JP Morgan Is at a Loss to See Where Oil Goes From Here

Energy News Beat | September 18, 2026JPMorgan’s commodities desk just did something Wall Street almost never does in public. It took the baseline off the table.

In the September 17 Oil Markets Weekly note, titled “Dry powder,” Natasha Kaneva, Lyuba Savinova, and Artem Fakhredtinov wrote the line that has been circulating all day: “For the first time since the start of the Iran conflict, we don’t have a baseline view. We simply don’t know how to model the endgame.”

That is not a colorful quote. It is an admission that the political variable that sets the oil price is no longer inside any bank model. The physical market still has a map. The war does not.HFI Research posted the first page of the note on X this morning, and that page is the cleanest primary source available to the public.

The rest of the argument has to be reconstructed from the same team’s June and early-September work, plus the figures Reuters and others extracted from the Thursday note.

What the note actually says

At the start of the conflict, JPMorgan assumed Washington would not cross a set of “economic red lines”: $100 oil, gasoline near $5 a gallon, 4% headline inflation, and a 5-handle on the 10-year Treasury yield. Those limits were supposed to force a deal to reopen the Strait of Hormuz by June.

Six months later, oil is above $100, and the 10-year has a 5-handle. U.S. regular gasoline is $4.37 a gallon — not $5, but at record seasonally adjusted levels after peak driving season. Diesel is the real problem: $6.31 a gallon, an all-time high, heading into winter with inventories at all-time lows.

The bank’s September fair value for Brent is $90. When the note was written, the market was near $106. JPMorgan’s rule of thumb is that every 1 million barrels per day of supply loss adds about $4 to the price. The $16 premium therefore prices the risk of another 4 million b/d of losses on top of the 10 million b/d already disrupted — not a confirmed, permanent cut.

Brent has since slipped back toward $103 as Saudi East-West pipeline repair talk took some heat out of the tape. That does not change the note. The $16 gap was a risk premium, not a forecast that crude must fall.

The new pressure points listed on page one are not hypothetical. Houthi advances toward Bab el-Mandeb put a second shipping artery at risk. An attack on Saudi Arabia’s East-West pipeline temporarily shut a key bypass for Gulf crude.

Ukrainian strikes hit Russia’s Slavyansk, Taneco (Tatarstan), and Syzran refineries. Russia answered against Ukrainian cities. Volatility is no longer a Hormuz-only story.

The next calendar marker in the note is September 24, when President Trump and President Xi are scheduled to meet in Washington. Absent a diplomatic breakthrough there, JPMorgan says the assumption that the disruption is temporary is “becoming increasingly difficult to sustain.”President Trump has separately said he does not expect the war to end until after the November midterms, and that oil would then “be tumbling downward.” That is a political claim, not a balance sheet. The Fed has already raised rates for the first time in more than three years because inflation has stayed too high. Oil is part of that.

Why prices did not explode

This is the part of the report the headlines skipped.

Since the war began, Brent has averaged about $94, not $130. Global inventories of crude and products are down about 555 million barrels — only one-third of what JPMorgan originally projected. Demand is running about 4.4 million b/d below year-ago levels. The market cleared the shock through demand destruction first, and stock draws second.

Kaneva’s earlier public work explained the price mechanics. When commercial inventories fall, prices usually rise because buyers compete for scarce barrels. When the market clears through weaker demand, the price response works the other way. China is the case study: JPMorgan estimated Chinese gasoline demand destruction at about 180,000 b/d and said 70% of that may not return even after markets normalize, which could cut China’s crude import need by as much as 1 million b/d.

That is why the desk titled the note “Dry powder.” Stocks in China, Europe, Japan, and South Korea are still large enough to cap a near-term spike. The warning is in the last three words: “for now.” If Middle East flows stay at current levels, JPMorgan says fourth-quarter and December prices could sit $7 and $8 above its earlier $80 and $78 markers.

How the oil market can stabilize

Stabilization is not the same thing as peace. The bank’s own September 11 curve piece already walked through that distinction.

