NG Breaks To $2.858 On Pipeline Startup As Storage Runs 167 Bcf Above The 5-Year Average

Front-month October natural gas (NGV26) fell 2.5% to a session low of $2.858 per MMBtu — the weakest print since the recent multi-week recovery attempt began — as Energy Transfer's Hugh Brinson Pipeline commenced service today.

The line moves approximately 2.2 billion cubic feet per day of gas from the Permian Basin to East Texas, giving Permian producers another route toward Erath, Louisiana, the delivery point where benchmark US gas futures settle. The market has known this date for weeks. What changed today is that a forward risk became an immediate physical reality.

The timing could not be worse for bulls. Lower 48 dry gas production is running at 113.0 Bcf/d, up 4.5% from a year ago. August output across the Lower 48 averaged approximately 111.5 Bcf/d, surpassing July's record of 110.7 Bcf/d. The Baker Hughes gas rig count rose by five last week to 132, a five-month high and just below February's three-year peak of 134.

Producers are adding rigs with Henry Hub beneath $3.00. That is not the behavior of a market clearing.

Storage confirms it. Working gas stood at 3,184 Bcf as of Friday, August 21 — 167 Bcf above the five-year average of 3,017 Bcf and roughly 6% above normal. Official projections put end-October inventories at a record 3,985 Bcf, 5% above the five-year average and the highest level heading into winter since 2016.

The official third-quarter Henry Hub forecast sits at $2.87 per MMBtu, cut 50 cents from the prior month's estimate on reduced LNG feedgas demand and record production. Futures contracts through September have been trading below $3.00.

The thesis for this forecast: the supply ceiling is rising faster than demand can absorb it, and 2.2 Bcf/d of new takeaway landed into an already saturated system. September heat is the only thing standing between this market and a faster slide into shoulder season.

There is one data series running the other way and almost nobody is trading it. Weekly injections have collapsed from 36 Bcf to 15 Bcf across three reports, and the five-year surplus has narrowed from 198 Bcf to 167 Bcf over the same stretch.

That deceleration is the entire bull case. It is real, and it is being ignored.

The 2.2 Bcf/d That Landed On An Already Saturated Market

The pipeline event deserves precise framing because its impact is structural rather than sentimental.

Hugh Brinson connects Permian Basin production to East Texas, ultimately routing supply toward Erath, Louisiana near the Henry Hub. Capacity is approximately 2.2 Bcf/d. Service began September 1.

Permian gas is associated gas — it comes out of the ground alongside crude oil, and its production is governed by oil economics rather than gas prices. With Brent at $92.04 and WTI at $87.96, crude is up 50% year to date and Permian drilling activity has every incentive to expand. That means the gas volumes flowing through Hugh Brinson are price-insensitive on the way down. Producers will keep shipping at $2.50 because the barrel pays for the well.

Before this line existed, a meaningful share of that associated gas was stranded behind takeaway constraints, and Waha basis frequently traded at steep discounts to Henry Hub. Removing the bottleneck converts stranded regional supply into Gulf Coast benchmark supply — which is precisely why the Henry Hub contract took the hit rather than a regional basis point.

The scale is worth sizing. Against Lower 48 dry production of 113.0 Bcf/d, 2.2 Bcf/d represents 1.9% of national supply. Against the roughly 15 to 16 Bcf/d of LNG feedgas demand, it is 14%. Against the injection rate that has been running 15 to 36 Bcf per week — roughly 2.1 to 5.1 Bcf/d — it is between 43% and more than 100% of the entire weekly storage build.

Put differently: this single pipeline can add more gas per day than the market has been injecting into storage in recent weeks.

That is why the commissioning triggered fresh selling rather than a shrug. The market had priced the announcement. It had not priced the molecules.

The offsetting consideration is that pipelines ramp rather than switch on. Initial flows are typically well below nameplate, and full utilization can take quarters. The 2.2 Bcf/d figure is capacity, not current throughput.

