Seasonal Patterns in Crude Oil Markets

Crude oil seasonal patterns are real but weaker than most traders assume. WTI tends to firm from March through May as refineries exit maintenance and ramp gasoline output, and it tends to sag in November and December. The catch: monthly volatility runs five to twenty times the size of the average seasonal edge, so seasonality works as a filter, never as a standalone signal.

That is the honest version, and it is worth stating up front because most of what gets written about oil seasonality is either a four-paragraph “prices rise in summer” sketch or a chart with no explanation of what is driving it. This article walks through the long-run data, the physical machinery behind the calendar — turnarounds, blend switches, driving season, heating demand — and then gets to the part that matters: what a trader can actually do with it. If you are still building your foundation, start with our complete guide to crude oil trading and come back; this piece assumes you know what a CL contract is.

What the Data Actually Shows About Crude Oil Seasonal Patterns

Strip out the folklore and look at the numbers. One widely cited long-run study of WTI spot prices from 1986 through 2020 found an average monthly return of about 0.6%, with a consistent cluster of strength in March through May and consistent weakness in November and December. Academic work covering Brent and WTI from 1983 to 2017 reached a similar conclusion: abnormally positive returns in March, April, and August; abnormally negative returns in October and November. Two independent datasets, two methodologies, same shape. The pattern exists.

Now the part that gets left out of the seasonal-chart sales pitch. In that same 1986–2020 sample, monthly standard deviations ranged from roughly 6.7% to 14.2%. Read that against a 0.6% average monthly return. The noise is an order of magnitude larger than the signal. A “reliable” seasonal month can and regularly does move against the pattern by 10% or more — May 2019 was a textbook seasonal-strength month that delivered multiple 4% down days as inventories built and the U.S.-China trade war hit demand expectations. Anyone who bought crude that spring purely because the seasonal chart pointed up got carried out.

Here is the calendar as the long-run data describes it, with the physical driver behind each phase:

One more wrinkle: the pattern you can observe in spot prices is not the pattern you can trade. A futures position has to be rolled, and roll yield — positive in backwardation, negative in contango — can overwhelm the seasonal tendency. The USO ETF is the classic demonstration: its month-by-month seasonal profile looks materially different from spot WTI’s because years of contango bleed reshaped its returns. If that sentence didn’t fully land, read our piece on how to read and trade the oil futures curve — curve shape and seasonality interact constantly, and you cannot evaluate one without the other.

The Machinery Behind the Calendar: Turnarounds and Blend Switches

Seasonality in crude is not mystical. It is the shadow cast by the refining system, which runs on a rigid annual schedule.

Start with maintenance. U.S. refineries schedule their heavy planned work — turnarounds that can take a unit offline for four to eight weeks — in the shoulder seasons when product demand is weakest: late winter into early spring, and again in early autumn. Planned maintenance typically peaks in late February and March, and a full turnaround comes roughly once every four years for a given refinery, so something like a quarter of the fleet cycles through major work each spring. While units are down, refineries buy less crude. That is one reason U.S. crude inventories reliably build in the first quarter: production keeps flowing, but the buyers of last resort are partially offline.

Then comes the switch. The EPA requires lower-volatility summer-blend gasoline at terminals by early May and at retail by June 1. Summer blend is more expensive to make and the changeover forces terminals to draw down winter-grade stocks first. Refineries exiting turnaround in April and May therefore ramp crude runs hard and all at once — every plant is racing to build summer gasoline supply before Memorial Day. That synchronized ramp in crude demand is the physical engine behind the March–May price strength in the seasonal data. It is not “summer is coming” sentiment; it is refiners bidding for cargoes.

The autumn mirror image runs the other way. Driving season ends at Labor Day, refineries take September–October maintenance, crude runs drop, and the market carries that reduced demand straight into the November–December seasonal trough. Layer on a purely financial effect — banks, funds, and physical traders trimming inventory and risk into year-end — and you get the most consistent seasonal weakness in the dataset.

