NG Front Month ($2.766) Recovers From $2.682 After a Smaller-Than-Expected 28 Bcf Injection

Natural gas futures trade at $2.766 per million British thermal units, up from Wednesday's $2.722 close, after the weekly storage report came in tighter than expected. Today's range runs $2.681 to $2.776 — a nine-and-a-half-cent band that captures both the fresh three-month low and the bounce off it. Price briefly slipped below $2.70 earlier in the week, extending losses to the weakest level since early May.

The annual context is a collapse. The 52-week range spans $2.483 to $7.827, and the contract is down roughly 10.95% over twelve months. That upper bound reflects the January polar-vortex episode; the lower bound is where this market found a floor in the spring. Front-month gas is now trading in the bottom 8% of its own annual range with the injection season still running.

The July path shows how it got here. The contract broke down from a $3.250 swing high early in the month, shedding more than 30 cents in a handful of sessions. That gave way to weeks of choppy range trade between roughly $2.800 and $2.950. Price then rallied to a $2.989 high in the week of July 20, gapped sharply lower on July 24 from the $2.950 region to $2.836 — erasing nearly a month of consolidation gains in a single move — and sliced through the $2.800 shelf that had held on multiple tests since mid-July. The low printed at $2.682.

For most of the summer the market had been considerably more comfortable. Front-month futures traded between $3.15 and $3.34 from mid-June through late July on strong cooling demand, and the June contract changed hands above $3.20 with technicians watching a 200-day average at $3.627. Roughly fifty cents has come out of the front month since, and it has come out through the supply side rather than the demand side.

The setup into August is a market carrying a large storage surplus, record production, softening export pull and a cooler weather pattern, trading at a level where producer economics start to bite. Every one of those inputs is bearish except the last. Today's smaller-than-forecast injection was the first genuine data point in weeks that cut the other way, and it produced a four-cent bounce off a three-month low. That ratio — a meaningful bullish surprise generating a marginal move — tells you where the burden of proof currently sits.

The 28 Bcf Build: A Seven-Bcf Miss That Bought Four Cents

The weekly report covering the seven days ended July 24 showed a net injection of 28 billion cubic feet, against market expectations for a 35 Bcf build. That is a 20% miss to the tight side and the smallest weekly build of the injection season to date. Total working gas in storage rose to 3,084 Bcf.

The sequence through July shows a steadily decelerating build, which is the constructive thread running underneath an otherwise bearish tape. The week ended July 3 delivered 61 Bcf. July 10 brought 41 Bcf. July 17 produced 32 Bcf against the same 35 Bcf consensus, a smaller miss in the same direction. July 24 came in at 28 Bcf. Four consecutive weeks of shrinking injections, with two consecutive undershoots of expectations, is the signature of a market where summer demand is finally starting to bite into surplus supply.

Set that against the June run rate and the deceleration is stark. The week ended June 26 saw 87 Bcf injected. The week ended May 29 saw 95 Bcf. Builds have fallen by roughly two-thirds since late spring, which is seasonally normal in direction but faster than the five-year pattern in magnitude — the five-year average build for the July 17 week was 30 Bcf against an actual 32 Bcf, and the comparable week in 2025 delivered 27 Bcf.

The year-over-year comparison has also flipped in the bulls' favour, quietly. Stocks were 15 Bcf below the prior year in the week ended July 3, 21 Bcf below on July 10, then 16 Bcf below on July 17, and 32 Bcf below on July 24 — roughly 1% under last year's level. Storage is now running a genuine year-over-year deficit that has been widening for two weeks.

The problem is that neither of those trends is doing much for price, and the reason is the surplus against the five-year average rather than against last year. A market can build a year-over-year deficit and still be structurally oversupplied if the prior year was itself tight. That is exactly the position here, and it explains why a 20% bullish surprise on the headline injection number generated a bounce of four cents rather than fifteen.