But directionally the market now knows where the ceiling on price sits, and it sits lower than it did last week.

Storage At 3,184 Bcf And 167 Bcf Above The Five-Year Average

The inventory picture is comfortable in absolute terms and tightening at the margin, and both facts matter.

Working gas in underground storage stood at 3,184 Bcf as of Friday, August 21, according to the latest weekly report. That was a net increase of 15 Bcf from the prior week. Stocks sat 30 Bcf below year-ago levels and 167 Bcf above the five-year average of 3,017 Bcf — roughly 6% above normal. Total working gas remains within the five-year historical range. All regions built except Pacific and South Central Salt.

The regional detail from the prior week is instructive. The East held 708 Bcf, the Midwest 848 Bcf, Mountain 237 Bcf, Pacific 296 Bcf and South Central 1,080 Bcf, with salt cavern storage down 18 Bcf and nonsalt up 5 Bcf.

South Central salt drawing down while the national total builds is the tell for late-summer power burn. Salt caverns cycle fastest and serve Gulf Coast gas-fired generation. Drawing salt in August means the region is running its generators hard.

The trajectory of injections is the most underappreciated data in this market. The sequence reads +32 Bcf for the week ending July 17, +28 for July 24, +33 for July 31, +36 for August 7, +16 for August 14 and +15 for August 21.

Injections more than halved across two weeks and have stayed halved.

The surplus narrative tracks it. The five-year surplus ran 183 Bcf, then 185, then 195, then 198 — and has since compressed to 185 and then 167. Thirty-one Bcf of surplus erosion in a fortnight.

Supply-side data explains part of it. Total US natural gas supply fell 0.9 Bcf/d, or 1%, over the reporting week. Net Canadian imports decreased 0.4 Bcf/d, or 8%, and dry gas production decreased 0.4 Bcf/d.

That is the bull case in full: injections decelerating, surplus narrowing, salt drawing, Canadian imports falling.

It is also fragile. A single 40 Bcf build reverses the narrative, and the next report lands Thursday, September 3.

The 3,985 Bcf Problem Heading Into Winter

The forward inventory projection is the number that caps every rally in this market.

Official forecasts put natural gas inventories at a record 3,985 Bcf at the end of October 2026 — an increase of 19 Bcf versus the prior month's estimate and 5% above the five-year average. That would be the highest pre-winter storage level since 2016.

Storage capacity in the Lower 48 is finite. Approaching 4,000 Bcf in late October means the system is close to operational limits in some regions, which forces gas to clear through price rather than through injection. That is the mechanism that produces autumn price collapses.

The price forecast follows directly. Henry Hub is expected to remain below $3.00/MMBtu until November and to average $3.03/MMBtu over the remaining five months of the year — nearly 50 cents lower than the prior month's projection. The third-quarter estimate sits at $2.87/MMBtu, cut 50 cents on reduced LNG feedgas demand and record production.

Trading at $2.858 puts the market slightly beneath that quarterly estimate with one month of the quarter remaining. And the forecast was completed before Hugh Brinson turned on.

The next official update lands September 9. It is the first revision that will incorporate the pipeline commissioning, the 113.0 Bcf/d production rate and the rig count at 132. A further cut to the third and fourth-quarter Henry Hub path from that release would validate the current selling.

The longer-arc setup is materially more constructive and worth holding in view. Supply growth outpaces demand growth by 0.5 Bcf/d in 2026, then falls behind by 1.6 Bcf/d in 2027 as demand growth of 2.5 Bcf/d exceeds supply growth of 0.9 Bcf/d. LNG exports grow 9%, or 1.3 Bcf/d, in 2026 and 11%, or 1.7 Bcf/d, in 2027 as Plaquemines LNG, Corpus Christi Stage 3 and Golden Pass LNG ramp.

Storage inventories are expected to move gradually below the rolling five-year average across that forecast period, and the Henry Hub annual average is projected to rise from just under $3.50 in 2026 to just under $4.60 in 2027.

