Natural Gas Holds $2.79 as 16 Bcf Build Undershoots the 5-Year Average by 13 Bcf

Natural gas traded around $2.79 per million British thermal units on Friday, up more than 2% and heading for a second consecutive weekly advance. That follows Thursday's slip to $2.72, a retreat from what had been a near four-week high, and Wednesday's push through $2.800 on hotter overnight weather models and a pullback in production.

The week's peak came Wednesday when the front month ran to $2.875 before stalling just short of the 50-day moving average at $2.945. That was the first time in weeks the main trend flipped to up on the daily chart.

The base underneath this bounce was ugly. On August 16 the contract was at $2.652, down 2.03% on the session and 4.26% across seven days. August 17 delivered another 2.07% decline. Prompt-month futures had repeatedly failed to sustain breakouts above the $2.80 to $2.83 resistance band, with systematic selling driving price back toward $2.70 to $2.72 each time.

That pattern has now been tested three times in ten sessions and has held on every attempt.

The context from July is why $2.79 feels like a recovery rather than a level. Henry Hub futures fell nearly 15% during that month, from $3.22 on July 1 to $2.75 on July 31. Day-ahead spot did worse, dropping 22.7% from $3.35 to $2.59. The decline came on record production and softer LNG feedgas demand, and it happened despite the first heatwave of the summer pushing power-sector demand higher.

Prices had been comparatively steady through mid-June, trading a $3.15 to $3.34 range on strong supply fundamentals. Everything below $3.00 has been carved out since.

The EIA's own forecast now sits below where the market trades. The agency cut its third-quarter Henry Hub projection to $2.87 per MMBtu in the August Short-Term Energy Outlook, down 50 cents from the July estimate, citing reduced LNG feedgas demand and record production.

At $2.79 the front month is trading eight cents beneath that quarterly average with five weeks left in the quarter.

The setup into the final stretch of summer is a genuine two-sided fight: warmer-than-normal forecasts running through September 3 against inventories that remain comfortably above seasonal norms and a pipeline expansion landing on September 1.

The 16 Bcf Build Was Bullish And The Tape Sold It Anyway

Thursday's storage print was the most constructive number this market has produced in months, and the reaction told you everything about the current balance of power.

The EIA reported a 16 Bcf injection into working gas storage for the week ending August 14. That came in below both the 19 Bcf build recorded during the same period last year and the five-year average of 29 Bcf. It nearly matched consensus expectations, with a hefty South Central withdrawal limiting the overall build.

A 13 Bcf undershoot against the five-year norm during peak cooling season is exactly what a bull needs. It says heat is finally pulling enough gas into the power sector to slow the pace of storage accumulation.

Futures floundered anyway. The front month slipped to $2.72 on Thursday, retreating from the near four-week high, as the market sized up a fundamental backdrop defined in large part by ample supply. Daily cash prices lost ground across the board. Comfortable supply expectations kept NYMEX futures in negative territory despite downwardly trending production and supportive heat in the Lower 48 forecast.

That is the tell. When a genuinely bullish inventory print produces a lower close, the marginal participant is not trading week-to-week balances. They are trading the aggregate.

The prior week's number explains why. The August 7 report delivered a larger-than-expected 36 Bcf injection, pushing total stockpiles to 3,153 Bcf and widening the surplus to 198 Bcf — 6.7% above the five-year norm. Adding this week's 16 Bcf puts working gas near 3,169 Bcf.

The comparison against last year has narrowed. Stocks now sit roughly 28 Bcf below the same week in 2025 while remaining approximately 185 Bcf above the five-year average. The year-over-year deficit is new; the five-year surplus is the problem.

The framing from the trading side ahead of the report was direct: buyers needed the number to come in tight or the weather trade keeps failing at the top of the range. It came in tight. The trade failed anyway on the day, then recovered Friday.

Two consecutive sub-average injections would change the arithmetic. One does not.

Storage At 3,169 Bcf And The Record 3,985 Waiting In October

The end-of-season projection is the single most bearish number on the board and it has not moved.

The EIA expects natural gas inventories to reach a record 3,985 Bcf at the end of October 2026 — an increase of 19 Bcf compared with the July outlook and 5% above the five-year average. That would leave the market with the highest storage level heading into winter since 2016.