185 Bcf Above Normal: The Surplus That Refuses to Shrink

Inventories stand 185 Bcf above the five-year average of 2,899 Bcf — a surplus of 6.4%. That percentage has been remarkably stable through the entire injection season, and its persistence is the single most bearish fact in this market.

The progression tells the story. At the end of May the surplus stood at 138 Bcf against a five-year average of 2,440 Bcf. By June 26 it was 175 Bcf against 2,747. By July 3, 185 Bcf against 2,798. By July 10, 181 Bcf against 2,843. By July 17, 183 Bcf against a base that had climbed further. And now 185 Bcf. The absolute surplus has widened by roughly 47 Bcf since the end of May, and the percentage above normal has held at 6.4% for three consecutive weeks despite injections running below both consensus and the five-year pace.

That combination — shrinking builds against a stable percentage surplus — is arithmetically consistent and strategically important. It means the market is injecting at roughly the five-year rate while carrying a cushion accumulated earlier in the season. Below-average builds narrow the surplus slowly; in-line builds hold it constant. The market needs sustained withdrawals or genuinely deficient injections to work the overhang off, and July has delivered neither.

The forward projection compounds it. The official outlook forecasts working inventories reaching 3,966 Bcf by the end of October, which would still be 5% above the five-year average heading into the withdrawal season. Storage was already 6% above normal at the end of June, so the model assumes only a modest narrowing across the remaining four months of injection. That is a market entering winter with a full tank.

The regional detail carries a caveat worth noting. Recent reports have shown Pacific storage withdrawing 5 Bcf and South Central salt withdrawing 7 Bcf in a single week, with South Central non-salt running 5.4% below last year. Salt-dome facilities cycle fastest and are the first to signal genuine regional tightness. A national surplus of 6.4% masks pockets where basis can move violently — which is why Gulf Coast, western and export-linked buyers cannot trade the headline number alone.

Production at 110.6 Bcf/d Matches the December Record

Supply is the reason the surplus will not clear. Lower 48 dry gas output has averaged 110.6 Bcf per day so far in July, up from 110.0 Bcf/d in June and matching the monthly record high set in December 2025. Daily readings have run higher still — one estimate put Thursday output at 111.7 Bcf/d, some 2.8% above the year-ago level.

The official production forecast was revised upward this month, to 111.2 Bcf/d for 2026 from 111.0 Bcf/d in the prior estimate. That is a market where the supply projection is being marked higher while the price is being marked lower, which is the definition of a structural rather than cyclical oversupply.

The basin detail explains where the barrels are coming from. Year-to-date production is up nearly 4%, with the gains concentrated in two liquids-adjacent plays and one dry-gas basin. Average output across the first seven months of 2026 is running approximately 10.8% higher in the Eagle Ford and 7.7% higher in the Haynesville than the same period last year. Permian production is up close to 6%, reaching an average of 23.7 Bcf per day. Month-to-date output through July 23 was 0.2% above the equivalent June period and 2.5% higher than a year earlier.

The Permian number is the one that matters most for price sensitivity, and it is the least responsive to gas economics. That gas is associated production — a byproduct of crude drilling, produced regardless of what Henry Hub does, because the oil economics drive the well decision. With Brent near $89 and West Texas Intermediate at $83, Permian oil activity has every incentive to continue, and 23.7 Bcf/d of gas comes out alongside it whether the price is $4 or $2.

The Haynesville gains are the more price-sensitive component, since that is a dry-gas play drilled specifically for the commodity and located close to the Gulf Coast export corridor. Producers there expanded into an expectation of export-driven demand growth that has arrived more slowly than the drilling did. That is the classic sequencing error in this industry, and it is why the current surplus exists despite genuinely strong structural demand growth.

127 Rigs and the Curtailment Threshold Nobody Has Hit

The rig count is the mechanism through which low prices are supposed to correct oversupply, and it is not responding. Active natural gas rigs stood at 127 in the week ended July 24, up by one. That is below February's three-year high of 134 but comfortably sufficient to keep output steady at record levels.