There are four offsets, and they have to work together:

• Bypass and reroute what can still move. Dark transits, ship-to-ship transfers, the UAE’s overland options, and a repaired Saudi East-West line are how Gulf barrels have already recovered to a large fraction of pre-war seaborne flow even while Hormuz remains contested. Reports this week said Saudi Arabia is aiming to get at least half of the damaged pipeline back in days and the rest in about six weeks. That is not a full reopening of the Gulf. It reduces the incremental shortage.
• Non-Gulf supply keeps growing. The United States, Brazil, Guyana, Canada, Argentina, and a post-sanctions Venezuela were already the 2027 supply story before the war. Goldman has been explicit that this cohort, plus the UAE outside OPEC discipline, is why a 2027 surplus is still the base case if the Strait normalizes.
• Demand stays the shock absorber. High prices, weaker growth, and China’s EV substitution have already done more work than SPR releases. That is ugly for refiners and truckers. It is how a 10 million b/d disruption did not produce a 2022-style product panic in crude itself.
• Distillate is the exception, and winter will test it.

Inventories stabilize near a stress floor rather than going to zero. JPMorgan put that floor around 7.6 billion barrels. In a “forever conflict” case — Middle East flows stuck near 13.5 million b/d, or about 10 million b/d below normal — the desk still saw 2027 balances close to flat because demand and non-Gulf supply take the rest of the strain. In that scenario, Brent averages about $87 in 2027. In a peace scenario, with Gulf barrels back and a surplus building, the same desk had 2027 Brent around $64.

The counterintuitive result is the one traders hate: a long war does not automatically mean $150 oil if demand keeps shrinking. A sudden peace does not automatically mean $50 oil in the first quarter if product stocks, especially diesel, have to be rebuilt first.

The market is already behaving as if both tails are live. Front-end prices look about $6 too high versus JPMorgan’s physical fair-value work. The back of the curve looks about $10 too cheap if the disruption lasts into 2027. That is a curve problem, not a slogan.LNG is tighter, and it may stay tighter longer

Oil still has barrels that can sneak out. LNG does not sneak as easily.

Before the war, about three LNG cargoes a day left through Hormuz. Qatar alone was roughly a fifth of global supply. After a Qatari tanker was hit in early July, Gulf LNG exports were at a virtual standstill for months. Spot Asian LNG more than doubled to the mid-$20s per MMBtu, the highest since late 2022. This week’s prints have been even higher on some assessments, with Northeast Asia near $28/MMBtu and talk that a cold European winter could push prices toward $40/MMBtu.

A few cargoes are trying again. Bloomberg tracking this week showed at least two LNG shipments making Hormuz and two more doing ship-to-ship transfers off Oman. That is a trickle, not a market. Mitsui O.S.K. Lines’ chairman said this week it is “almost impossible” to carry LNG through Hormuz while the situation lasts.

How LNG stabilizes is a different equation than oil:

U.S., African, and Australian projects have to cover Qatar. IEA work earlier this summer said new non-Gulf supply could add close to 50 bcm in 2026, enough to keep global LNG trade roughly flat if Gulf volumes start coming back in Q3–Q4. They have not, on that schedule. Any further delay is the first down year in global LNG supply.

Demand destruction in Asia is already here. Reuters reporting this week said Asian LNG demand is set to fall for a second year. Chinese discretionary restocking is being pushed into late December or Q1 2027. That is how the market rations scarce molecules.

Europe enters winter with storage well below the five-year norm. Sites are under 70% full versus a five-year average above 80%. Europe and Asia will bid against each other for Atlantic cargoes. That bid is the price.

The glut is delayed, not canceled.

Pre-war consensus had an LNG surplus arriving in 2027–28. Gas Outlook and others now put “abundance” closer to 2029. Qatar’s North Field expansion and undamaged trains can still flood the market later this decade. The war changed the timing. It did not erase U.S. liquefaction that is already built.

Kpler and Rystad, assuming a Q1 2027 Gulf restart and some lost Qatari capacity, see Asian demand recovering toward 280 million tonnes next year with 2027 spot still elevated: Kpler around $14.90/MMBtu, Rystad around $17, Wood Mackenzie $15–20 if Hormuz reopens and above $20 if it does not. Those are not pre-war prices. They are a new plateau until the non-Gulf wave arrives.

Oil can stabilize on reroutes and demand destruction within months. LNG needs either a political reopening of Hormuz or another two winters of U.S. export growth. That is why gas is the tighter market into 2027 even if crude mean-reverts.How far out the rest of the Street is willing to look

JPMorgan is the first major house to say it no longer has a base case. Everyone else still publishes one. The dates on those numbers matter, because many were written in June during the brief U.S.–Iran interim deal that later collapsed.

