Key takeaways
- Brent futures settled about $90.87 on Monday (Aug. 17), WTI about $84.50, per the Reuters settlement tape. Screens were still around $91 Tuesday. That is not EIA’s number. EIA’s August Short-Term Energy Outlook (released Aug. 11) has Brent averaging about $85 in Q3 2026 and $69 in 2027.
- The gap is a risk premium on seaborne crude. EIA raised Middle East shut-in estimates because Strait of Hormuz transits stay “severely constrained” through August in their base case. WTI prices Permian barrels on U.S. pipe. Different geography, $6-plus spread.
- BLS already printed energy +14.7% year over year in July CPI. Retail gasoline in the STEO averages $3.78/gal for 2026. That is how $91 oil shows up in COLA guesses and in the long end of the Treasury curve.
Brent crude oil price is the ticker people type when gasoline jumps and the 30-year makes a 19-year high on the same morning. It is a North Sea marker that became the world’s seaborne benchmark. It is not the Fed funds rate, and it is not West Texas Intermediate. Mixing those three is how you get a take that “Warsh hiked oil.”
The useful split this week: what the screen is paying for a barrel that has to float, versus what EIA thinks the quarter will average if Hormuz traffic normalizes on their timetable.
Spot vs forecast
Monday settlement (Reuters): Brent +$2.35 to $90.87, WTI +$2.10 to $84.50. Tuesday the Brent CFD/front month was still hugging $91. Three sessions without a down close on some desks. Fine. That is a day.
EIA, forecast completed Aug. 6, published Aug. 11: Brent spot average around $85/b in 3Q26 — $11 higher than the July STEO. Then $78 by 4Q26 in the PDF detail, $69 as the 2027 annual average. Full-year 2026 Brent in the overview table: $87. Source: EIA Short-Term Energy Outlook, August 2026 and the August STEO PDF.
$91 spot versus $85 quarterly average means the market is paying for more disruption than the agency’s mean path, or the $85 already embeds a quarter that started cheaper. Both can be true. EIA also said global inventories fell 4.2 million b/d on average in 2Q26 and will fall another 3.8 million b/d in 3Q26. A draw at that pace is why they lifted the price path at all.
Next STEO: Sept. 9. Do not treat a Tuesday tick as the new official forecast.
Brent is not WTI
Brent (and the ICE Brent complex) is waterborne crude linked to the Atlantic basin and, by arbitrage, to barrels that used to exit the Gulf via Hormuz. WTI Cushing is a landlocked U.S. grade. U.S. production in the STEO: 13.8 million b/d in 2026, 14.2 million in 2027. Exports have been strong enough that EIA has U.S. commercial crude inventories below the five-year (2021–2025) low through end-2026.
A $6–$7 Brent–WTI spread is the map. Seaborne risk in the Gulf of Oman does not strand Midland in the same hour. Refiners on the Gulf Coast still feel product cracks and export nets. Midwest retail can lag Brent and lead WTI. If someone quotes “oil” without naming the contract, they are not helping you.
Kpler ship-tracking, as relayed on the wires: five commodity vessels through Hormuz on Saturday, zero on Sunday, versus 31 the prior weekend. That is a flow print, not a UN resolution. EIA’s language is “severe constraints on Strait of Hormuz transits,” assumed through August, most regional production back near pre-conflict averages in early 2027, leftover disruption about 0.6 million b/d through end-2027. Use that. Do not write that the Strait is a binary open/closed switch because a politician said the word “open.”
What EIA actually assumed
The August STEO is an assumption stack, not a satellite.
- Hormuz constraints persist through August, then traffic increases later (their words).
- Most shut-in Gulf production restored in early 2027.
- Residual ~0.6 mb/d disruption through 2027.
- U.S. net crude imports stay below average through 2027 because the world still wants U.S. barrels.
If night-time Hormuz loadings stop, or Bab el-Mandeb becomes the next choke, the $85 quarterly mean is scrap. If a ceasefire puts VLCCs back on the usual daylight schedule, the $91 risk premium dies faster than the 2027 $69 line. Options markets are priced for that two-way jump. A flat “oil to $120” thread is not an STEO.
OGJ’s read of the same STEO had Hormuz crude+liquids at 4.9 mb/d in 2Q26 versus 21.6 mb/d in 4Q25, and July shut-ins around 5.5 mb/d. Saudi East-West pipe to Yanbu and higher Bab el-Mandeb volumes are the reroute, not a new ocean.
Gasoline, CPI, the 30-year
STEO retail gasoline: $3.78/gal average for 2026, $3.29 in 2027 (was $3.10 in 2025). Wholesale gasoline and diesel were revised up 5.9% and 8.5% for 2026 versus the July book. That is the pump, with a lag, not the ICE pit.
BLS July CPI: all-items +3.4% year over year; energy index +14.7%; gasoline (all types) +24.6% in the TED breakdown. The FOMC’s July 29 statement already blamed energy supply shocks and Middle East uncertainty for inflation still above 2%. Same week the 30-year CMT printed 5.31% on Aug. 17. Long bonds do not need a hike if gasoline is running 25% year over year. That curve: 10-year Treasury yield.
COLA 2027 uses third-quarter CPI-W, not Brent. Energy still sits in that basket. August and September CPI will decide whether the 3.1% July-only sketch holds: Social Security COLA 2027.
Jackson Hole’s printed topic is payments, not oil. Warsh can still mention energy. The vote is still Sept. 15–16: Jackson Hole 2026.
U.S. oil vs U.S. gas
Do not mash Henry Hub into Brent. STEO Henry Hub: $2.87/MMBtu in 3Q26, $3.44 for calendar 2026, inventories headed for 3,985 Bcf in October — most gas in storage before winter since 2016, per EIA’s Aug. 11 press note. LNG maintenance (Freeport) cut feedgas. Oil is a Strait problem. Gas is a U.S. storage problem. Opposite tapes.
Electricity is the AI footnote. EIA cut Texas load growth in 2027 to 6% from 14% after the governor’s Aug. 3 pause on new data centers. Hyperscaler capex still needs megawatts. A GPU quarter does not ship if the interconnect and the substation are both late: AI semiconductor and optical stocks.
What to watch
- Weekly EIA petroleum status (inventures, imports, SPR). Commercial crude below the five-year low is the domestic squeeze; SPR draws are a separate political lever.
- Actual Hormuz transits, not speeches. Five-then-zero weekends matter more than a “waterway is open” clip.
- Brent–WTI spread. Compression = seaborne fear fading or Cushing tightness. Blowout = Gulf barrels still stuck.
- Sept. 9 STEO. If they keep $85 Q3 while spot lives at $91, the agency is fading the premium on purpose.
- Crack spreads (gasoline/diesel vs crude). CPI energy can stay hot even if Brent mean-reverts, if refiners are short product.
$91 is a contract. $85 is a quarterly mean with Hormuz still impaired in the model. $69 is 2027 if the model’s recovery happens. Trade the difference, or ignore the ticker and watch the pump. Those are not the same oil.
Education, not a futures recommendation. Settlements move. EIA forecasts are assumptions. Gasoline in your zip code is AAA/EIA weekly, not ICE Brent.