The disconnect between price and rig activity is the crux of the near-term problem. Front-month gas has fallen roughly fifty cents since mid-June, and the rig count has gone up. Producers are drilling into a three-month price low, which tells you either that their hedge books are protecting current economics, or that they are drilling for the winter strip rather than the prompt, or both. All three are true across different operators.

The threshold at which behaviour changes is identifiable and it sits below current price. Market analysis has consistently placed the practical floor for this cycle around $2.83 per MMBtu absent extreme warm weather, with the observation that at roughly $2.00 production curtailments and rig count reductions would emerge within weeks, tightening the market faster than seasonal trends suggest. Front-month at $2.766 is beneath the first of those markers and well above the second. That is the awkward middle: low enough to hurt, not low enough to trigger the supply response that would fix it.

One desk has framed the resolution as an autumn story rather than a summer one — lower prices eventually creating room for a recovery as producers pull back and utilities switch from coal to gas. Both mechanisms are real and both operate with a lag measured in months rather than weeks. Rig decisions made in August affect production in the fourth quarter. Coal-to-gas switching in the power stack responds within days but requires the price differential to hold long enough for dispatchers to reconfigure.

The practical read is that the supply side offers no near-term relief. The market has to work through the gas it already has, and there is a great deal of it. The correction mechanism exists, it is well understood, and it will not deliver inside the current quarter. Anyone modelling a supply-driven rally before October is modelling a lag that has not yet started running.

LNG Feedgas Softened at Exactly the Wrong Moment

Export demand has been the structural floor under this market for three years, and it wobbled in July. Flows to major export terminals have averaged 17.2 Bcf per day month to date, down from 17.4 Bcf/d in June. One weekly measurement put deliveries at 16.9 Bcf/d for the period ending July 22, a decline of 1.6% week over week — though still 6.8% above the same week last year.

The proximate cause is maintenance rather than demand destruction. Scheduled work at a major Texas liquefaction facility has taken capacity offline, with intermittent reductions at other terminals compounding the effect. Feedgas flows remain below the record levels reached earlier in 2026, and that gap is the difference between a market with a firm floor and one without.

The recovery has already started, which matters for the forward view. Deliveries were reported at 18.1 Bcf/d on Tuesday, up 2.4% from the prior week, and one daily estimate put net flows to export terminals as high as 19.3 Bcf/d, an increase of 1.5% week over week. Expectations earlier in the season had feedgas climbing from around 18 Bcf/d in late May toward record levels north of 20 Bcf/d as new capacity ramped. Exports to Mexico have also been running strong at roughly 7.3 Bcf/d.

The arithmetic of why this matters: a swing from 17.2 to 20 Bcf/d is 2.8 Bcf/d of incremental demand, which over a four-week period is roughly 78 Bcf — meaningful against a 185 Bcf surplus. Export capacity is the only demand category capable of moving the storage balance at that scale inside a quarter, because weather-driven power burn is already near its seasonal peak and industrial demand is close to inelastic.

That makes the feedgas number the single most important daily data point for anyone trading this contract. When flows return durably above 19 Bcf/d, the injection pace should decelerate further and the surplus should finally start compressing. Until then, a market carrying record production and a 6.4% storage overhang has lost its most reliable source of incremental pull at precisely the point in the season when it needed it most. The maintenance is temporary. The timing was not helpful.

The Weather Turned and Took the Last Bullish Input With It

The final leg of July's decline was meteorological. Updated forecasts calling for cooler temperatures across the central and eastern United States in the coming weeks reduced expected air-conditioning demand and removed the one variable that had been supporting price through the first half of the month.

The contrast with early July is sharp. Demand for electric power generation averaged 45.6 Bcf per day in the week ending July 7, more than 15% higher than the preceding week, as heat built over the holiday weekend. Total demand including exports was running above year-ago levels. That heat is what kept front-month futures inside the $3.15 to $3.34 band from mid-June through most of July despite the storage surplus. When the forecast models shifted, that support disappeared inside a week.

