NG Breaks $2.90 With Production at 113 Bcf/d and LNG Feedgas Capped at 18.3 Bcf/d

Natural gas fell to $2.81 per MMBtu Thursday, down 3.73% from the previous session — the sharpest single-day decline in weeks and a break below the $2.90 shelf that had held all week. Over the past month the contract is up just 0.47%, and it sits 7.32% below where it traded a year ago.

On the same day, European natural gas rose above €80 per MWh, its highest level since December 2022, extending a run that has now stretched across five sessions.

Two benchmarks for the same molecule, moving in opposite directions by nearly four percentage points in a single session. That divergence is the entire story of this market and it explains why every American bull case has failed at the same price for six weeks.

The recent Henry Hub record shows the pattern clearly. Friday, September 4 brought a 2% gain to $2.97. Monday's October contract added 1.63% on triple-digit temperature forecasts. Tuesday, the contract briefly topped $3.00 before reversing hard, settling near $2.855 after opening at $2.904 with an intraday range of $2.845 to $2.904 — the third time in four trading sessions that a move above $3.00 proved short-lived. Wednesday it held above $2.90, not far from two-month highs. Thursday it broke.

The macro backdrop was hostile across commodities. West Texas Intermediate touched $100.10 a barrel, up 4.2%. Brent reached $105.37. The 10-year Treasury yield climbed to 4.90%. August producer prices came in at 5.4% annually. Odds of a Federal Reserve hike at the September 15–16 meeting sit between 62% and 64%.

Crude ripping 3.44% while domestic gas falls 3.73% on the same day is unusual, and it isolates what is driving each. Oil is trading a geopolitical supply shock. Henry Hub is trading a domestic storage surplus, and no amount of Middle East escalation changes what is sitting in American salt caverns.

The October contract carries roughly 304,680 contracts of open interest at 10,000 MMBtu each, with next settlement on September 28.

The Fourth Failed Attempt at $3.00 in Six Sessions

The $3.00 handle has become the defining feature of this market, and the repeated rejections tell you more than any single print.

October futures have now pushed above $3.00 and failed four separate times inside six trading sessions. Tuesday's reversal was the most instructive: the contract opened at $2.904, cleared the round number intraday, and settled at $2.855 — a full session round trip that ended below the open despite the market spending time above resistance.

Each attempt has been driven by the same input and defeated by the same one. The bull catalyst has been weather. Forecasts for above-average temperatures across the U.S. South through mid-September, near-record highs in the southern and eastern United States, triple-digit readings in Texas, and peak power demand projections from the largest Texas grid operator that exceeded the all-time record set in July.

The bear catalyst has been supply, and it has won every time. Lower-48 dry gas production has been running near record levels — 113.1 Bcf per day at one measurement, up 4.4% year over year, with a near-record 115.0 Bcf per day recorded on a single Sunday. Production above 112 Bcf/d absorbs any weather-driven demand spike within days.

The pattern has a name in the physical market: a weather trade that runs straight into the storage wall. Hot forecasts trigger buying. They do not settle the market unless the storage numbers start tightening, and they have not tightened.

Spot physical prices have behaved differently from futures. Cash strengthened across much of the country during the recent heat, with gains spreading from the Permian Basin and Rockies into Appalachia and the Northeast, while elevated Gulf Coast and Southeast prices reversed. Regional price spikes have been driven by pipeline operational constraints rather than by a national supply shortage.

Thursday's 3.73% break below $2.90 confirms what four failed attempts implied. The seasonal window for a weather-driven rally is closing, and the market knows it.

Production at 113 Bcf/d and a Record 3,985 Bcf Storage Target

The supply picture is the reason $3.00 keeps failing, and the numbers are unambiguous.

Lower-48 dry gas production has been running at or near record levels throughout the summer, with readings between 112.7 and 115.0 Bcf per day and year-over-year growth in the 3% to 4.4% range. That production has continued through a period when the front-month contract has spent most of its time below $3.00 — meaning producers are not responding to price, because associated gas from oil drilling and Appalachian economics both work well below current levels.

The EIA projects natural gas inventories will reach 3,985 billion cubic feet at the end of October 2026. That would be the highest end-of-injection-season level since 2016 — a ten-year record.