U.S. EIA (September 2026 STEO): Assumes Middle East flows stay constrained through Q4 2026, with shut-ins averaging 5.7 million b/d in that quarter. Brent averages about $91 for 2026 and $90 in the second half, then $77 by Q2 2027 and $67 in H2 2027 as shut-in barrels return and stocks rebuild. Full-year 2027 average: $74. The EIA is explicit that week-to-week Hormuz volatility will be larger than the forecast path.

IEA versus OPEC on demand, not price: IEA sees 2026 demand falling about 1.6 million b/d, then growing 2.4 million b/d in 2027. OPEC still has 2026 demand up 0.38–0.6 million b/d depending on the vintage, and 2027 up about 2.2 million b/d. That gap is the whole argument about whether this shock is cyclical or structural.

• Goldman Sachs: Has been the most scenario-driven. After the June deal, it cut 2027 Brent toward $75–$80. In September it raised December 2026 to $85 and 2027 to $80, with a $120 upside case if 2027 Gulf output stays 4 million b/d below pre-war levels and an $80 downside if exports normalize. A still-worse path in the June work had 2027 at $140 if Hormuz stayed closed all year. Goldman is not “at a loss.” It is publishing a fan chart. JPMorgan just refused to pick a center line.
• Morgan Stanley (June vintage): $90 in Q3 2026 and $80 in Q4, with a slower physical recovery: 50% of lost production back by September, 80% by December, the rest in early 2027. That timetable is already late.
• Citi (June vintage): The dove. Q3 2026 $75, Q4 $70, 2027 average $65. Those numbers assumed the peace deal held. They are the lower bound of the old consensus, not a live war forecast.
• JPMorgan’s own last published path, before it withdrew the baseline: Brent $86 in Q3 2026, $80 in Q4, $78 at year-end, then a slide toward the low $60s in H2 2027 if peace and surplus arrive. The September 11 “forever war” overlay raised 2027 to $87. The September 17 note says even that overlay is now guesswork.

Put the dates in one place, and the Street is not disagreeing about chemistry. It is disagreeing about a calendar:Weeks:

• Saudi pipeline repairs, the September 24 Trump–Xi meeting, and whether a few more LNG cargoes get out of Ras Laffan.
• Through winter 2026–27: Distillate stocks and European gas storage decide whether crude’s “dry powder” argument survives.
• Q1 2027: The earliest date most LNG houses will underwrite a meaningful Qatari restart.
• H2 2027: EIA and the pre-collapse bank consensus still see prices rolling over if Gulf barrels and new non-OPEC supply show up together.
• 2028–29: The delayed LNG glut. That is the first window where the market can be long molecules again without a political deal.

Anyone printing a single 2027 average today is choosing a peace date. JPMorgan’s point is that the peace date is not a market input anymore.

The HFI post, and what to do with it

The HFI Research post that kicked this around the energy timeline is useful because it published the actual first page, not a paraphrase. The authors on that page are Kaneva, Savinova, and Fakhredtinov. The date is September 17, 2026. The title is “Dry powder.” If you only read “JPMorgan panics” quote-tweets, you missed the inventory math that still argues against an immediate squeeze.

The replies under that post are the market’s mood, not the research. One user noted oil was still down several dollars over two days despite the note. Another remembered JPMorgan talking $30 oil before last Thanksgiving. That is the business. Banks will be early, late, and occasionally honest. Thursday was the honest day.

Bottom line

The oil market can stabilize without a treaty if three things hold: Gulf bypass flows stop getting worse, non-Gulf producers keep adding, and demand stays the residual. In that world, $90 is closer to fair value than $106, winter diesel is the spike risk, and 2027 lives in a $64–$87 band depending on whether the guns stop.

The LNG market cannot use that script. It needs either Hormuz or another year of American liquefaction. Until one of those arrives, gas stays the tighter commodity and the one that can still print crisis prices in a cold quarter.JPMorgan did not say oil is going to $40 or $140. It said the input that used to pick between those numbers is no longer modelable. That is a more serious sentence than a revised target. The rest of the Street will keep publishing targets. Treat them as scenarios with dates attached. The date is the forecast.

Making Appendices Great Again

Check out the World’s Greatest Podcast Show Notes at EnergyNewsBeat.co or EnergyNewsBeat.com.

The post JP Morgan Is at a Loss to See Where Oil Goes From Here appeared first on Energy News Beat.