Natural Gas Futures at $2.77 Against a 195 Bcf Surplus as the EIA Revises Its $3.70 Path Today

Front-month Henry Hub natural gas traded at $2.77 per million British thermal units on Tuesday, down 0.80% on the session and 4.33% over the past month, sitting 1.30% below where it traded a year ago. The contract has recovered from below $2.70 last week, a level that marked the lowest print since April, but the recovery has produced no structural change.

The reason the market cannot hold a bid is arithmetic. Working gas in storage reached 3,117 billion cubic feet as of Friday, July 31, which is 195 Bcf above the five-year average of 2,922 Bcf, a surplus of 6.7%. Lower 48 dry gas production averaged a record 111.2 Bcf/d during August, up from 110.7 Bcf/d in July. Inventories have held above the five-year average continuously since March.

Against that, the bullish inputs are real and smaller. Daily flows to the nine major US LNG export plants were on track to reach a one-month high of 17.9 Bcf/d, recovering from 16.9 Bcf/d earlier in the month when Freeport LNG operations were reduced and Golden Pass ran a single liquefaction train. Weather forecasts point to continued above-normal temperatures through August 25, which sustains power-generation demand as air-conditioning load stays elevated.

The most recent storage report broke the market's short recovery. Energy firms injected 33 Bcf during the week ended July 31, above the 28 to 31 Bcf consensus range and above survey estimates centered near 29 to 30 Bcf. That print compared with a 13 Bcf build in the same week last year and a five-year average of 23 Bcf, and it offset bullish surprises from the prior two weeks.

The single most consequential item on today's calendar is not a price level. The US Energy Information Administration publishes its August Short-Term Energy Outlook, and the July edition placed the Henry Hub spot price averaging close to $3.70 per MMBtu in 2026. Front-month gas at $2.77 sits 33.6% below that projection.

That gap is the analytical core of the current market. Either the agency's price path collapses in today's revision, or the front month is mispricing a fourth quarter the agency expects at $3.57.

The August storage report arrives Thursday, with the surplus expected to narrow slightly to 6.6% above normal for the week ending August 7.

$2.77 After Testing The Lowest Level Since April

The price action across two weeks describes a market that found a floor without finding a reason to leave it.

The NYMEX September contract closed at $2.69 on August 5 and traded down to $2.62 the following session, a $0.06 decline that took the front month to its lowest level in more than three months. September averaged $2.69 that week, down from $2.74 the prior week.

The recovery to $2.77 represents a 5.7% advance from the $2.62 low, which is a normal countertrend move rather than a breakout. Every element of that bounce traces to weather forecasts extending above-normal temperatures through August 25 and to LNG feedgas recovering toward 17.9 Bcf/d.

The technical structure is narrow. The $2.70 level, which contained the market on the way down and again on the way back up, functions as the operative pivot. Below it sits $2.62, the recent low, and beneath that the April lows that have not been tested in the current cycle. Above $2.77, the $2.90 area represents the first meaningful resistance, with $3.00 as the psychological barrier the front month has not held since the winter withdrawal season ended.

The distance to each reference is small in absolute terms and large in percentage terms, which is the defining characteristic of trading a sub-$3 gas market. A $0.13 move from $2.77 to $2.90 is 4.7%. A move to $3.00 is 8.3%. A break to $2.62 is 5.4% lower. Position sizing in this market has to account for a commodity where a routine daily range consumes what would be a monthly move in most assets.

The twelve-month comparison is the most revealing single figure. Gas at $2.77 sits 1.30% below where it traded a year ago, which means the entire 2026 story of record LNG exports, record power burn, and record data center electricity demand has produced no net price appreciation. Production growth absorbed all of it.

3,117 Bcf Is 195 Bcf Above The Five-Year Average

The storage position is the reason this market cannot rally, and the trajectory of the surplus tells the story better than the level.

Working gas in storage stood at 3,117 Bcf as of July 31, a net increase of 33 Bcf from the prior week. Stocks were 12 Bcf below the same point last year and 195 Bcf above the five-year average of 2,922 Bcf. Total working gas remained within the five-year historical range.

