The WTI–Brent Spread (1987–2026): 39 Years of Divergences Show US Logistics, Not Geopolitics, Drove the Gap

Chart: The WTI–Brent Spread, May 1987 to March 2026

Monthly WTI–Brent spread, in US dollars per barrel

WTI traded at a $1 premium to Brent for 23 years. Then shale broke Oklahoma’s pipelines. The premium became a $16 average discount for 5 years.

Key Takeaway

The structural break occurs at the boundary between two production regimes: from May 1987 through December 2010 the spread averaged +$1.37, with WTI trading at a small premium driven by quality (lighter, sweeter crude) and US refinery proximity. Beginning January 2011 the spread inverted abruptly to −$7.35 in a single month and continued to widen for the next eight months, reaching −$27.31 by September 2011.

Sources: U.S. Energy Information Administration (MCOILWTICO, MCOILBRENTEU). Chart: Eco3min Research.

PNG
SVG
PDF
Free to use with attribution · Cite

How to Read This Chart

The y-axis shows the WTI minus Brent monthly spread in US dollars per barrel. Positive values mean WTI traded above Brent; negative values mean WTI traded at a discount. The horizontal zero line marks parity. The shaded band marks the 2011-01 to 2015-12 shale-glut window, the only period in the sample during which the absolute spread exceeded $13. Spikes outside this window — visible in 2018–2019, 2022, and 2026 — correspond to documented geopolitical episodes that pushed Brent above WTI but did not reach the structural magnitudes observed during the bottleneck years. The dataset itself is described on our WTI Crude Oil price dataset and Brent Crude Oil price dataset pages.

The Geopolitical Misreading of an American Phenomenon

The dominant interpretation in financial commentary holds that when WTI and Brent diverge sharply, the explanation lies in geopolitical events: a Middle East conflict, OPEC+ supply policy, or Russian sanctions. This framing treats the spread as a signal of global supply stress, with the assumption that Brent — as the seaborne international benchmark — reacts faster to international disruptions than WTI.

The 39-year dataset tells a different story. The single largest sustained divergence in the entire sample occurred between January 2011 and June 2013, when the WTI–Brent spread averaged −$16.24 per barrel for 30 consecutive months and reached an extreme of −$27.31 in September 2011. During this window the spread crossed below −$15 in eighteen separate months. In the remaining 437 months of the sample — covering the Gulf War, the Iraq invasion, the Iran nuclear standoff, the Russian invasion of Ukraine, and multiple OPEC+ supply decisions — the spread never crossed −$13. Not once.

Anatomy of the Oklahoma Bottleneck (2011–2014)

The mechanism behind the 2011–2013 spread compression is mechanical and documented in EIA storage reports. Cushing crude oil ending stocks, which had averaged 23.1 million barrels in the EIA Weekly Petroleum Status Report series between 2004 and 2010, rose to 38.1 million barrels by December 2010 and continued to climb through 2012, peaking at 51.8 million barrels in January 2013. . Cushing’s nameplate capacity at the time was approximately 65–75 million barrels of operational storage, according to EIA’s working capacity reports.

The constraint was not absolute capacity. It was the absence of southward outlet pipelines toward the US Gulf Coast refining and export complex, combined with a federal ban on US crude oil exports dating from the 1975 Energy Policy and Conservation Act. Crude that arrived at Cushing from the Bakken and the Permian had two viable destinations: midwest refineries (with limited additional intake capacity) or storage. As storage filled, the price required to clear new barrels into Cushing fell relative to Brent, which traded on water-borne markets with global outlet access.

The first major outlet — the reversal of the Seaway Pipeline from south-flowing to north-flowing in May 2012 — added approximately 150,000 barrels per day of southbound capacity. Subsequent additions (Seaway twin, BridgeTex, Permian Express, Cactus, Keystone XL Phase II) progressively relieved the bottleneck through 2013 and 2014. The spread tracked these additions: from −$22.53 in November 2012, it tightened to −$3.26 by July 2013, six weeks after the Seaway expansion completed. By June 2014, the spread had compressed to −$6.01 despite a Brent price of $111.

A Legitimate Analytical Qualification

A reasonable counter-argument is that the 2011–2013 period coincided with extraordinary geopolitical events that drove Brent higher: the Arab Spring uprisings in late 2010 and 2011, the Libyan civil war that began in February 2011 and removed approximately 1.6 million barrels per day of Brent-priced supply from global markets, and the imposition of Iran nuclear sanctions in mid-2012. The argument runs that these events lifted Brent specifically, widening the spread mechanically.