The bull case is a 2027 story. The bear case is happening now.

Production At 113 Bcf/d And A Rig Count That Keeps Climbing

The supply side is the cleanest bearish argument on the board and it has no near-term resolution.

Lower 48 dry gas production is running at 113.0 Bcf/d, up 4.5% from a year ago. August output averaged approximately 111.5 Bcf/d across the month, surpassing July's record of 110.7 Bcf/d. Both figures are all-time highs.

The rig count says more is coming. Baker Hughes reported the US gas rig count rose by five last week to 132 — a five-month high, and only two rigs below February's three-year peak of 134.

Producers are adding rigs with the front-month contract below $3.00. That is the signature of a market where marginal supply economics are not being set by the gas price.

Two things explain it. First, associated gas from the Permian is a byproduct of oil drilling, and with crude up 50% year to date the oil economics fund the well regardless of what gas does. Hugh Brinson gives that gas a route to market it did not have, which increases the volume reaching Henry Hub without requiring a single new gas-directed rig.

Second, dry gas producers in Appalachia and Haynesville are drilling against 2027 demand rather than 2026 prices. With LNG export capacity ramping across three facilities and the supply-demand balance flipping to a 1.6 Bcf/d deficit in 2027, forward strip economics justify activity that spot prices do not.

Pipeline exports add to the demand side but not enough to offset. Total US pipeline exports are estimated at 9.6 Bcf/d in 2026, rising to 10.0 Bcf/d in 2027 from 9.5 Bcf/d in 2025, driven by the Energia Costa Azul LNG terminal on Mexico's Pacific Coast, which shipped its first cargo on July 8 and brings 0.4 Bcf/d of nominal capacity online. Mexican gas-fired generation additions are absorbing more US molecules.

Against 113.0 Bcf/d of production, an incremental 0.5 Bcf/d of Mexican demand is a rounding error.

The structural resolution is LNG. US LNG exports are projected to average 16.5 Bcf/d in the third quarter, and every incremental Bcf/d of feedgas is a direct offset to record production. That ramp is the only demand source scaled to match the supply growth.

It arrives in 2027, not September.

The LNG Arbitrage: $23 In Asia Against $2.86 At Henry Hub

The international price picture is the most striking number in this market and the strongest argument for the structural bull case.

The Title Transfer Facility price in Europe averaged $22.70/MMBtu for the week ending August 26, up $1.59 from the prior week. The Japan-Korea Marker averaged $23.07/MMBtu, an increase of $1.47. Both weekly averages were the highest since the weeks ending December 29, 2022 and January 18, 2023 respectively, when they reached $25.48 and $24.85.

Henry Hub trades $2.858.

That is a spread of roughly $20 per MMBtu between the US benchmark and both international markers — an arbitrage of approximately eight times. Even accounting for liquefaction costs, shipping and regasification, the economics of moving American molecules to Europe and Asia are extraordinary.

The cause is the same conflict driving crude to $92.04 Brent. European gas prices spiked with the war in Iran, and the eurozone inflation impulse from energy has already pushed August flash HICP to 3.3% year over year. Asia is competing for the same cargoes.

For US producers and exporters, that spread is the entire investment thesis. Every liquefaction train that comes online captures it. Plaquemines LNG and Corpus Christi Stage 3 are ramping toward full operations, and Golden Pass LNG is expected to begin operations in 2026.

The constraint is liquefaction capacity, not price signal. The United States cannot export the gas fast enough to close the arbitrage, which is why Henry Hub can sit at $2.86 while Asia pays $23.07.

Freeport LNG in Texas completed maintenance and returned online, pushing feedgas flows to their highest level since late June. That is a genuine demand addition and it happened during the same week the price fell — which tells you how much supply is overwhelming it.