The weekly progression through July shows the surplus building rather than eroding. The week ended July 10 delivered a 43 Bcf injection to 3,024 Bcf. July 17 added 32 Bcf to 3,056 Bcf, leaving stocks 183 Bcf above a five-year average of 2,873 Bcf. July 24 added 28 Bcf to 3,084 Bcf, 185 Bcf above a 2,899 Bcf average, or 6.4% above. July 31 added 33 Bcf to 3,117 Bcf, 195 Bcf above average at 6.7%.

The surplus expanded from 183 Bcf to 195 Bcf across three weeks during the peak of cooling season. That is the period when the surplus should compress, since air-conditioning demand pulls gas out of the injection stream. It did not compress.

The context that makes this more significant is where the market started the year. Working gas stood at 1,848 Bcf as of March 6, which was 17 Bcf below the five-year average of 1,865 Bcf despite being 141 Bcf higher than the prior year. The market entered spring in a deficit position and has spent five months converting that into a 195 Bcf surplus.

Strong output combined with relatively mild weather earlier this year kept inventories above the five-year average since March. That is the specific mechanism: production growth outrunning weather-driven demand across the shoulder season and into summer.

The forward expectation is that the surplus narrows slightly to 6.6% above normal for the week ending August 7. A 0.1 percentage point compression from 6.7% is not a change in condition.

The +33 Bcf Injection Beat Consensus And Broke Two Bullish Prints

The most recent report was the specific event that ended the market's attempt to base above $2.70.

The 33 Bcf injection for the week ending July 31 came in above the 28 to 31 Bcf consensus range and above survey estimates centered around 29 to 30 Bcf. It compared with a 13 Bcf build in the same week last year and a five-year average build of 23 Bcf.

The magnitude of the surprise relative to seasonal norms is what mattered. A 33 Bcf build against a 23 Bcf five-year average is 43% above normal. Against last year's 13 Bcf, it is 154% above. During the hottest weeks of the year, the market injected at a rate more consistent with October than with late July.

The sequencing amplified the effect. The prior two weeks delivered bullish prints, with the 28 Bcf build for July 24 coming below expectations for 35 Bcf and the 32 Bcf build for July 17 landing slightly below the 35 Bcf consensus. Two consecutive misses to the tight side had begun to establish a narrative that summer demand was outrunning supply.

The July 31 report reversed that in a single release, which is why the front month fell to $2.62. Markets that have started pricing a change in fundamentals react more violently to contradicting data than markets holding an established view.

Production data during the same period was mixed rather than uniformly bearish. Output was revised to 110.1 Bcf/d on Permian and Northeast intraday revisions, with volumes pulling back further as Transco maintenance continued in the Northeast. Canadian imports averaged 4.9 Bcf/d on the week. LNG feedgas reached 18.6 Bcf/d, with Golden Pass nominations up 0.3 Bcf/d to 0.558 Bcf/d, a monthly high on early-cycle data.

Those were bullish inputs on the same day as the bearish storage print, and the storage number won. That hierarchy is instructive: in a market carrying a 195 Bcf surplus, inventory data dominates flow data.

Thursday's report for the week ending August 7 is the next test.

Lower 48 Production At A Record 111.2 Bcf/d Caps Every Rally

Supply is the structural constraint on this market, and the production figures have kept setting records through a price decline.

Lower 48 dry gas production averaged a record 111.2 Bcf/d during August, up from 110.7 Bcf/d in July. Earlier in the month the running average sat at 110.6 Bcf/d, and intraday revisions placed output at 110.1 Bcf/d during a period of Transco maintenance in the Northeast.

Production rising from 110.7 to 111.2 Bcf/d while the front month fell from $2.74 to $2.62 is the definition of a supply-driven market. Producers are not responding to price because the marginal molecule is associated gas from oil-directed drilling and Permian development where the gas is a byproduct rather than the target.