The Summer Driving Season: Myth vs. Data

Ask a casual observer when to be long oil and you will hear “summer, obviously — driving season.” The data says the opposite is closer to the truth. By the time the first minivan hits the interstate on Memorial Day weekend, the crude story is largely over.

The reason is timing. Refineries buy crude four to eight weeks before the gasoline made from it reaches a pump. The crude market prices driving season in March, April, and May, while refiners are stocking up — which is exactly where the seasonal strength shows up. June through August returns are far less impressive and far more erratic. In the long-run monthly data, part of the summer’s positive average comes from a handful of extreme years rather than a steady tendency, and August in particular carries some of the highest monthly volatility of the year, thanks to Gulf of Mexico hurricane risk sitting right on top of peak demand.

So the driving-season trade, to the extent it exists, is a spring trade. Buying crude in July because “people are driving” is buying a story the market priced in months earlier. U.S. gasoline demand does peak in July and August — the BLS’s work on gasoline price cyclicality documents the demand curve clearly — but crude prices anticipate; they do not react to things everyone can see on a calendar.

Gasoline Carries the Seasonality. Crude Borrows It.

Here is the cleaner way to think about the whole subject: the strong, reliable seasonality in the petroleum complex lives in the products, not in crude itself. Gasoline demand swings hard with the calendar. Heating oil demand swings with winter. Crude demand — global, diversified across gasoline, diesel, jet, petrochemicals, and power — is comparatively smooth. Crude inherits a diluted version of its products’ seasonality.

You can see this directly in the crack spread. The 3-2-1 crack (three barrels of crude against two of gasoline and one of distillate) reliably widens into spring and early summer as gasoline strengthens, and narrows in autumn. RBOB gasoline futures carry their own seasonal quirks on top — CME Group’s primer on gasoline futures flags the summer-blend specification change and hurricane exposure as drivers that exist independently of crude. This is why experienced energy traders who want seasonal exposure often express it in RBOB, heating oil, or crack spreads rather than flat-price crude: the seasonal signal-to-noise ratio is simply better one step down the barrel.

The same logic, amplified, applies next door in gas. If you want to see what commodity seasonality looks like when it is genuinely strong, read our companion piece on seasonal patterns in natural gas — a market where winter demand can run double summer demand and the injection/withdrawal calendar is the whole game. Crude’s seasonality is a whisper by comparison, and calibrating your expectations to that difference will save you money.

Inventory Seasonality and the Shape of the Curve

Crude inventories follow their own calendar, and the futures curve prices it. U.S. crude stocks typically build through the first quarter — refinery maintenance season — peak somewhere in spring, then draw through summer as refinery runs hit their highs. Gasoline inventories run the opposite cycle: built up through winter and early spring, drawn down through driving season. The EIA’s overview of what drives crude prices is worth reading on this point: inventories are the balancing item between supply and demand, and their seasonal rhythm is one of the few genuinely predictable things in this market.

Because the curve prices storage, seasonal inventory swings leave fingerprints on calendar spreads. Front-month WTI spreads tend to be at their weakest relative to deferred months during the Q1 build (ample prompt supply, storage filling) and firmest during the summer draw season. Traders who watch the CL1-CL2 spread through the year are effectively watching seasonality expressed in its purest tradable form — spreads strip out most of the flat-price noise that makes outright seasonal trades so treacherous. The spread market also tells you when seasonality has been overrun by something bigger: when the whole curve flips shape in a week, as it did during the 2026 Hormuz crisis, the calendar is no longer driving.

Why Crude Oil Seasonality Has Weakened

Compare a seasonal chart built on 1986–2005 data with one built on the last fifteen years and the newer one looks flatter and messier. Several structural changes did that, and they are worth understanding because they tell you how much weight the old averages deserve.

Shale changed the supply response. Before the shale era, supply was slow and lumpy — megaprojects with decade lead times. U.S. shale brought a supply source that responds to price in months, not years. When spring demand firms, incremental barrels show up faster than they used to, shaving the seasonal amplitude.