The specific shift matters more than the direction. Forecasts moved toward normal seasonal weather across the eastern two-thirds of the country, which is not a cold pattern — it is simply the absence of an anomalously hot one. In a market this oversupplied, normal weather is bearish, because the surplus was accumulated under normal conditions and only above-normal cooling demand can work it off inside the injection season.

Lower 48 gas demand of roughly 79.1 Bcf/d in a recent session, up 4.7% year over year, confirms that consumption is genuinely growing. It is growing more slowly than production, which is up 2.8% year over year off a much larger base. That is the entire imbalance in two numbers: demand growth of 4.7% on 79 Bcf/d against supply growth of 2.8% on 111 Bcf/d produces roughly 3.7 Bcf/d of new demand against 3.0 Bcf/d of new supply — nearly balanced, which is not enough to clear an existing 185 Bcf overhang.

August weather is therefore the dominant near-term swing factor and it is unforecastable beyond two weeks. A genuine heat dome across the population centres of the Midwest and Northeast would flip injections to withdrawals and compress the surplus fast. Normal August weather leaves the market grinding toward the $2.65 measured-move target that technicians have identified. The asymmetry favours the downside simply because the surplus is the starting condition and normal weather does not remove it.

The Data Centre Demand Story Is Real and It Is Slipping Right

The strongest structural argument for owning natural gas is electrification, and the official projections are unambiguous about its scale. Monthly consumption is forecast to reach 50.6 Bcf/d in July 2027, which would be the highest of any month on record. Gas-fired generating capacity is projected to reach 508 gigawatts by the end of 2027, up 3% from 2025 levels. That capacity build is driven by rising overall electricity demand, additions to the gas generation fleet, and gas remaining the marginal dispatch fuel across most hours.

The demand driver behind those numbers is increasingly artificial intelligence infrastructure. Data centre campuses require firm, dispatchable power at scale, and gas turbines are the only technology that can be permitted, built and commissioned on the timeline hyperscalers are working to. That is why gas is the second-order beneficiary of the same capital expenditure cycle currently reshaping the semiconductor complex.

The timing risk is the problem, and this week produced a concrete example. Federal regulators moved a New Mexico gas lateral — a project intended to supply a major data centre campus being developed for two of the largest names in artificial intelligence — out of fast-track review, potentially delaying a decision until December. That is one pipeline on one campus, but it is representative. The demand exists on paper long before the molecules can physically reach it, and the constraint is permitting and pipe rather than production.

The consequence for price is a widening gap between the forward narrative and the prompt reality. Gas producers drilled into an expectation of export and power-sector demand growth that is genuine and is arriving late. Every quarter of delay in infrastructure commissioning is a quarter in which record production meets demand that has not yet connected, and the surplus persists.

That is the honest framing of the bull case: correct on direction, wrong on timing, and the timing error is currently being expressed as a 6.4% storage overhang and a $2.766 front month. Investors positioning for the electrification thesis should be doing so through the winter strip and 2027 contracts rather than the prompt, because the prompt is trading a physical balance that the thesis does not affect for another twelve to eighteen months.

The Chart: $2.682 Floor, $2.799 First Fib, $2.872 the Ceiling

The technical structure is cleanly bearish with a well-defined bounce zone. A descending trendline connects the early-July swing high near $3.30 to the more recent consolidation highs around $2.99, and price has not produced a decisive break above that ceiling at any point this month. The commodity broke down sharply from that area late in July, tumbling to $2.682 before staging the modest recovery to current levels.

The Fibonacci grid drawn from the $2.989 swing high to the $2.682 low gives the retracement map. The 38.2% level sits at $2.799, the 50% at $2.836, and the 61.8% at $2.872. That last level lines up with the descending trendline and should act as a strong ceiling on any corrective bounce. The $2.836 to $2.872 band is also the broken consolidation floor from earlier in the month — a zone that previously acted as support and has now flipped to resistance, giving sellers a natural place to reload.