That structure removes the price mechanism that historically balanced this market. When gas traded below $3 in previous cycles, dry gas rigs came down and production followed within two quarters. Associated gas from Permian oil wells does not respond to gas prices, and with West Texas Intermediate at $83.33 there is no signal telling operators to slow down.

The EIA has identified the Permian specifically as the region leading the production growth that keeps inventories high. Record output, led by that growth, is what the agency expects to meet rising demand while limiting upward price pressure.

The forward supply path reinforces it. The agency's framework has supply growing by 1.1 Bcf/d in 2026, or nearly 1%, against demand growth of 0.6 Bcf/d, or less than 1%. Supply growth outpaces demand growth by 0.5 Bcf/d in 2026.

That 0.5 Bcf/d surplus is roughly 3.5 Bcf per week, or 182 Bcf annualized. It is close to the exact size of the storage surplus that has developed, which confirms the balance rather than the weather is producing the inventory build.

Canadian imports at 4.9 Bcf/d add to the total available supply, and maintenance events like the Transco work in the Northeast produce temporary reductions that resolve within days rather than changing the annual balance.

LNG Feedgas At 17.9 Bcf/d Is The Only Bullish Variable Working

Export demand is the single growth engine in this market, and its recent volatility has been the main source of price movement.

Daily flows to the nine major US LNG export plants were on track to reach a one-month high of 17.9 Bcf/d. Average feedgas demand stood at 17.2 Bcf/d in July, just below June's monthly record of 17.4 Bcf/d. Earlier in August the average had fallen to 16.9 Bcf/d, driven by reduced operations at Freeport LNG in Texas and a single operating liquefaction train at Golden Pass.

The recovery from 16.9 to 17.9 Bcf/d represents 1.0 Bcf/d of restored demand, which is 5.9% of the total feedgas figure and roughly 7 Bcf per week of additional storage withdrawal pressure. That is meaningful against weekly injections running 28 to 43 Bcf.

The Golden Pass ramp is the variable to watch. Nominations rose 0.3 Bcf/d to 0.558 Bcf/d on early-cycle data, a monthly high, and the facility operating a single train means substantial capacity remains uncommissioned. Each additional train converts directly into Henry Hub demand.

The multi-year trajectory is the structural bull case. US LNG exports averaged 11.9 Bcf/d in 2024 and 15.1 Bcf/d in 2025, and the agency's projection places them at 17.4 Bcf/d in 2026 and 18.6 Bcf/d in 2027. That is growth of 9%, or 1.3 Bcf/d, in 2026 and 11%, or 1.7 Bcf/d, in 2027.

The problem for the front month is timing. A 1.3 Bcf/d increase spread across 2026 has already largely occurred, since feedgas reached a 17.4 Bcf/d monthly record in June. The 1.7 Bcf/d of 2027 growth arrives after the current storage surplus has to be worked off.

Maintenance risk cuts both ways and has been the dominant short-term driver. A Freeport outage removes roughly 2 Bcf/d and moves the front month 5% in a session. A return restores it. Neither changes the annual balance, which is why these moves reverse.

The nine-plant configuration means concentration risk is high. Two facilities running below capacity accounts for the entire 0.5 Bcf/d gap between July's average and June's record.

Above-Normal Temperatures Through August 25 Are Already Priced

Weather is the variable the market trades daily and the one that has provided the least durable support.

Forecasts point to continued above-normal temperatures through August 25, which should sustain gas demand from power generators as air-conditioning use remains elevated. Earlier in the month the picture was the opposite, with forecasts pointing to moderating temperatures across much of the country and reducing the likelihood of a significant increase in consumption.

That reversal from moderating to above-normal within days produced the bounce from $2.62 to $2.77. It also demonstrates how little the weather signal is worth in a market carrying a 195 Bcf surplus.

The arithmetic explains why. A hot August week adds roughly 3 to 5 Bcf/d of power burn against normal, or 21 to 35 Bcf across a week. The surplus is 195 Bcf. Six consecutive weeks of above-normal heat would be required to eliminate it, and the injection season ends before that could complete.