Sit with that figure. Storage entering the withdrawal season at a decade high is a structural cap on winter prices, because it means the market can absorb an unusually cold December without the scarcity pricing that produces spikes. The agency framed its own downgrade in exactly those terms, citing reduced LNG feedgas demand and record natural gas production as the reasons inventories will be at their highest level heading into winter since 2016.

Demand has not kept pace. Lower-48 gas demand has been running around 80.9 to 82.2 Bcf per day, with year-over-year readings ranging from up 2.7% to down 3.1% depending on the week. Production growing 4% against demand growing 2% or falling is the surplus arithmetic in its simplest form.

The one genuinely bullish demand data point is electricity. U.S. electricity output rose 6.1% year over year in one August week to 100,895 gigawatt hours, with output across the trailing 52 weeks up 2.2% to 4,365,212 GWh. Data center load is real and growing.

It is not yet growing fast enough to offset 113 Bcf/d of production. The political dimension is also shifting — a trade group representing the Marcellus Shale industry recently voiced disappointment over a Pennsylvania political pivot on data centers, which is the first sign that the load-growth story faces siting resistance.

Storage at 5.2% Above the Five-Year Average

The weekly storage record through the summer has been relentlessly in line or bearish, and that consistency is what has broken the bull case.

U.S. gas inventories stood 5.2% above the five-year seasonal average as of August 28. Operators injected 30 Bcf into storage during that week — a result that met expectations exactly and, at least initially, deflated futures.

The prior weeks tell the same story. One injection came in at 33 Bcf against a 30 Bcf estimate and a five-year average of 23 Bcf, with warm weather already in place. That is the damning combination: a build 43% above the seasonal norm during a heat wave. Another week produced 59 Bcf against consensus forecasts clustered at 51 to 54 Bcf. Earlier in the season, storage sat 6.7% above the five-year average.

The surplus has narrowed slightly, from 6.7% to 5.2%, which is the closest thing to a bullish trend the data has produced. It has narrowed because of exceptional cooling demand, and cooling demand is about to disappear.

The weekly storage report is due Thursday, and it carries more weight than usual for two reasons. It is the last report covering peak cooling demand, and it is the first read on whether LNG feedgas at 18.3 Bcf/d is pulling enough gas out of the South Central region to offset the production overhang.

Historical context is worth holding. Working gas stocks at 3,065 Bcf during the last withdrawal season stood 177 Bcf — 6% — above the five-year average and 141 Bcf above the prior year. The market has been carrying a surplus for most of a year.

A build meaningfully below 30 Bcf would be the first genuine tightening signal of the season. A build above it confirms the path to 3,985 Bcf and caps the contract below $3.00 into the seasonal turn.

LNG Feedgas at 18.3 Bcf/d Is the Arbitrage That Cannot Widen

Here is the mechanism that keeps the American and European markets apart, and it is physical rather than financial.

Average feedgas flows to the nine major U.S. LNG export facilities climbed to 18.3 Bcf per day in early September, up from 17.2 Bcf per day in August, as Texas facilities returned to full operations after maintenance. Freeport LNG completing its maintenance programme restored roughly 2 Bcf per day of feedgas demand that had been offline. One September reading put flows at 18.1 Bcf/d, and a separate measurement showed feedgas slipping 1.4% week over week to 18.3 Bcf/d during Freeport downtime.

Demand for U.S. LNG from Europe and Asia has risen sharply as buyers seek to replace disrupted Middle Eastern supplies and replenish inventories ahead of winter. That demand is genuine, urgent and price-insensitive.

And it cannot be met, because liquefaction capacity is fixed in the short run. Whatever Europe is willing to pay, the United States can export roughly 18.3 Bcf per day and no more until new trains come online. The arbitrage between $2.81 Henry Hub and €80/MWh TTF — a spread of well over $10 per MMBtu — is uncapturable in size.

That is why the two benchmarks have decoupled so violently. In a market with unlimited export capacity, Henry Hub would be pulled toward European pricing. In a market with fixed capacity, the export terminals run flat out, capture the entire spread as margin, and the domestic benchmark clears against domestic supply and demand alone.

The corporate winner is the liquefaction operator. Cheniere and the terminal owners earn the spread. The producer earns $2.81.