The moving average configuration confirms the bias. The 100-period simple average remains below the 200-period, both are sloping lower, and they are converging just above current price action. Price trades beneath both, which reinforces them as dynamic resistance on any recovery attempt. Earlier in the month the two averages had begun flattening and converging in a way that hinted at a base; the late-July breakdown reversed that improvement.

Momentum readings describe a market that is stretched but not exhausted. The stochastic oscillator dipped into oversold territory during the selloff, reflecting how extended the recent selling became, and any upturn from there could fuel a near-term bounce toward the Fibonacci confluence. The relative strength index has been hovering in the mid-40s with room to fall further, which suggests sellers retain the upper hand and could resume control once a corrective pullback runs its course.

The trade is mechanical. The $2.800 shelf that held on multiple tests since mid-July has broken, and the height of the prior range projects a measured move toward $2.650 or lower. Sell rallies into $2.836 to $2.872 with a stop above the descending trendline. Losing $2.682 targets $2.650 and then the 52-week low at $2.483. Only a daily close above $2.872 that clears both the trendline and the converging averages changes the structure, and nothing in the current fundamental set supports that.

Contango: The Market Is Pricing a Winter It Does Not Yet Own

The forward curve is where this market's genuine tension lives. The front month sits at $2.766 while the December contract has been trading above $4.00 per MMBtu — a spread of well over a dollar, or roughly 45%, between the prompt and the heating season. That is steep contango, and it is the curve telling you that the physical surplus is a summer condition rather than a structural one.

Regional forward markets are more emphatic still. Winter forwards at the Mid-Atlantic hub have been described as running hot, and Southern California continues to command a substantial premium over benchmark pricing, with local structural constraints suggesting cheap gas into that market may be a thing of the past. Those are basis markets pricing delivery constraints rather than commodity scarcity, and they diverge from Henry Hub in exactly the conditions where pipeline capacity binds.

The economics of that contango are important for anyone considering the exchange-traded vehicles. A curve in steep contango imposes a negative roll yield on any long position that must sell the expiring contract and buy the more expensive next month. Over a summer of this shape, the fund tracking the front month can lose meaningful value even if spot price is flat. That is the structural headwind embedded in the most commonly used retail expression of a long gas view, and it is why holding those instruments through a contango season has historically been expensive.

The constructive read on the curve is that it validates the structural thesis while pricing the cyclical problem. The market believes in export growth, power-sector demand and winter tightness. It simply refuses to pay for them in August, because in August there is 3,084 Bcf sitting in the ground and 110.6 Bcf/d coming out of it.

For a trader, that argues strongly for expressing any bullish view through the winter contracts or calendar spreads rather than the prompt. Buying December at $4.00 requires a cold winter to pay. Buying the front month at $2.766 requires the surplus to clear inside eight weeks, which nothing in the supply, weather or export data currently supports. Those are different trades with different odds, and conflating them is the most common error in this market.

The Official Forecast at $3.70 Against a Market at $2.766

The gap between the published outlook and the traded price is roughly a dollar, and it deserves scrutiny because it is unusually wide. The official short-term outlook, completed on July 1 and released July 7, has the benchmark spot price averaging close to $3.70 per MMBtu across 2026 before declining below $3.50 in 2027. A separate passage in the same document puts the average closer to $3.60 across both years — about 10% below the 2016-to-2025 average in inflation-adjusted terms.

For a full-year average of $3.70 to hold with the front month at $2.766 in late July, the remaining five months must average materially above $4.00. That requires a substantial fourth-quarter rally, which is precisely what the forward curve prices and what the fundamental setup does not currently support. The forecast is not implausible — it is simply front-loaded on months that have already printed at higher levels and back-loaded on a winter that has not happened.

The same outlook is explicit that inventories remaining above the five-year average through much of the forecast period will limit upward price pressure, and that inventories stay high because record production led by Permian growth is meeting rising demand. That is the bearish half of the official view, and it is the half currently being validated by the tape.