Working from the current 3,169 Bcf, reaching 3,985 Bcf by the end of October requires roughly 816 Bcf of net injection across eleven weeks — an average of 74 Bcf per week. Against injections of 16 Bcf and 36 Bcf in the last two reports, that pace looks unreachable on current fundamentals, which suggests either the forecast is stale or the market expects a substantial cooling in demand as the shoulder season arrives.

The consequence for price is explicit in the agency's own language. With high storage heading into winter, the Henry Hub spot price is expected to remain below $3.00 until November and average $3.03 over the remaining five months of the year — nearly 50 cents lower than the prior month's forecast.

That projection strips away the winter-scarcity premium bulls would otherwise lean on. Futures prices show a similar pattern, with contracts through September 2026 remaining subdued.

The historical anchor is worth carrying. Inventories at the end of October 2025 stood at 3,915 Bcf, 4% above the five-year average and near a high point for the decade. The market went on to see January 2026 futures rally toward $5.50 anyway, because Winter Storm Fern produced the largest weekly net withdrawal in the history of the EIA's storage report and January spot averaged $7.72.

That episode is the reason nobody should treat an October surplus as a settled outcome. Storage sets the starting point; weather sets the path.

The current forecast assumes closer-to-normal temperatures, which is what produced the reduction in residential and commercial consumption estimates versus 2025's colder-than-normal winter months.

A repeat of Fern with 3,985 Bcf in the ground is a very different market than the same storm with 3,500 Bcf.

Production At 113.5 Bcf/d Is The Whole Bear Case

Everything that has capped this market since June traces back to one number.

Lower-48 dry gas production hit 113.5 Bcf per day in the most recent full reading, up 4.4% from a year earlier. Demand over the same window was 83.1 Bcf per day, down 0.5%. That is a 30.4 Bcf per day gross surplus before exports, and it is why every heat-driven rally has stalled at the top of the range.

Output has been climbing all year. Production hovered near 107 Bcf per day through the autumn of 2025 and topped 114 Bcf per day at points this summer — the highest levels in more than two and a half months at the time. Strong output across the major shale basins, particularly associated gas from the Permian alongside stable Haynesville drilling, has kept the physical market oversupplied.

The associated gas dynamic is the structural problem and it is getting worse, not better. Crude at $86.94 WTI and $93.96 Brent encourages more Permian drilling, and more Permian drilling produces more associated gas regardless of whether gas prices justify the volume. Higher oil prices can increase natural gas supply directly. With the Iran conflict keeping crude elevated, that mechanism runs continuously.

Consolidation is accelerating it. Continental Resources announced it will significantly expand its Permian position through the acquisition of private equity-backed FireBird Energy II — capital moving into the basin that produces the gas that caps the price.

There have been signs of stabilization. The Wednesday rally was helped by a pullback in production, and Thursday's session saw downwardly trending output alongside supportive heat. Those are one-day effects rather than a trend.

The EIA has been raising its production estimates through the year. The agency's forecast reduction for third-quarter Henry Hub prices was attributed explicitly to reduced LNG feedgas demand and record natural gas production.

For price to break structurally higher, either production has to roll over or demand has to step up materially. The Appalachian curtailment discussion is the first evidence of the former, and it is a fall story rather than a summer one.

LNG Feedgas At 17.2 Bcf/d — Freeport Is The Swing Factor

The demand side that was supposed to absorb this production has been offline.

Average gas flows to the nine major US LNG export facilities held at 17.2 Bcf per day in August, slightly below June's record of 17.4 Bcf per day. Individual daily readings have run between 17.1 and 17.9 Bcf per day depending on maintenance schedules, with one recent print at 17.9 Bcf per day representing a 1.8% weekly decline.

The single largest swing factor is Freeport LNG. Maintenance began July 10 and is expected to complete in late August, affecting 2.0 Bcf per day of nominal export capacity. Roughly 2 Bcf per day of offline capacity at a single terminal is more than 11% of total feedgas demand, and every molecule not moving into that facility flows back into domestic pipelines and storage.

That maintenance window ends within days. Freeport returning to full rates would add back demand equivalent to more than two weeks of average storage injections at the current pace, arriving just as the shoulder season begins.

The EIA expects US LNG exports to average 16.5 Bcf per day in the third quarter, down 0.2 Bcf per day from the prior forecast. Even with Freeport fully operational, exports remain limited by slow growth in additional capacity despite elevated US price spreads to Europe and Asia.