The seasonal transition is the harder constraint. Cooling demand peaks in late July and declines through August and September. Above-normal temperatures through August 25 support the tape for two more weeks, after which the market enters the shoulder period where injections accelerate mechanically as power burn falls and heating demand has not yet begun.

That transition is when the surplus typically expands fastest. A market entering the shoulder season with 195 Bcf of excess inventory and record production faces its weakest fundamental window in September and October.

The offsetting consideration is that weather also drives the withdrawal season, and a 3,966 Bcf October starting point going into a cold winter draws down faster than a normal one. That is a fourth-quarter story that the front month does not price.

Wholesale electricity prices are forecast to average about $45 per megawatt hour nationally this summer, lower than last summer, primarily because of lower delivered natural gas costs. Heatwaves during the summer could still cause price spikes, with the largest declines occurring in western hubs and the Midcontinent ISO region.

The EIA's $3.70 Forecast Sits 34% Above The Front Month

The gap between the official projection and the traded market is the largest single anomaly in this commodity.

Per the EIA's Short-Term Energy Outlook released July 7, 2026, record US natural gas production helps meet rising demand, putting moderate downward pressure on prices, with the Henry Hub spot price averaging close to $3.70 per MMBtu in 2026 before declining below $3.50 next year. The fourth-quarter 2026 average is projected at $3.57 per MMBtu, 5% less than the same quarter last year.

The full-year overview places the spot price at $2.19 in 2024, $3.53 in 2025, $3.67 projected for 2026, and $3.49 projected for 2027.

Front-month gas at $2.77 sits 33.6% below the $3.70 annual projection and 28.9% below the $3.57 fourth-quarter figure. That is not a modest divergence between a model and a market. It is a structural disagreement.

Reconciling it requires one of three things. The first is that the year-to-date realized average has been high enough to carry the annual figure despite a weak second half, which the winter withdrawal season and the January price spike would support. The second is that the agency expects a substantial fourth-quarter rally the forward curve does not price. The third is that the projection is stale and today's revision cuts it.

The first explanation carries the most weight. Natural gas traded above $4.00 at most US hubs during January, with the February NYMEX contract reaching $4.875 during the winter squeeze. A first quarter averaging near $4.50 against a third quarter near $2.80 produces an annual average well above the current front month without requiring any fourth-quarter rally.

That decomposition matters for positioning. Traders comparing $2.77 to a $3.70 annual forecast and concluding the market is 34% cheap are comparing a summer contract to an annual average that includes a winter spike. The relevant comparison is $2.77 against the $3.57 fourth-quarter projection, and that 29% gap is what the December and January contracts have to close.

Today's August Outlook Has To Address The Third Quarter

The revision arriving today is the most consequential scheduled event in this market, and the direction is not obvious.

The July edition was released July 7 with the forecast completed July 1, and the next release date is August 11. That means the current published projection predates the entire July storage build sequence, the record 110.7 Bcf/d July production figure, the record 111.2 Bcf/d August average, and the break of the front month below $2.70.

The storage assumption is where the revision pressure sits. The July outlook stated that at the end of June, working inventories were 6% above the five-year average, and forecast inventories reaching 3,966 Bcf by the end of October, 5% above the five-year average.

Inventories are now 6.7% above the average rather than 6%, and the surplus expanded during the period the forecast expected it to narrow toward 5%. That trajectory argues for an upward revision to the October figure and a downward revision to the price path.

The mechanics of reaching 3,966 Bcf are worth checking. From 3,117 Bcf on July 31 to 3,966 Bcf by October 31 requires 849 Bcf of injections across roughly 13 weeks, an average of 65 Bcf per week. Recent builds have run 28 to 43 Bcf, which means the forecast requires injection rates to roughly double as cooling demand fades.

Seasonal patterns support that acceleration. September and October builds routinely run 60 to 90 Bcf per week once power burn falls away. The 65 Bcf weekly average is achievable, and if production holds at 111.2 Bcf/d the market could exceed it.

An October inventory figure above 4,000 Bcf would place the surplus wider than 5% and would justify cutting the fourth-quarter $3.57 projection toward $3.20 or lower. That revision would validate the front month rather than contradict it.