Feedgas rising from 17.2 to 18.3 Bcf/d is a 6.4% increase in export demand — meaningful, and roughly one-tenth of the daily production surplus that needs absorbing.

Europe Above €80/MWh With 20% of Global Gas Flows Shut

The European market is pricing a genuine emergency, and the mechanics deserve laying out because they explain the entire global gas complex.

European natural gas rose above €80 per MWh Thursday, the highest since December 2022 and a fifth consecutive session of gains. The driver is the Strait of Hormuz, which handles roughly 20% of global gas flows, principally Qatari LNG.

Iran said it had struck more than a dozen vessels attempting to pass through the strait and warned it would escalate if Washington continued strikes on its territory. The U.S. military destroyed five Iranian oil tankers near Kharg Island in response to attacks on its warships, and Tehran retaliated against U.S. targets in Jordan. Iran-backed Houthi militants launched drones and missiles at energy facilities in southern Saudi Arabia.

The sustained disruption has forced Qatar to suspend shipments. Qatar is the world's largest LNG exporter by several measures and its output is effectively stranded behind a contested waterway.

Two additional supply constraints are compounding it. Maintenance in Norway is limiting pipeline flows into northwest Europe, and lower Algerian flows to Italy are reducing southern supply.

The timing is the worst possible. Europe is nearing the end of its summer storage injection season with inventories below historical norms — the period when the continent must fill tanks or face winter short. Every week that Hormuz stays closed reduces the volume available for injection and intensifies competition for the cargoes that do move.

For the American market, the read-through is asymmetric and small. European scarcity supports U.S. LNG export economics and terminal utilization, both of which are already maxed. It does not create incremental American demand, because the pipes to the terminals are full.

The European contract at €80/MWh converts to roughly $27 per MMBtu at prevailing exchange rates. Henry Hub is $2.81.

The EIA's Own Forecast: $2.87 for Q3, $3.14 for Q4, $3.62 for Q1

The government's baseline is the cleanest reference available and it has been revised down twice.

The most recent Short-Term Energy Outlook puts the Henry Hub spot price averaging $3.44 per MMBtu in 2026 and $3.31 in 2027. The prior outlook had projected $3.67 this year and $3.49 next. Before that, the agency had forecast a 2% decline in 2026 followed by a 33% increase in 2027 to just under $4.60.

The quarterly path matters more than the annual average. The agency sees Henry Hub at $2.87 in the third quarter of 2026, $3.14 in the fourth, $3.62 in the first quarter of 2027 and $2.79 in the second. The third-quarter figure was cut by 50 cents from the prior forecast, with the explanation citing reduced LNG feedgas demand and record production leaving inventories at their highest level heading into winter since 2016.

Henry Hub averaged $3.53 in 2025. The August 2026 spot price came in at $2.78.

Read against the tape, the current contract at $2.81 sits almost exactly on the official third-quarter estimate of $2.87 and 10.5% below the fourth-quarter projection of $3.14. That gap is the seasonal recovery the forecast expects and the market has not yet delivered.

The first-quarter 2027 estimate of $3.62 is where the structural story reasserts itself. The agency has been consistent that 2027 demand growth will outpace supply growth, driven mainly by feed gas demand from U.S. liquefaction facilities, drawing down storage and lifting prices. As inventories move toward or below the five-year average, the forecast price rises.

That framing identifies the single variable to track: storage relative to the five-year average. At 5.2% above it, prices stay near $3.00. Below it, the curve reprices.

Nothing in the current data suggests that crossover happens before 2027.

The Calendar Problem: Cooling Fades Before Heating Arrives

The seasonal mechanics are working against buyers and the window is narrow.

Cooling demand starts fading in September and winter heating demand does not pick up until late October or November. That gap — roughly six to eight weeks — is the period when weather provides no support in either direction and the market trades pure supply and storage.

The market is entering that gap now. Forecasts for above-average temperatures across the U.S. South through mid-September provide a final stretch of air-conditioning load, and after that the bull case has nothing until the first genuine cold front.

The historical seasonal pattern reinforces it. Storage builds run from spring through October, with injections typically continuing at full pace through September. A market with production at 113 Bcf/d, demand falling to seasonal lows, and eight weeks of uninterrupted injections ahead is a market where the surplus grows rather than shrinks.