Those spreads are elevated for a specific reason. LNG vessel traffic through the Strait of Hormuz slowed considerably after strikes on vessels resumed on July 7, pushing international prices in July to levels last reached in early April. Qatar is the world's second-largest LNG exporter and its entire export volume except deliveries to Kuwait transits Hormuz. With transits running 73 in the week ended August 16 against roughly 130 daily pre-war, European and Asian buyers have been pushed toward US supply.

The structural picture supports the winter curve. Pipeline exports are forecast at 9.6 Bcf per day in 2026 rising to 10.0 Bcf per day in 2027, up from 9.5 Bcf per day in 2025, driven partly by the new Energia Costa Azul terminal that shipped its first cargo on July 8 and adds 0.4 Bcf per day of nominal capacity on Mexico's Pacific coast.

Roughly 15% of all US dry gas production is now committed to overseas buyers.

The EIA Cut Its Third-Quarter Forecast To $2.87 And Price Beat It

The official forecast has been chasing the market lower all year, and the August revision was the largest yet.

The EIA now sees the Henry Hub spot price averaging $2.87 per MMBtu in the third quarter of 2026 — a 50-cent reduction from the prior month's estimate. The quarterly path runs $3.14 in the fourth quarter, $3.62 in the first quarter of 2027, $2.79 in the second, and $3.20 in the third.

Annual averages were cut to $3.44 for 2026 from $3.67, and to $3.31 for 2027 from $3.49. Both benchmarks sit below the $3.53 actual recorded in 2025.

The revision history is a case study in how quickly this market has repriced. In December 2025 the agency projected Henry Hub averaging $4.30 across the November-to-March heating season and $4.01 for all of 2026. In February 2026, after Winter Storm Fern produced record withdrawals and January spot averaged $7.72, the forecast went to $4.30 for 2026 and $4.40 for 2027. In January the projection had been just under $3.50 for 2026 and just under $4.60 for 2027.

Eight months later the 2026 number is $3.44 and the 2027 number is $3.31. That is a $0.87 and $1.09 downward revision from the February peak.

The driver of every cut has been the same pairing: record production and softer LNG feedgas demand. The August STEO named both explicitly.

There is a longer-term counterweight in the same document set. The agency's long-term outlook has Henry Hub reaching $3.80 per MMBtu by 2030 as LNG exports scale past 20 Bcf per day and AI data centre demand adds a persistent new load floor, climbing to $4.20 by 2040. More aggressive independent forecasts put 2030 at $5.40 and 2040 at $6.35 on sustained Asian and European LNG demand plus US data centre power needs outpacing supply growth.

The structural floor has risen. With US LNG at 17 Bcf per day and climbing, sub-$2 prices are increasingly a historical anomaly rather than a live risk.

What that framework does not resolve is the next six months.

Hugh Brinson Adds 1.5 Bcf/d On September 1

There is a dated supply event landing in eleven days that the market has been discounting since July.

Energy Transfer's Hugh Brinson pipeline reaches full capacity of 1.5 Bcf per day on September 1. That routes additional Permian gas directly toward Henry Hub in the month when cooling demand starts rolling off and before winter heating picks up.

The timing is the problem. Peak air-conditioning burn typically fades through the first half of September. Winter space-heating demand does not begin meaningfully until late October or November. That leaves a six-to-eight-week window where seasonal demand is at its annual trough — and 1.5 Bcf per day of incremental takeaway capacity arrives precisely inside it.

At current injection rates, 1.5 Bcf per day compounds to roughly 10.5 Bcf per week of additional supply reaching the market. Against an August 14 injection of 16 Bcf, that is a material shift in the weekly balance.

The mechanism is worth understanding. Permian associated gas has been constrained by takeaway capacity rather than production capability — wells produce the gas as a byproduct of oil, and when pipelines fill, the gas gets flared or the well gets curtailed. Adding takeaway does not create new gas; it delivers gas that was previously stranded to a pricing point where it competes with everything else.

That is why regional forwards have been splitting. During the August 13 to 19 trading period, western hubs added more than 25 cents while West Texas and Northeast points shed close to 20 cents. The market is pricing basis differentials around exactly this dynamic.

The offset is Freeport returning from maintenance in late August with 2.0 Bcf per day of nominal capacity. If both events land within days of each other, they roughly cancel — 1.5 Bcf per day of new supply against 2.0 Bcf per day of restored export demand.

That sequencing is the single most important thing to watch over the next fortnight, and the outcome determines whether the September balance tightens or loosens.