The export ban ended. Until December 2015, U.S. crude was largely trapped onshore, which made WTI hostage to the domestic refinery calendar. With exports flowing — routinely around 4 million barrels per day in recent years — a soft domestic maintenance season no longer strands barrels at Cushing to the same degree. WTI has become a globally arbitraged grade, and global demand is less seasonal than U.S. gasoline demand.

Demand’s center of gravity moved. The seasonal folklore was written when OECD gasoline and heating demand dominated. Growth since has come from non-OECD transport, petrochemicals, and jet — flows with different or offsetting calendars. Two hemispheres with opposite summers also net against each other more than they did when the U.S. was an outright majority of demand growth.

Financialization arbitrages the obvious. Once a calendar pattern is published in every commodity handbook, systematic money positions ahead of it. Whatever edge survives is the residual after CTAs and index flows have had their bite. The most-cited academic finding on oil seasonality notes that in many specifications the month-of-year effect is statistically insignificant — which is what you would expect of a pattern that has been partially traded away.

And shocks simply dominate. The last several years are a museum of seasonal patterns flattened by events: the March–April 2020 price war and COVID collapse (including WTI’s −$37.63 settlement on April 20, 2020 — in the middle of the “seasonally strong” window), the 2022 invasion of Ukraine, and most recently the February–June 2026 Strait of Hormuz crisis, which took Brent from around $70 to well north of $100 in weeks and then gave most of it back by late June under the ceasefire. No seasonal average survives contact with a supply shock of that size. In 2026, the seasonally strong spring happened to coincide with a war premium — anyone crediting the seasonal chart for that rally is fooling themselves.

Trading Seasonal Patterns: Two Worked Examples

Suppose you still want the exposure. Fine — but structure it so the seasonal edge is the tiebreaker, not the thesis, and size it so a 10% adverse month doesn’t end the conversation. Two illustrations with honest math.

Example 1: The spring crack spread

The most defensible seasonal trade in the complex is long gasoline cracks into spring, because it isolates the strongest seasonal driver (gasoline supply tightness during turnaround season) from flat-price risk. Say it is mid-February. RBOB for May delivery trades at $2.10 per gallon; May WTI trades at $70.00. The gasoline crack is $2.10 × 42 − $70.00 = $18.20 per barrel. You buy one RB contract (42,000 gallons) and sell one CL contract (1,000 barrels). If the crack widens to $25.00 by late April — the kind of move turnaround season regularly produces — you have made $6.80 per barrel on a 1,000-barrel spread: $6,800 before costs. If a recession scare knocks $10 off flat price along the way, you barely feel it; both legs fall together. What kills this trade is refinery capacity coming back faster than expected or a demand shock specific to gasoline — real risks, but narrower and more analyzable than “where will oil go?”

Example 2: Fading Q4 weakness, with numbers

The November–December softness is the most consistent pattern in the data, so consider the short side. You sell one December CL contract at $78.00 in early November with a stop at $81.00. CL moves $10 per $0.01, so a $1.00 move is $1,000 per contract: you are risking $3,000 to capture a seasonal tendency whose long-run average monthly edge is well under $1,000 per contract per month. Read that sentence again — that is the arithmetic problem with outright seasonal trades. The average edge is small relative to any survivable stop distance, which means the seasonal tendency cannot carry the trade alone. It needs confirmation: a curve in contango and building inventories, bearish momentum on the charts (our guide to crude oil technical analysis covers the tools), and no live geopolitical bid. With three of those aligned, the seasonal lean is a legitimate tailwind. Alone, it is a coin flip with a story attached.

Hurricane Season: The Wildcard Written Into the Calendar

One seasonal factor deserves its own treatment because it works differently from the demand cycle: Atlantic hurricane season, June 1 through November 30, with the dangerous peak from mid-August to mid-October. The Gulf Coast is the densest concentration of energy infrastructure in the hemisphere — offshore production platforms, roughly half of U.S. refining capacity, and the export docks that move millions of barrels a day. A major storm can hit supply and demand at once, and not always in the direction people expect.