The forward implication is that this arbitrage does not persist. Either international prices fall as the conflict resolves, or US export capacity catches up and drags Henry Hub higher. The 2027 forecast of just under $4.60 assumes the second.

Between now and November, the spread is a fact about the world rather than a tradeable signal for the front month.

September Heat Is The Only Thing Holding This Market Up

Weather is the sole bullish input in the near term, and it has a defined expiry.

Above-average temperatures are expected across the eastern two-thirds of the United States from September 2 through September 11, with forecasts calling for 90s to 110s across the southern two-thirds through the first week of the month. That keeps air-conditioning demand elevated and supports gas consumption from power generators.

Warmer-than-usual conditions could persist through mid-September, although one forecast model showed a reduction in cooling-degree days, pointing to moderation in demand.

The power burn channel is what matters. Gas-fired generation is the swing consumer in September, and with the South Central salt caverns drawing down while national storage builds, the data already shows generators running hard.

The problem is arithmetic. If forecasts for 90s to 110s across the southern two-thirds hold through the first week of September, power burns can keep gas near current levels. The moment that heat breaks, the floor gets substantially harder to defend.

Shoulder season is the structural issue. Between the end of cooling demand and the start of heating demand, gas has no natural consumer other than LNG feedgas and industrial load. That window typically runs mid-September through late October, and it arrives this year with production at 113.0 Bcf/d, a new 2.2 Bcf/d pipeline running, and storage on track for a record 3,985 Bcf.

Residential and commercial consumption is forecast to decrease 4% in 2026 to 22.1 Bcf/d on closer-to-normal temperatures compared with 2025's colder-than-normal winter months. Industrial consumption is also forecast lower on decreased activity as measured by the gas-weighted manufacturing index.

Both demand categories shrinking into a record supply year is the definition of a loose market.

The trade implication is a calendar rather than a level. Heat carries the price through roughly September 11. Between September 11 and the first cold snap, there is a window where supply has no offset. That is where the risk of a fast move below $2.75 lives.

Watch the six-to-ten-day forecast more closely than the storage number.

Technical Structure: The 50-Day Rejected The Spike

The chart is unambiguously bearish on the longer timeframes and mixed on the short ones.

The 50-day moving average rejected last Thursday's spike and sellers came in immediately behind it. That rejection is the most important recent technical event, because it marked the failure of the multi-week recovery attempt that took the market to a five-week high above $2.92 on Monday.

Today's session low at $2.858 is the weakest print since that recovery began. Losing it decisively opens the $2.80 handle, and beneath that the $2.75 area becomes the reference. Below $2.75, the market has limited tested support until the summer lows.

On the upside, the immediate references are stacked and close. The Monday high above $2.92 is the first meaningful resistance, followed by the $3.00 psychological handle that futures contracts through September have not sustained. The 50-day moving average sits above as the level that already rejected one attempt.

The multi-timeframe signal picture reflects the split. Hourly readings show Strong Sell and weekly and monthly readings show Strong Sell, while the daily reading has been Strong Buy — a configuration where a short-term oversold bounce sits inside a bearish structure.

That combination argues for selling rallies rather than shorting breakdowns. The daily oversold reading means a bounce toward $2.92 is plausible. The weekly and monthly readings say that bounce is a selling opportunity rather than a reversal.

Volume and open interest confirm the positioning shift. The October contract rose as much as $0.05/MMBtu during the reporting week while most 2027 contracts edged lower — a curve steepening in the front and flattening in the back that reflects near-term heat premium against long-term supply expectation.

For the exchange-traded vehicles, the structural point remains that contango in the forward curve erodes long-side returns over time. UNG carries roll costs in a contango market, and the leveraged products compound that decay. BOIL is a trading instrument for days, not weeks; KOLD works in the opposite direction and benefits from the same structure.

The equity complex trades the 2027 story rather than the September price. EQT, Chesapeake, Coterra and Cheniere all price off forward strip and LNG capacity, not off $2.858 spot.

