivery in a stressed market.

Metallgesellschaft, 1993: the roll is a position

The German conglomerate’s U.S. arm sold customers long-term fixed-price fuel contracts and hedged them by stacking futures in the front months, rolling forward every expiry. The hedge was directionally sound; the structure was a giant implicit calendar-spread position. When the curve flipped into persistent contango, every monthly roll meant selling the cheap expiring month and buying the dearer next month — a steady bleed on more than 100 million barrels of exposure that, combined with the cash-flow strain of margining losses, unwound the firm to the tune of over a billion dollars. The lesson generalizes to anyone who holds futures positions across expiries, from hedge desks to ETF investors: if you roll, you are trading the calendar spread whether you meant to or not. Price the roll before it prices you.

Risk rules for spread traders

Adapted from how energy desks actually manage spread books, scaled to a retail account:

• Size off the tail, not the average. Take the worst multi-week move the spread has made in the past several years — not the typical daily wiggle — and size so that a repeat costs a tolerable fraction of the account. For WTI calendars that tail includes 2020; treat it as admissible evidence, not an outlier to be excluded.
• One thesis, one position. Long the front calendar, long the crack, and short the arb can easily be the same bet (tight U.S. market) worn three ways. Correlated spreads compound like any other correlated positions.
• Stop on the spread, act on the spread. Define exits in differential terms ($−1.55 target, +$0.20 stop) and honor them by lifting the whole structure. Taking off the profitable leg and “letting the other one run” converts a hedged position into a naked directional bet at the worst possible moment — it is the single most common way retail traders destroy a spread book.
• Respect liquidity asymmetry. Front spreads are deep; back-month and product legs are thinner, and exit costs widen exactly when volatility arrives. If the position is big relative to the book’s quiet-day depth, it is too big.
• Event calendar discipline. The Wednesday EIA report, OPEC+ meetings, and expiry week are when spreads gap. Carrying maximum size through a known binary is a choice, not an accident — make it consciously or not at all.

Where crude oil spread trading fits in your operation

Spreads are not a separate discipline bolted onto oil trading; they are the market’s own language for its most tradeable questions. Is storage filling or draining? That’s the calendar spread. Are refiners making money? That’s the crack. Is the risk seaborne or domestic? That’s the arb. A trader who only watches flat price is reading the headline and skipping the article. Start small — one adjacent-month WTI calendar spread is the standard first trade for good reason: deep, cheap to margin, and moved by data you can actually follow weekly. Build the habit of writing down the thesis, the differential entry, the target, and the stop before the order goes in. Six months of that journal, reviewed honestly, will teach you more about how the crude curve actually behaves than any amount of reading — including this. And keep the broader machinery — contract specs, order types, position sizing, the full toolkit — within reach in our complete guide to crude oil trading, because spread trading rewards exactly the traders who treat it as a craft rather than a margin discount.