The oil side of the same document faces the opposite problem, with Brent at $88.89 against a $74 third-quarter projection built on a Hormuz reopening that never occurred.

The October Target Of 3,966 Bcf Decides The Winter

The end-of-injection-season inventory level is the number that determines whether this market has a fourth quarter.

With above-average inventories heading into winter, the agency expects the Henry Hub spot price in the fourth quarter to average $3.57 per MMBtu. That projection assumes 3,966 Bcf at end-October and 5% above the five-year average.

The relationship between starting inventory and winter price is close to mechanical. Periods with higher-than-average inventories are generally associated with lower prices, while lower storage levels correspond with higher prices and tighter market conditions. A 5% surplus entering November caps the upside available from cold weather because the market has the cushion to absorb a demand shock.

The comparison with the prior two years frames it. Storage levels had been relatively high in 2024 and 2025, with inventories remaining 1.7% above the five-year 2020 to 2024 average at the close of December 2025. A 5% surplus at end-October 2026 would be a wider cushion than the market carried into either of the last two winters.

The offsetting projection is that inventories gradually move below the rolling five-year average across the forecast period as demand outpaces supply. That transition occurs in 2027 rather than 2026, which places the tightening one full injection cycle away.

For the front month, the implication is that the current $2.77 print is trading the 2026 balance and the curve's upward slope is trading the 2027 balance. The December and January contracts carry the winter risk premium, and the September contract carries none of it.

The winter squeeze precedent is recent and severe. During the January 2026 episode, the February NYMEX contract rose $1.76 in a single report week, from $3.120 to $4.875 per MMBtu, with the 12-month strip climbing 65 cents to $3.970. That move was mostly a reaction to anticipated changes in 2026 storage balances rather than to spot conditions.

A market that repriced 56% in one week on storage expectations can do so again. Entering winter with 3,966 Bcf makes that less likely than entering with 3,600 Bcf.

Power Burn Reaching 38.1 Bcf/d In 2027 Is The Structural Bid

The demand growth that eventually absorbs this surplus comes from electricity generation, and the scale is substantial.

US natural gas consumption in the electric power sector is forecast to increase in 2026 and 2027, reaching a record next year. Average consumption in the sector rises by 2% in 2026 and by another 4% in 2027 to 38.1 Bcf/d. On a monthly basis, consumption is forecast to reach 50.6 Bcf/d in July 2027, which would be the most in any month on record.

The drivers are rising overall electricity demand, additions to the natural gas generation fleet, and relatively low natural gas prices. Data reported by generators indicate 508 gigawatts of natural gas-fired generating capacity in the United States by the end of 2027.

That 508 GW figure is the physical constraint on how much gas power generation can consume, and the fleet is growing. Every gigawatt added at roughly 7,000 British thermal units per kilowatt hour heat rate at full utilization consumes approximately 0.17 Bcf/d.

The reason this matters more in 2026 than in prior cycles is the source of the electricity demand. Data center construction tied to artificial intelligence infrastructure has become the marginal load growth in multiple regions, and gas generation is the dispatchable supply meeting it. Listed operators have announced more than $70 billion in AI and high-performance computing contracts, and that capacity requires firm power.

The 2% growth in 2026 against a 195 Bcf storage surplus is the timing problem. Power sector consumption rising 2% adds roughly 0.7 Bcf/d, or 5 Bcf per week, which is less than the surplus expands during a single bearish storage print.

Industrial demand runs the other way. Consumption in the industrial sector decreases in both 2026 and 2027 because of closer-to-normal weather and decreased industrial activity as measured by the gas-weighted manufacturing index. That contraction partially offsets the power sector growth.

The net effect for 2026 is demand growth below 1% against supply growth near 1%, which is the balance producing the current surplus.

A 2027 Deficit Of 1.6 Bcf/d Is Why The Curve Prices Higher

The forward market is trading a different year than the front month, and the projected balance shift explains the shape.