That is precisely how inventories reach 3,985 Bcf by the end of October.

The counterweight is real but slow. LNG feedgas at 18.3 Bcf/d provides a demand floor that did not exist in prior shoulder seasons, and European scarcity guarantees export terminals run at maximum utilization. Data center load growth adds baseload power demand that is weather-independent. Neither offsets 113 Bcf/d of production in a shoulder season.

Fourth-quarter positioning is where the risk turns. Once storage peaks and the withdrawal season begins with a decade-high inventory, the market needs a genuinely cold winter to draw stocks toward the five-year average. Absent that, the surplus carries into spring 2027 and the $2.79 second-quarter forecast becomes the destination rather than the floor.

A cold December against 3,985 Bcf of storage is a $3.50 to $4.00 market. A mild one is a $2.50 market.

That binary is what October and November trade.

Technicals: $2.81, the $3.00 Wall, and a Base at $2.60

The chart has a well-defined structure and Thursday's break changes the near-term read.

The immediate reference is $2.81, Thursday's print after a 3.73% decline. Below it, $2.78 marks the August spot average and the first shelf. Beneath that, $2.60 is the base of the pattern that formed in late August, and it anchors the bull-flag structure that traders identified running from a $3.20 flagpole high down to $2.60.

Above, the levels stack tightly. $2.855 was Tuesday's settlement. $2.904 was Tuesday's open and the session high — a level that has now capped price on multiple attempts. $2.97 was Friday's close. And $3.00 is the round number that has rejected four separate attempts in six sessions.

Beyond $3.00, the structure opens considerably. $3.20 is the flagpole high and the first meaningful target on a genuine breakout. The 2026 high sits far above at $4.55, a two-year peak tested earlier in the year when a winter storage scare briefly repriced the entire curve.

Daily technical composites read neutral, which fits a market that has spent six weeks oscillating in a 40-cent band.

Thursday's break below $2.90 on a 3.73% move is the first decisive directional session in that stretch, and it happened on a day when crude gained 3.44% — meaning it was not a broad energy liquidation but a gas-specific repricing.

The trading setup that follows is mechanical. Rallies toward $2.90 and $2.904 are sellable while storage runs above the five-year average. A close below $2.78 targets $2.60. Only a settlement above $3.00 held for consecutive sessions changes the structure, and that requires either a storage print well below 30 Bcf or an unexpected weather event.

Each contract represents 10,000 MMBtu with a tick size of 0.001 worth $10, and open interest of roughly 304,680 contracts against daily volume near 66,000 indicates positioning that is held rather than churned.

Today's Storage Report and the Number That Matters

The weekly EIA storage release is the highest-frequency signal in this market and Thursday's carries specific weight.

Recent prints have clustered around 30 Bcf. The week ended August 28 delivered exactly 30 Bcf, meeting consensus and deflating futures on release. The week before came in at 33 Bcf against a 30 Bcf estimate and a five-year average of 23 Bcf. An earlier April week produced 59 Bcf against forecasts of 51 to 54 Bcf.

The pattern across those three prints is consistent: builds have met or exceeded expectations, and they have exceeded the five-year average even during heat waves.

Working gas inventories now sit 5.2% above the five-year seasonal average. The path to the projected 3,985 Bcf end-of-October figure requires roughly 30 Bcf per week to continue through the remainder of the injection season, which is exactly what the market has been delivering.

Three outcomes and what each does to price. A build below 20 Bcf would be the first genuine surprise of the season, would narrow the surplus toward 4%, and would put $3.00 back in play immediately. A build of 25 to 35 Bcf confirms the trajectory, keeps the surplus near 5%, and leaves the contract trading $2.78 to $2.90. A build above 40 Bcf accelerates the path to a decade-high inventory and targets $2.60.

The variable that could produce the bullish surprise is LNG. Feedgas at 18.3 Bcf/d against 17.2 Bcf/d in August represents an incremental 1.1 Bcf/d of demand — roughly 7.7 Bcf per week — pulling primarily from the South Central storage region where the export terminals sit. If that volume shows up in the regional breakdown, the tightening thesis gains its first real evidence.