Three observations qualify but do not overturn the structural reading. First, the WTI–Brent spread began to widen sharply in January 2011 — moving from −$2.30 in December 2010 to −$7.35 in January 2011 — before the Libyan civil war started in February. Second, the Saudi-led OPEC response to the Libyan outage, announced in June 2011, replaced approximately 1.5 million barrels per day of light sweet crude — a capacity match designed precisely to neutralize the Brent supply shock. Third, the spread remained below −$15 well into 2013, more than a year after Libyan production had partly recovered and a full year after Iran sanctions had been priced into Brent. The persistence of the spread cannot be explained by transient geopolitical premia. Internal links: the global pricing reference, including Iran sanctions episodes, is documented on our Brent dataset page, and US dollar effects on commodity pricing are covered in our US Dollar Index dataset.

Has the Spread Normalized? Post-2015 Evidence

The US crude oil export ban was lifted by the Consolidated Appropriations Act of 2016, signed on 18 December 2015. Combined with the buildout of Gulf Coast export terminals (Corpus Christi, Houston Ship Channel) through 2018, this removed the regulatory constraint that had isolated Cushing from global markets.

The post-export-ban era (January 2016 onward, 123 months) shows partial but not complete normalization. The mean spread sits at −$4.19, with a standard deviation of $2.55 — far narrower than the shale-glut era ($7.02) but distinct from the pre-shale era when WTI traded at a +$1.37 premium. The 5.6-dollar gap between pre-shale mean and post-ban mean reflects the structural cost of moving WTI from Cushing to global markets: pipeline tariffs to Gulf Coast terminals, sea freight from US ports to European refiners, and the freight discount required to compensate buyers of US light tight oil relative to Brent-grade specifications.

Three sustained divergence episodes occurred during this normalized era, each with a clear geopolitical driver:

These three geopolitical episodes share a quantitative signature that distinguishes them from the shale-glut period. Each peaked in a narrow band between −$10.31 and −$11.75, a range of less than $2 across three events spread over seven years. None approached the −$27 extreme of September 2011 or even the −$22.53 of November 2012. The geopolitical channel exists, but its amplitude is bounded.

The Cushing–Spread Relationship Is Non-Linear and Regime-Dependent

A naive reading of the bottleneck thesis would expect a simple negative correlation between Cushing crude oil inventories and the WTI–Brent spread: more storage filling, wider WTI discount. The data shows something more nuanced.

The pre-shale fragment behaves as a textbook would predict: when Cushing inventories rose, the WTI discount widened (r = −0.62). The shale-glut era shows a near-zero contemporaneous correlation. The post-ban era shows the opposite sign: lower Cushing inventories correlate with a wider WTI discount.

The interpretation that reconciles these regimes is straightforward. Before 2011, Cushing operated as a normal storage hub: occasional fill-ups produced occasional WTI discounts, with the level of inventory mapping directly onto a small price differential. After 2011, the relevant variable was no longer the absolute inventory level but the available outlet capacity — the ability to move crude out of Cushing toward the coasts. Inventories peaked at 69.4 million barrels in March 2017, well after the bottleneck ended, with the spread sitting at only −$2.26 because pipelines and export terminals were by then absorbing the flow. After 2018, the post-ban spread fluctuates around −$4 as a structural cost of US-to-international logistics, and inventory variation reflects export pulls rather than constraints.

The simple regression of spread on Cushing stocks over the full 2004–2026 overlap yields an R² of 0.08 — the lowest-effort statistical test fails by the standard measure of explanatory power. The structural narrative is supported not by linear correlation but by episode-level documentation, mechanism-level analysis (pipeline capacity, export policy, refinery geography), and the empirical observation that no other 60-month window has produced anything close to the 2011–2015 spread distribution.

Distribution by Era: Three Stable Worlds, One Outlier

Spread distribution by era (1987–2026)

The pre-shale, shale-glut, and post-ban eras occupy distinct, non-overlapping price differential regimes.

Sources: U.S. Energy Information Administration (MCOILWTICO, MCOILBRENTEU). Chart: Eco3min Research.

PNG
SVG
PDF
Free to use with attribution · Cite

Eleven Sustained Divergence Episodes (1987–2026)

The following table lists every continuous monthly window in which the absolute spread exceeded $5 for three or more consecutive months. Driver classifications follow the dominant documented mechanism; “hybrid” labels indicate episodes where both logistical and geopolitical drivers were operating simultaneously.

Eleven sustained divergence episodes, 1987–2026

Logistical episodes (gold) include the −$27 extreme. Geopolitical episodes (red) cap below −$12.