Annual average spot prices are forecast to decrease by 2% in 2026 and then increase by 33% in 2027. Supply growth outpaces demand growth by 0.5 Bcf/d in 2026 but then falls behind by 1.6 Bcf/d in 2027, putting upward pressure on prices.

The 2027 composition is where the shift originates. Demand growth of 2.5 Bcf/d exceeds supply growth of 0.9 Bcf/d, a reversal from 2026 where supply added 1.1 Bcf/d against 0.6 Bcf/d of demand. LNG exports grow 11%, or 1.7 Bcf/d, in 2027 after 9%, or 1.3 Bcf/d, in 2026.

A 1.6 Bcf/d deficit sustained across a year removes roughly 584 Bcf from storage, which is three times the current 195 Bcf surplus. That is the mechanism by which the market transitions from comfortable to tight, and it operates entirely through export capacity coming online rather than through weather.

The complication in that framework is that the agency's own price projections do not reflect the 33% increase. The overview table places 2026 at $3.67 and 2027 at $3.49, which is a decline rather than an increase, and the narrative text projects the Henry Hub averaging below $3.50 next year after close to $3.70 in 2026.

Those two statements are reconcilable only if the 33% figure refers to a different baseline period or if the projections were revised between publications. The internal tension is itself informative: the agency's balance work points toward tightening while its price work points toward softening, and record production is the variable resolving it in favor of lower prices.

For a trader, the practical read is that the 2027 curve carries the tightening thesis and the 2026 curve carries the surplus. The calendar spread between them is the cleanest expression of the structural view, and the front month is not the instrument for it.

Europe At Roughly $20.80 Pays 7.5 Times Henry Hub

The international price differential is the widest structural feature in global gas and it explains the export growth trajectory.

European gas at the Title Transfer Facility traded at €61.64 per megawatt hour, up 1.39%. At the prevailing euro rate near 1.154 and the standard conversion, that equates to roughly $20.80 per MMBtu. Henry Hub at $2.77 means European buyers pay approximately 7.5 times the US benchmark.

That spread is the economic engine behind every liquefaction facility under construction on the Gulf Coast. Liquefaction, shipping, and regasification costs run roughly $2.50 to $3.50 per MMBtu on the Atlantic route, which means the arbitrage is open by more than $14 per MMBtu at current prices.

An arbitrage that wide guarantees maximum utilization of every available train. It also explains why feedgas at 17.9 Bcf/d represents a facility-constrained figure rather than a demand-constrained one: US export capacity is running as hard as the physical plant permits.

The European price level traces to the same geopolitical event driving crude. The Strait of Hormuz remains closed, Qatari LNG cargoes that would normally transit it are disrupted, and Brent at $88.89 with oil-indexed Asian contracts pulls incremental cargoes toward the Pacific. Europe is paying to compete for the marginal cargo.

Historical context shows how much conditions have changed. During the week ending January 22, 2025, East Asian LNG front-month prices averaged $14.01 per MMBtu and TTF averaged $14.57. Both benchmarks now sit substantially higher while Henry Hub trades below where it did.

The consequence for US prices is asymmetric and limited. Export capacity is the binding constraint, so a European price of $20 or $30 makes no difference to US demand until new trains commission. The spread creates enormous margin for exporters without transmitting into Henry Hub.

That insulation is why US gas at $2.77 coexists with European gas at $20.80. The two markets are connected by a pipe that is already full.

Brent At $88.89 Does Not Transmit Into US Gas

The relationship between the oil complex and Henry Hub has been the most misread variable in this market.

West Texas Intermediate traded at $83.33, up 1.45%, and Brent at $88.89, up 1.33%, in a fourth consecutive session of gains as US-Iran negotiations moved backward. President Trump introduced new demands on Tehran while stating a preference for allowing economic pressure to accumulate. Iran and Oman have not finalized an agreement to reopen the strait.

Earlier in August, improving prospects for a US-Iran agreement and potential reopening of Hormuz reduced energy market concerns and added downward pressure on natural gas prices. That transmission channel operated through sentiment rather than through physical balances, and it has since reversed without lifting Henry Hub.