Freeport LNG's return from maintenance is the specific driver, having restored roughly 2 Bcf/d of feedgas that had been absent from prior reports.

Regional detail will matter more than the headline this week.

The Equity Read-Through: Producers, Midstream and the Export Terminals

The equity complex splits cleanly along the same fault line as the two benchmarks, and it is a better expression of the thesis than the futures contract.

Pure-play producers — EQT in Appalachia, Coterra across the Marcellus and Permian, Chesapeake in the shale basins — earn the domestic benchmark. At $2.81 Henry Hub with production at 113 Bcf/d and storage headed to a decade high, their realized pricing is capped regardless of what Europe pays. Appalachian producers additionally face basis differentials that put their netbacks below the headline figure.

Cheniere and the liquefaction operators earn the spread. With Henry Hub at $2.81 and European gas above €80/MWh, the tolling economics on every cargo leaving a Gulf Coast terminal are extraordinary, and terminal utilization is maxed. Feedgas at 18.3 Bcf/d across nine facilities is close to nameplate capacity for the existing fleet.

That distinction is the trade. The commodity is cheap because America has too much of it. The export infrastructure is valuable because the world does not have enough and the pipes are full.

The broader energy tape was mixed Thursday. Liberty Energy fell 5.58% to $20.83 and Transocean traded $5.71, down 0.87%, on a day WTI gained 3.44% — services names get no benefit from a geopolitical price spike because it does not produce a drilling cycle. Energy has still led all eleven S&P 500 sectors in 2026 with a gain of roughly 42%, on multiple expansion rather than earnings, which are down 10.9% year over year at the sector level.

The M&A signal is small but present. A small-scale LNG asset acquisition in the Desert Southwest closed this week, in a region where scorching weather has caused gas demand and prices to spike periodically through summer.

Data center demand remains the structural bull case for producers, and the Pennsylvania political pivot that drew industry criticism is the first evidence that siting, not gas supply, may be the binding constraint.

Verdict and Forecast: Capped at $3.00 Until Storage Turns

Natural gas at $2.81, down 3.73% on a day crude rose 3.44% and European gas hit a four-year high, is a market that has decoupled from the global energy story and is trading its own surplus.

The bear evidence is overwhelming and it is domestic. Lower-48 production running 112.7 to 115.0 Bcf per day, near record and up 3% to 4.4% year over year. Storage 5.2% above the five-year average with weekly injections of 30 to 33 Bcf that have met or exceeded consensus and beaten the 23 Bcf seasonal norm. An official projection of 3,985 Bcf at the end of October, the highest in a decade. Four failed attempts at $3.00 in six sessions. And a seasonal calendar where cooling demand fades before heating arrives.

The bull evidence is real and it is offshore. European gas above €80/MWh, the highest since December 2022. Qatar suspending shipments with Hormuz — 20% of global gas flows — contested. Norwegian maintenance and reduced Algerian flows compounding it. Europe entering winter with inventories below historical norms. LNG feedgas rising from 17.2 to 18.3 Bcf/d as Texas terminals return from maintenance.

The reason the second list has not lifted the first is capacity. American liquefaction runs at roughly 18.3 Bcf/d and cannot flex higher this year. The spread between $2.81 and €80/MWh is captured by terminal owners, not by the domestic benchmark.

The forecast follows. Into the fourth quarter, Henry Hub trades $2.60 to $3.05, with $3.00 the ceiling that requires a storage print below 20 Bcf or a genuine cold event to clear. Thursday's break below $2.90 targets $2.78 first and $2.60 on continuation. The official $3.14 fourth-quarter average implies a recovery the current data does not support, and the risk to that number is downward.

The turn comes in 2027, not 2026. The structural case — demand growth outpacing supply growth on new liquefaction capacity, storage drawing back toward the five-year average, and a $3.62 first-quarter print — depends on export capacity expanding beyond 18.3 Bcf/d. That is a 2027 event.

Selling rallies toward $2.90 to $3.00 remains the trade while storage runs above the five-year average. Buying below $2.60 with a winter horizon is where the asymmetry sits, because 3,985 Bcf of storage against a cold December is the only path back to $4.00.

That's TradingNEWS