Downside Map: $2.80, $2.75 And The Shoulder Season Gap

The bear case is well supported and the path is short.

The first reference is today's session low at $2.858. Closing beneath it confirms the recovery attempt has failed and puts the $2.80 handle in play immediately.

Below $2.80, the market has minimal tested support until $2.75 and then the summer lows. The reason is the same reason that produced the rally: this market spent weeks compressed in a narrow band and then spiked to $2.92 before reversing. Spikes do not build volume shelves.

The catalyst sequence is dense and lands over ten days. Thursday, September 3 brings the weekly storage report covering the week ending August 28. Injections have decelerated to 15 and 16 Bcf across the last two prints. A build back above 30 Bcf would break the tightening narrative that is currently the only thing arguing against a slide and would confirm Hugh Brinson volumes are already reaching storage.

September 9 brings the next official energy outlook, which will be the first to incorporate the pipeline commissioning, the 113.0 Bcf/d production rate and the 132 rig count. A further reduction to the third and fourth-quarter Henry Hub path would validate the selling.

September 11 marks the end of the currently forecast above-average temperature window. That is the calendar date where power burn support expires.

The structural pressure underneath all three: record production, a new 2.2 Bcf/d takeaway route into the benchmark, a rig count at a five-month high, and end-October storage projected at a record 3,985 Bcf.

The mitigating factors are genuine but thin. Injections have more than halved. The five-year surplus narrowed from 198 Bcf to 167 Bcf in a fortnight. South Central salt is drawing. Canadian imports fell 8%. Total supply dropped 0.9 Bcf/d on the week. Freeport returned from maintenance and lifted feedgas to its highest since late June.

If those trends hold through two more reports while heat persists, the market can defend $2.80 and grind sideways into October.

If a single large build prints on Thursday, the $2.75 test happens within the week.

Upside Map: $2.92, $3.00 And What Would Actually Change This

The bull case requires specific conditions and none of them are currently present.

The first level is Monday's high above $2.92, the five-week peak. Reclaiming it would repair the immediate damage from today's 2.5% decline and put the 50-day moving average back in play.

Above that, $3.00 is the psychological and structural line. Futures contracts through September have been trading below it and the official forecast expects Henry Hub to remain beneath $3.00 until November. Sustaining a close above $3.00 would contradict the base case that has been guiding the market since the summer.

What would produce it is a specific combination. Injections continuing to decelerate below 15 Bcf per week, which would compress the 167 Bcf surplus toward parity faster than the market expects. LNG feedgas rising through 16.5 Bcf/d as Freeport runs at full rate and the newer trains ramp. Weather models extending the above-average temperature forecast past mid-September. And a production stumble — freeze-off risk does not apply in September, but maintenance and interruptions do.

The single largest potential catalyst is the surplus math. If weekly injections stay near 15 Bcf while the five-year average injection for those weeks runs closer to 29 Bcf, the surplus erodes at roughly 14 Bcf per week. Nine weeks of that pace takes 167 Bcf to zero — right at the start of the withdrawal season.

That is the scenario where the record 3,985 Bcf forecast does not materialize, and it is the trade that has the best asymmetry in this market because nobody is positioned for it.

The longer-term case is stronger than the near-term one and worth separating clearly. Demand growth of 2.5 Bcf/d exceeds supply growth of 0.9 Bcf/d in 2027. LNG exports grow 1.7 Bcf/d. Storage moves below the five-year average across the forecast horizon. The annual Henry Hub average rises from just under $3.50 in 2026 to just under $4.60 in 2027.

International prices at $22.70 in Europe and $23.07 in Asia show what the molecule is worth once it can reach a buyer.

None of that helps the October contract.

The Equity Complex Trades A Different Curve

The producer and infrastructure names respond to a different set of variables than the front-month contract, and the divergence is worth understanding.