The reason the correlation is weak is structural. Associated gas from Permian oil development means higher crude prices increase US gas supply rather than decreasing it. A $10 rise in WTI accelerates oil-directed drilling, which produces more associated gas, which raises Lower 48 output. Oil strength is gas-bearish on the supply side.

On the demand side, oil-to-gas substitution in US power generation is minimal because residual fuel oil has been displaced from the generation stack. The substitution effect that operates in Asian and European markets does not exist domestically.

The one channel that does transmit is LNG netbacks. Oil-indexed Asian LNG contracts price off Brent, and Brent at $88.89 raises those netbacks, which pulls cargoes toward the Pacific basin and supports European prices. That increases the value of US export capacity without increasing US export volume, since volume is plant-limited.

The practical implication is that traders using crude strength as a bullish gas signal are trading a relationship that has inverted. Record Permian oil activity at $83 WTI is the reason Lower 48 gas production reached 111.2 Bcf/d.

The macro overlay is more relevant than the crude one. July CPI arrives Wednesday at 8:30 a.m. Eastern Time, with headline expected at 3.4% and core at 2.5%, and the 10-year Treasury at 4.726% sets the cost of capital for the entire midstream buildout.

Natural Gas Futures Price Forecast: Levels And Invalidation

The base case holds the front month between $2.55 and $3.00 through the shoulder season, with $2.70 as the operative pivot and the 195 Bcf storage surplus as the structural cap.

The bullish path requires three confirmations. First, a close above $2.90, which clears the resistance that has capped the recovery from $2.62. Second, a close above $3.00, the psychological level the front month has not held since the withdrawal season ended. Third, a storage print for the week ending August 7 that narrows the surplus below 6.0% rather than the 6.6% currently expected, which would establish that summer demand is finally outrunning the production surplus. Clearing all three opens $3.20 and the fourth-quarter projection at $3.57, with the January contract carrying the winter premium beyond that.

The bearish path requires two. A close below $2.70, which returns the front month to the range that produced the April-equivalent low. Then a break of $2.62, which opens the April lows with no recent structure beneath. That sequence targets $2.50, and a September and October injection sequence running at the 65 Bcf weekly pace the October forecast requires would push end-of-season inventories above 4,000 Bcf and justify a move into the low $2.40s.

Invalidation for the bullish case is a close below $2.62. Invalidation for the bearish case is a close above $3.00.

The medium-term structure is bearish through the injection season and constructive into 2027, and the two do not conflict. Working gas at 3,117 Bcf sits 195 Bcf and 6.7% above the five-year average, having widened from 183 Bcf across three weeks of peak cooling demand. Lower 48 production reached a record 111.2 Bcf/d in August against 110.7 Bcf/d in July. Supply growth outpaces demand growth by 0.5 Bcf/d in 2026. Industrial consumption declines in both years. The October target of 3,966 Bcf leaves a 5% surplus entering winter.

The 2027 case is the mirror image. Demand growth of 2.5 Bcf/d against supply growth of 0.9 Bcf/d produces a 1.6 Bcf/d deficit that removes roughly 584 Bcf from storage across the year. LNG exports rise 11%, or 1.7 Bcf/d, to 18.6 Bcf/d. Power sector consumption reaches a record 38.1 Bcf/d with a monthly peak of 50.6 Bcf/d in July 2027 against 508 GW of installed gas capacity. Feedgas at 17.9 Bcf/d is already a one-month high with Golden Pass running a single train.

The resolution runs through today's August Short-Term Energy Outlook and Thursday's storage report. The July outlook placed the 2026 Henry Hub average at close to $3.70 and the fourth quarter at $3.57 against a front month at $2.77, and the revision either closes that 29% gap or confirms the front month was right.

The trade is the $2.70 to $2.90 box. Above $2.90 with a supportive storage print, $3.00 and $3.20 come into play. Below $2.62, the April lows open and $2.50 becomes the target. Europe at roughly $20.80 per MMBtu will keep every export train full either way, and it will not lift Henry Hub until new capacity commissions.