Sources: U.S. Energy Information Administration (MCOILWTICO, MCOILBRENTEU). Driver classification: Eco3min Research.

PNG
SVG
PDF
Free to use with attribution · Cite

Driver classifications are based on documented sources cited in the Methodology section. “Hybrid” indicates the simultaneous presence of meaningful logistical and geopolitical factors during the episode window. Episodes are not predictive: past distributions describe historical patterns, not expected future outcomes.

Historical Turning Points

January 2011 — The Spread Inverts

The WTI–Brent spread moved from −$2.30 in December 2010 to −$7.35 in January 2011 — the first crossing of the −$5 threshold in the entire sample to that point. Cushing inventories that month stood at 38,839 thousand barrels, already at the high end of historical norms. This single-month move marked the beginning of what would become the largest sustained divergence in 39 years of data.

September 2011 — The Extreme

The spread reached −$27.31, with WTI at $85.52 per barrel and Brent at $112.83 per barrel. Cushing inventories had pulled back to 30,058 thousand barrels — counter-intuitively, the spread peaked when storage was draining locally but global Brent prices remained elevated following the Libyan supply disruption and Saudi spare-capacity offset. The mechanism was forward expectation: market participants priced the structural inability to move incremental shale crude out of Cushing toward refining demand, not the contemporaneous storage level. The analytical counterpart to this passage sits in the mapping of the ounce-to-barrel ratio.

May 2012 — The Seaway Reversal

The Seaway Pipeline, originally built to transport crude north from the Gulf Coast to Cushing, was reversed to southbound flow on 17 May 2012, adding approximately 150,000 barrels per day of outlet capacity. The spread that month was −$15.68 — already well above the September 2011 extreme. Subsequent expansions through 2013 and 2014 (Seaway Twin, BridgeTex, Keystone Phase II) progressively relieved the constraint. By July 2013 the spread had compressed to −$3.26.

December 2015 — The Export Ban Lifts

The Consolidated Appropriations Act of 2016, signed 18 December 2015, repealed the four-decade restriction on US crude oil exports established by the 1975 Energy Policy and Conservation Act. The spread that month was −$0.82. While the structural pipeline capacity continued to expand thereafter, the regulatory constraint that had isolated US-priced crude from international markets ceased to operate.

November 2018 — Iran Sanctions and the Post-Ban Geopolitical Channel

The Trump administration reimposed nuclear-related sanctions on Iran effective 5 November 2018, pushing Brent to a sustained premium over WTI. The spread that month was −$7.79. The episode would persist for 11 months, peaking at −$10.49 in May 2019. This was the first sustained divergence in the post-ban era and demonstrated that a Brent-specific supply shock (sanctioned Middle Eastern crude is Brent-priced) can pull the spread into the −$10 to −$11 range — but no further.

March 2026 — Current Observation

The spread closed March 2026 at −$11.75, with WTI at $91.38 and Brent at $103.13. Cushing inventories at the most recent reading (January 2026) stood at 24,498 thousand barrels — well below the post-ban mean of 40,593 Kb. The combination signals a geopolitical-driven episode (consistent with documented Iran/Hormuz transit disruptions in early 2026) rather than a logistical re-emergence. The episode has lasted three months and is the third post-ban geopolitical sustained divergence in the dataset. It approaches but has not crossed the −$13 threshold that has held in the 407 months outside the 2011–2015 window.

Methodology

The dataset combines two FRED price series at monthly frequency, with a third EIA series for inventory context. All observations are end-of-month averages of daily spot prices, retrieved from the public FRED and EIA databases between May 7 and May 12, 2026. Reproduction code is provided below.

Episode Selection Criteria

Sensitivity analysis. Varying the threshold and duration produces the following episode counts: at $5 / 3 months → 11 episodes (84 months total); at $5 / 6 months → 5 episodes; at $7 / 3 months → 3 episodes; at $10 / 3 months → 3 episodes. The headline finding “spread never exceeded $13 outside 2011–2015” survives sensitivity at thresholds $13, $15, and $20 — all return zero months outside the shale-glut window. At threshold $12, also zero months outside. At threshold $10, four months outside (the three geopolitical-episode peaks documented in Section 3 plus October 2018 during the Iran sanctions period). The headline finding is therefore robust at thresholds at or above $12.