EQT, Chesapeake and Coterra price off the forward strip and reserve economics rather than off $2.858 spot. A gas producer's net asset value is a function of the 2027 and 2028 curve, where the supply-demand balance flips to a 1.6 Bcf/d deficit and the annual average is projected just under $4.60. Front-month weakness compresses near-term cash flow without changing the reserve valuation.

Cheniere sits on the other side of the arbitrage entirely. Low Henry Hub input costs against $22.70 TTF and $23.07 JKM output prices widen liquefaction margins. Cheap domestic gas is a tailwind for an LNG exporter, not a headwind — which is why the equity and the commodity can move in opposite directions on the same day.

The rig count at 132 is the operational signal from the producer side. Adding five rigs in a week with the front month below $3.00 tells you producers are underwriting the forward curve rather than the spot. That is either discipline breaking down or conviction in the 2027 balance, and the answer determines whether the current oversupply resolves or extends.

For the retail exchange-traded products, the mechanics matter more than the direction. UNG holds front-month futures and rolls them, which means a contango curve produces persistent drag independent of the spot price. BOIL applies leverage to that same structure and decays faster. KOLD inverts the exposure and benefits from both the decline and the roll in a falling contango market.

None of these are buy-and-hold instruments. The curve structure penalizes duration on the long side.

Energy equities broadly caught a bid Tuesday on the oil side, with the Energy Select Sector SPDR gaining more than 1% premarket alongside Brent at $92.04. Gas-levered names have not participated in that move, which is the correct differentiation — the oil complex is trading a Hormuz supply shock while the gas complex is trading a domestic supply surplus.

Two energy markets, opposite directions, same session.

Forecast: Range Between $2.75 And $3.00 With Downside Bias

Weighting the evidence produces a defined distribution and a clear tilt.

The base case is a range between $2.75 and $3.00 through the second week of September, and it carries the highest probability. Above-average temperatures across the eastern two-thirds of the country from September 2 through September 11 hold power burn elevated enough to defend $2.80, while record production at 113.0 Bcf/d, the 2.2 Bcf/d Hugh Brinson addition and a 132 rig count cap every attempt at $3.00. The official third-quarter estimate of $2.87 sits inside that band and the market is trading beneath it.

The bear case triggers on a close below $2.858 followed by a break of $2.80. That opens $2.75 and then the summer lows. The catalyst is a storage build back above 30 Bcf on Thursday, September 3, a further Henry Hub reduction in the September 9 outlook, or the temperature forecast breaking after September 11. With end-October inventories projected at a record 3,985 Bcf — the highest pre-winter level since 2016 — the structural pressure has no offset until the withdrawal season begins.

The bull case requires reclaiming $2.92 and then closing above $3.00. That needs weekly injections to stay beneath 15 Bcf, compressing the 167 Bcf surplus at roughly 14 Bcf per week, plus LNG feedgas rising through 16.5 Bcf/d and the heat window extending past mid-September. The surplus has already narrowed from 198 Bcf to 167 Bcf across two reports while injections halved from 36 Bcf to 15 Bcf — that trend is the most under-owned data in the market.

Structural support: injections decelerating, five-year surplus compressing 31 Bcf in a fortnight, South Central salt drawing down, Canadian imports off 8%, total supply down 0.9 Bcf/d, Freeport back from maintenance with feedgas at its highest since late June, and a $20 international arbitrage.

Structural risk: 113.0 Bcf/d production up 4.5% year over year, 2.2 Bcf/d of new takeaway commissioned today, a rig count at a five-month high, record 3,985 Bcf end-October storage, and residential, commercial and industrial demand all forecast lower in 2026.

Verdict: bearish into September 11, neutral after. Sell rallies into $2.92 to $3.00 with a stop above $3.05. Buy $2.75 to $2.80 only if Thursday's injection prints below 20 Bcf. Stand aside on a close beneath $2.75 — the shoulder-season gap between the end of cooling load and the first cold snap has no natural buyer, and this is the year with the most gas in storage since 2016 to fill it.

That's TradingNEWS