Neither event is weather-dependent, which makes them the rare fundamentals in this market that can actually be forecast.

Europe At 59% Against A 76% Norm

The international demand pull is real but it has not translated into US price support yet.

European storage stood at 59% as of August 9, well below the five-year average of 76% for that date. That 17-percentage-point gap represents a substantial rebuild requirement before the Northern Hemisphere heating season, and Europe has no domestic production to fill it.

The catch is timing. That deficit supports future US LNG demand once terminal maintenance ends, but it is not pulling enough gas out of the domestic system right now. European buyers cannot take cargoes that Freeport is not producing.

The Hormuz constraint amplifies the structural pull. Qatar exported over 112 billion cubic metres of LNG in 2025 as the world's second-largest supplier, and the entirety of that volume except Kuwait deliveries transits the strait. Transits have collapsed to 73 in the week ended August 16 from roughly 130 daily crossings before the conflict. Every cargo Qatar cannot ship is a cargo Europe and Asia have to source elsewhere, and the United States is the marginal supplier.

Soaring European natural gas prices have been flagged as an upside inflation risk for the euro area specifically because of Middle East supply shortages.

The arbitrage is open. US price spreads to Europe and Asia remain elevated due to ongoing disruptions, which is exactly the condition under which every available molecule should flow to export terminals.

The constraint is not economics. It is capacity. Even with Freeport fully operational, exports remain limited by slow growth in additional export infrastructure. Terminals cannot liquefy more than their nameplate allows regardless of how wide the spread runs.

That is the structural bull case for the winter and 2027 contracts, and it is why the forward curve holds a premium the front month cannot access. The summer contract sees the supply. The winter contract sees export demand eating into it month after month.

For the front month at $2.79, European storage at 59% is a reason the price is not lower rather than a reason it should be higher.

Heat Through September 3 Versus The Shoulder Season

The weather trade is what produced this week's bounce and it has a defined expiry.

Forecasts point to predominantly warmer-than-normal temperatures through September 3, which could boost gas consumption from power generators and provide further support to prices. That extension of the heat window was the direct catalyst for Wednesday's push through $2.800 — hotter overnight model runs combined with a production pullback strengthened the near-term setup.

Natural gas-fired electricity generation has been rising through 2026, up 2% in the first half against year-over-year declines in 2025 when gas prices were sharply higher. That switching dynamic works in the bulls' favour: cheaper gas displaces coal in the dispatch stack, which raises baseline burn independent of temperature.

The generation mix is moving against gas in other respects. Solar generation rose 21% in the first half and wind rose 6%, with continued renewable capacity additions expected to sustain that growth through 2027. Hydropower was up 9% in the first half but is forecast to decline 3% in the second half on intensifying drought across the western United States — which is marginally gas-supportive.

The problem is what comes after September 3. Updated models have indicated moderating temperatures across the Midwest and Northeast into late August, threatening to break late-summer cooling burns. Institutional positioning has been trimming long exposure as the seasonal window for heat-driven power demand closes.

Diminishing power-sector burn accelerates the transition into the low-demand autumn shoulder season well before winter heating begins. That is the calendar problem this market cannot solve with weather.

The seasonal arithmetic: roughly six weeks between the end of cooling demand and the start of heating demand, during which storage builds regardless of price, and during which 1.5 Bcf per day of new Permian takeaway arrives.

The counterweight is the short position. Speculator shorts have been building through the year, and a heavily short market can move fast on the next hot revision. The market is caught between short-covering risk and a supply problem, and the supply side has been winning every time the rally runs into the data.

Technicals: $2.945 Is The Line That Matters

The chart has flipped constructive on the short timeframe and remains capped on the medium one.

The main trend changed to up on Wednesday for the first time in weeks when the front month cleared $2.800 and ran to $2.875. The rally stalled just short of the 50-day moving average at $2.945, which is now the defining resistance for whether this becomes a genuine trend change.

The resistance structure below it is layered. The $2.80 to $2.83 band has produced systematic selling on every approach, and the Fibonacci level near $2.84 sits inside it. Taking out $2.84 with conviction opens room toward the 50-day average.

Support has been equally consistent. The $2.70 to $2.72 zone has caught every decline, with the 50% retracement at $2.723 acting as the pivot. A break below $2.723 exposes the recent swing lows and opens the downside again.