Filter definitions. “Pre-shale era” = date < 2011-01-01 (284 months). “Shale-glut era” = 2011-01-01 ≤ date < 2016-01-01 (60 months). “Post-export-ban era” = date ≥ 2016-01-01 (123 months in current sample, through 2026-03). These boundaries follow regulatory and structural breakpoints (US Bakken/Eagle Ford production ramp; Consolidated Appropriations Act of 2016), not optimized statistical thresholds.

Dataset Design

Python Reproduction Code

# Reproduce this dataset from primary sources
import pandas as pd
from fredapi import Fred

fred = Fred(api_key="YOUR_FRED_API_KEY")

# Fetch WTI and Brent monthly
wti = fred.get_series("MCOILWTICO").rename("wti")
brent = fred.get_series("MCOILBRENTEU").rename("brent")

# Merge and compute spread
df = pd.concat([wti, brent], axis=1).dropna()
df["spread"] = df.wti - df.brent
df["abs_spread"] = df.spread.abs()

# Identify sustained divergence runs
df["div_flag"] = (df.abs_spread > 5).astype(int)
# [...full code in CSV documentation]

Dataset Download & Reproducibility

467 observations · Monthly · May 1987 – March 2026 · Licensed under CC BY 4.0.

Data Sources & References

• Primary U.S. Energy Information Administration. Crude Oil Prices: West Texas Intermediate (WTI) — Cushing, Oklahoma. FRED series MCOILWTICO. Retrieved 12 May 2026 from https://fred.stlouisfed.org/series/MCOILWTICO.
• Primary U.S. Energy Information Administration. Crude Oil Prices: Brent — Europe. FRED series MCOILBRENTEU. Retrieved 12 May 2026 from https://fred.stlouisfed.org/series/MCOILBRENTEU.
• Primary U.S. Energy Information Administration. Cushing, OK Ending Stocks of Crude Oil. Series MCRST_YCUOK_1. Retrieved 12 May 2026 from EIA Petroleum Data Navigator.
• Research Borenstein, S. and Kellogg, R. (2014). “The Incidence of an Oil Glut: Who Benefits from Cheap Crude Oil in the Midwest?” The Energy Journal, Vol. 35, No. 1.
• Research Kilian, L. (2016). “The Impact of the Shale Oil Revolution on U.S. Oil and Gasoline Prices.” Review of Environmental Economics and Policy, Vol. 10, No. 2, pp. 185–205.
• Reference U.S. Energy Information Administration. “U.S. Crude Oil Production Forecast: Analysis of Crude Types.” May 2014 Special Report.
• Reference Consolidated Appropriations Act of 2016, Public Law 114-113, Division O, Section 101 (repeal of crude oil export ban). 18 December 2015.
• Reference Energy Policy and Conservation Act of 1975, 42 U.S.C. §6212 (original crude oil export restrictions).

Methodological Limitations

• Monthly frequency averages out daily volatility. The April 2020 negative WTI futures settlement (−$37.63 on 20 April 2020) does not appear in the monthly series, which shows WTI at $16.55 for that month. Daily granularity would change the recent-extreme picture and is available via DCOILWTICO and DCOILBRENTEU for replication.
• Cushing inventory data is available only from January 2004 onward in the monthly series, limiting overlap with the spread to 265 observations. Pre-2004 Cushing inventory analysis is qualitative, based on EIA Working Capacity Reports.
• Driver classification of episodes is necessarily ex-post and draws on documented mechanisms identified after the fact. The “hybrid” label, in particular, is judgmental — though restricted to episodes where multiple documented factors operated within the window.
• The sample includes a single shale-glut episode (2011–2015), giving n=1 for the central observation. Strong inference about the bottleneck mechanism rests on the documented physical and regulatory constraints, not statistical replication.
• The dataset reflects spot prices at delivery points. It does not capture freight differentials, quality adjustments for specific buyers, or the WTI Midland and other regional pricing nodes that have developed in the post-shale era.
• The current observation (March 2026) is the most recent monthly close available. Updates to the spread during ongoing geopolitical events (e.g., the Iran/Hormuz episode) are visible in the daily WTI and Brent series before the monthly close is recorded.

Frequently Asked Questions

Source

Related Eco3min Research

• WTI Shocks Amplify Existing Inflation — How oil price changes interact with the inflation regime
• US Inflation Is Not Linear — Five episodes account for 80% of cumulative purchasing power loss
• WTI Crude Oil Price Dataset — Daily and monthly time series for the US benchmark
• Brent Crude Oil Price Dataset — Daily and monthly time series for the international benchmark
• US Dollar Index Dataset — The dollar’s role in commodity pricing
• Real Crude Oil Price Dataset — CPI-adjusted long-run oil price history