Here is the asymmetry that trips up newer traders: a hurricane that shuts offshore production is bullish crude, but a hurricane that primarily floods refineries is often bearish crude — refineries offline means less crude buying — while being violently bullish gasoline. Harvey in 2017 was the canonical case: WTI barely moved and briefly sagged while gasoline cracks exploded, because the storm took out demand for crude (refining) rather than supply of it. The seasonal implication is that August–October carries a persistent risk premium and elevated implied volatility rather than a directional bias. It shows up in the data as high monthly standard deviations, not high average returns. Traders position for it with options and crack spreads, not outright length bought “because it’s hurricane season.”

The Recurring Calendar Beyond Weather

Weather and demand aren’t the only rhythms. The oil market runs on a repeating institutional calendar that creates its own micro-seasonality, and professionals plan their weeks around it.

The EIA’s Weekly Petroleum Status Report lands every Wednesday at 10:30 a.m. Eastern (Thursdays after Monday holidays), with the API’s survey the prior afternoon. Intraday volatility clusters around that release with clockwork regularity — spend a month watching the 10:30 candle and you will never schedule anything else for Wednesday morning again. Monthly, the OPEC, IEA, and EIA outlook reports drop in a cluster mid-month, each capable of resetting the balance narrative. OPEC+ ministerial meetings — historically clustered in early June and late November/early December — put decision risk right on top of the seasonal inflection points, which is not a coincidence: the group meets ahead of the summer and winter demand seasons to set supply for them. A November OPEC+ meeting that disappoints on cuts, landing in the seasonally weakest stretch of the year, is how you get the kind of December selloffs the long-run data remembers.

Futures mechanics add one more layer. WTI contracts expire around the 20th of each month, three business days before the 25th-of-the-month delivery cycle cutoff, and spread activity concentrates in the days before expiry as physical players square positions and index money rolls on its published schedule. None of this is seasonality in the almanac sense, but it is the same phenomenon at higher frequency: predictable calendar structure creating predictable windows of flow and volatility. The traders who make money from calendars mostly make it here, at the weekly and monthly scale, not from owning crude in April because a chart of forty-year averages slopes upward.

A note on January, since it confuses people: winter is peak season for heating fuels, so why is crude soft? Because heating demand is a distillate and natural gas story, and it is regional. Distillate cracks and heating oil futures carry the winter bid; crude gets only a diluted pass-through, and the gasoline side of the barrel is in its weakest demand stretch at the same time. The two roughly cancel. Winter is when product traders earn their year while crude flat-price traders mostly wait for spring.

How Professionals Actually Use Seasonality

On a commodities desk, seasonality functions as context, not conviction. In practice that means a few specific habits.

First, seasonal expectations are built into the analysis baseline. When the EIA reports a 4-million-barrel crude build in March, the desk question is not “is that bullish or bearish?” but “how does that compare to the normal March build?” A build smaller than the seasonal norm is bullish even though the headline says supply grew. Deseasonalizing the data is the single most practical use of everything in this article.

Second, seasonality sets the burden of proof. A bullish crude idea in April needs less supporting evidence than the same idea in November, because one swims with the historical current and one against it. That is a sizing and patience input, not an entry signal.

Third, the seasonal trade gets expressed in the instrument where the pattern actually lives — cracks, product spreads, calendar spreads — rather than outright flat price, for the signal-to-noise reasons covered above.

And fourth: sample-size humility. Forty years of data is forty observations of each calendar month. Remove the three or four macro-shock years that dominate the averages and the “pattern” in several months quietly disappears. Anyone selling you a seasonal system with two-decimal win rates is selling curve-fit noise.

The Bottom Line

Crude oil seasonal patterns are real, physically grounded, and modest. Spring strength driven by the refinery calendar is the most tradable tendency; November–December weakness is the most statistically consistent; everything else is noise-dominated. The patterns have weakened as the market globalized and financialized, and any given year’s price path is set by supply shocks, OPEC decisions, and macro — with the calendar contributing a lean, not a law. Treat seasonality as one input in a stack, express it in spreads where possible, and never let a seasonal chart talk you out of respecting a stop. For the full framework this fits into — fundamentals, curve, technicals, and execution — go back to our complete guide to crude oil trading.