Momentum indicators read neutral rather than directional. MACD has been printing between 0.046 and 0.052 — effectively flat and offering no signal. RSI has held between 52.6 and 52.7, squarely in the neutral zone. Williams %R has swung from 78.774 on August 17, a sell reading, to 10.549 on August 19, an overbought reading, inside forty-eight hours. That oscillation captures how choppy this tape has been.

One framework reads the setup as a sell signal with average strength, noting that relative strength just crossed above 30% after bouncing from oversold territory and flagging the risk of mean reversion.

Another reads the daily buy/sell signal as a buy based on moving averages and indicators.

That disagreement is honest. The contract is trading in the middle of a range it has occupied since late July, with resistance stacked above and support that has held three times.

The structural read: natural gas has been in a downtrend since the $3.22 print on July 1, and every rally has failed at a lower high. Clearing $2.945 would be the first higher high in seven weeks and would break that sequence.

Until then, the trend favours sellers unless the storage data starts telling a different story.

The Levels: $2.84 Overhead, $2.723 And $2.65 Underneath

Immediate resistance. $2.80 is the round number and the base of the band that has capped every attempt. Above it, the Fibonacci level near $2.84 and the swing top from Wednesday at $2.875 mark the next tier.

Primary resistance. The 50-day moving average at $2.945 is the line that defines the trend on the medium timeframe. Clearing it opens $3.00 to $3.05 and would represent the first genuine breakout since June.

Above that. $3.15 to $3.34 was the range that held from mid-June through early July on strong supply fundamentals. Reclaiming it requires either a production rollover or a demand shock, and would put the front month back above the EIA's fourth-quarter forecast of $3.14.

First support. $2.723 is the 50% retracement and the pivot on the intraday framework. Holding it keeps the constructive structure alive.

Second support. $2.70 is the round number at the base of the zone that has caught every decline through August. Losing it exposes the lows.

Third support. $2.652 was the August 16 print, and $2.59 was the July 31 day-ahead spot low. That band is where the market went when production and feedgas were both working against it.

Structural floor. A break and sustained close below $2.50 would open the path toward $2.00, a level most frameworks view as unsustainable given LNG export dynamics but possible in a severe warm-winter scenario. Sub-$2 would eventually trigger producer curtailments.

The framing for the next fortnight: above $2.723 the bias is neutral-to-higher with $2.945 the gate. Below $2.70 the range breaks and $2.59 comes into play.

The calendar makes the resolution likely to arrive quickly. Freeport maintenance completes in late August. Hugh Brinson reaches full capacity September 1. Warmer-than-normal forecasts run through September 3. Two more storage prints land before then.

The next EIA number tells the market which side has the stronger case heading into the final stretch of summer.

Appalachia Is Already Talking About Curtailments

The supply-side response is starting, and it is the mechanism that eventually resolves an oversupplied market.

It appears increasingly likely that Appalachian producers could again curtail some natural gas production this fall as regional storage inventories build and forward prices slide. That is the classic sequence — regional basis blows out, producers cannot clear the price, and volumes get shut in until the calendar or the pipeline capacity improves.

The regional price data supports it. Appalachian day-ahead spot ran at a $0.75 discount to Henry Hub on July 1 and widened to $0.83 below the benchmark by July 31. The Northeast regional average swung from a $0.92 premium to Henry Hub at the start of July to a $0.22 discount by month end — a $1.14 reversal in thirty days.

That is not a marginal move. A producer in a basin trading nearly a dollar below a $2.79 benchmark is receiving under $2.00 for the molecule, which is below the economics for a meaningful share of dry gas acreage.

The forwards split during the August 13 to 19 window reinforced the pattern: western hubs added more than 25 cents while West Texas and Northeast points shed close to 20 cents.

Curtailment matters because it is the only supply response that works quickly. Rig count reductions take six to nine months to show in production. Shutting in existing wells removes volume within days.

The precedent exists. Appalachian producers curtailed in prior oversupply episodes, and the phrase "again" in the current reporting signals this is a repeat rather than a novel development.

What it does not fix is the Permian. Associated gas comes out of oil wells whether the gas price justifies it or not, and with WTI at $86.94 the drilling economics for crude are compelling. Hugh Brinson reaching full capacity on September 1 delivers more of exactly that gas to market.

So the supply response is regional and partial. Dry gas basins curtail; associated gas basins do not.

That asymmetry is why the surplus has proved so durable, and why every rally has been sold. Until oil weakens or takeaway fills, the Permian keeps supplying the marginal molecule regardless of what happens at Henry Hub.

The Winter Curve Is Where The Conviction Sits

The front month tells one story and the deferred contracts tell another, and the spread between them is the cleanest read on what this market actually believes.

Through the summer the August-to-February spread stayed wide even with front-month production running at highs. At one point the August contract traded near $3.20 while February held close to $3.95 — a 75-cent premium for the winter month.

The EIA's own quarterly path carries the same shape: $2.87 in the third quarter, $3.14 in the fourth, and $3.62 in the first quarter of 2027. That is a 75-cent step from the current quarter to the winter peak.

The summer contract sees the supply. The winter contract sees the export demand eating into it month after month. Both are correct.

The structural demand case for the deferred months is genuine. LNG exports are climbing, roughly 15% of US dry gas production is committed to overseas buyers, and the Hormuz disruption has pushed European and Asian buyers toward American cargoes with no near-term resolution visible. European storage at 59% against a 76% norm has to be rebuilt.

Data centre load is the other structural leg. Electricity generation growth is being driven primarily by increasing demand from large customers including data centres, and natural gas-fired generation is forecast to increase in 2027 as prices stay relatively low while coal generation continues to decline.

What caps the winter contract is the starting inventory. A record 3,985 Bcf at the end of October — 5% above the five-year average and the highest heading into winter since 2016 — provides an enormous buffer against a cold shock.

The Fern precedent cuts against complacency. Last January delivered the largest weekly net withdrawal in the history of the EIA's storage report, spot averaged $7.72 in January, and the February contract traded $4.875 against $3.120 a week earlier. The 12-month strip climbed 65 cents in a single report week.

Storage sets the buffer. Weather sets whether the buffer matters.

The mid-range of the longer-term channel sits near $4.00 per MMBtu, which represents the most likely destination heading into winter 2026-27. The upper boundary near $5.00 aligns with structural targets and the EIA's earlier 2027 forecast, now reachable only under a significantly colder-than-normal winter.

Natural Gas Price Forecast: Base, Bull And Bear Into Q4

Base case. The front month holds $2.65 to $2.95 through September while the market works through the shoulder season. This is the highest-probability path. Production at 113.5 Bcf per day against demand at 83.1 Bcf per day, a 198 Bcf surplus to the five-year average, and Hugh Brinson adding 1.5 Bcf per day on September 1 cap every rally. Warmer-than-normal forecasts through September 3 and Freeport returning 2.0 Bcf per day of export demand in late August put a floor under it. The EIA's $2.87 third-quarter average sits inside that range and is the reference to trade against. Watch the daily close relative to $2.723 as the cleanest read on control.

Bull case. A close above the 50-day moving average at $2.945 breaks the seven-week sequence of lower highs and opens $3.00 to $3.05, then the $3.15 to $3.34 band. That path requires three things: two consecutive storage injections below the five-year average, Freeport returning to full 17.4 Bcf per day feedgas rates without offsetting terminal outages, and heat extending materially past September 3. Add Appalachian curtailments actually materializing and the surplus starts closing rather than widening. The extended target sits near $4.00 into winter 2026-27, which is where the mid-range of the longer-term channel and the EIA's first-quarter 2027 forecast of $3.62 converge.

Bear case. Moderating temperatures across the Midwest and Northeast break the late-summer burn, Hugh Brinson delivers 1.5 Bcf per day into a demand trough, and injections revert toward the 29 Bcf five-year average. Losing $2.70 exposes $2.652 and then $2.59. A push toward the record 3,985 Bcf October target strips out any winter-scarcity premium and puts $2.50 in play. Below that, $2.00 becomes technically reachable in a severe warm-winter scenario, though it would trigger producer curtailments before it settled there.

What actually decides it. Three variables, in order. The Freeport restart against the Hugh Brinson startup — 2.0 Bcf per day of returning demand versus 1.5 Bcf per day of new supply, both landing within days of each other. The next two storage prints, because one 16 Bcf undershoot proves nothing and two changes the trajectory toward October. And the 50-day moving average at $2.945, which has capped this contract since July and which defines whether the August bounce is a trend change or the third failed attempt at the top of the same range.

At $2.79 the market is pricing an oversupplied summer with a tight winter behind it. The October storage number decides which of those two views is right.

That's TradingNEWS