Independent modelling clusters lower and closer to the market. One framework assigns a 55% base case of $2.50 to $4.50, with third-quarter prices near $2.80 to $3.00 before firming toward $4.00 to $4.50 in the fourth quarter as export demand peaks and heating demand returns — landing the full-year average near $3.50. A 25% cold-winter scenario carries $5.00 to $8.00, with a polar vortex repeat capable of revisiting the $7-plus range seen in January 2026. A drop below $2.50 in the second half is considered possible under a warm-weather scenario but unsustainable given the structural export floor.

That last judgment is the one to internalise. The 52-week low at $2.483 is not a random number — it is roughly where export economics and producer breakevens converge to stop the decline. The market is currently 11% above it with no supply response yet visible.

Regional Basis and the Curtailment Floor Beneath $2.50

The most useful way to frame downside from here is through producer economics rather than chart levels. Front-month gas at $2.766 sits below the practical floor most analysis places around $2.83 for this cycle, meaning the marginal producer is already operating at or beneath the threshold where new drilling stops making sense on prompt economics alone. That has not yet translated into rig reductions because hedge books and the $4.00 winter strip are carrying the decision.

The genuine curtailment trigger sits considerably lower. At roughly $2.00 per MMBtu, shut-ins and rig cuts would emerge within weeks and would tighten the market faster than the seasonal pattern implies. Between $2.50 and $2.80 the response is gradual and expressed through completion deferrals rather than outright curtailment — operators drill the well and leave it uncompleted, waiting for the winter strip. That behaviour builds an inventory of deferred production that caps any subsequent rally, because the first sustained move above $3.50 brings it online.

The coal-to-gas switching channel provides the other floor and it operates faster. When delivered gas falls far enough relative to coal, dispatchers reorder the stack and gas-fired units run more hours. That mechanism is already partly engaged and adds incremental power-sector burn without requiring any weather event. Wholesale electricity prices are forecast to average about $45 per megawatt hour nationally this summer, lower than last summer specifically because of cheaper gas delivered to power plants — which is the switching mechanism visible in the output data.

Regional dispersion is where the actual money is being made this season. South Central non-salt storage running 5.4% below last year, Pacific withdrawals, a persistent Southern California premium and firm Mid-Atlantic winter forwards all point to a market where the national number understates local tightness. Basis traders have had a considerably better summer than flat-price traders.

The synthesis for a flat-price view: downside is real but bounded, with $2.650 the measured-move target, $2.483 the annual low, and genuine supply response arriving well before $2.00. Upside requires either August heat or an export surge above 19 Bcf/d, and neither is currently in the data. That is a market with an asymmetric but shallow downside and a deferred upside — which is precisely the configuration in which the forward curve, not the prompt, is the correct instrument.

The Forecast: $2.90 Base, $3.60 Bull, $2.48 Bear Into the Fourth Quarter

The base case, at roughly 50% probability, is a range between $2.65 and $3.00 through August and September, resolving toward $2.90 as the injection season winds down and the market begins pricing winter risk. This requires normal weather, production holding near 110.6 Bcf/d, and export flows recovering toward the 18 to 19 Bcf/d range as maintenance completes. Under this path the front month works back toward the $2.836 to $2.872 Fibonacci and trendline confluence, fails there at least once, and only clears it in September when the curve rolls into contracts with heating-season exposure. Trade the range and respect the descending trendline.

The bull case, around 25%, requires August heat plus an export surge. A genuine heat dome over the eastern population centres would flip weekly injections toward zero and start compressing the 185 Bcf surplus quickly, and feedgas moving durably above 20 Bcf/d would add roughly 3 Bcf/d of pull that the storage balance cannot absorb. That combination clears $2.872, then $2.989, and opens the path toward $3.30 and eventually the 200-day area near $3.60. Note that even this scenario leaves the front month below the official $3.70 full-year average, which tells you how much work that forecast requires from the fourth quarter.

The bear case, around 25%, is normal-to-cool August weather with production drifting higher. Losing $2.682 opens the $2.650 measured-move target, and beneath that the 52-week low at $2.483 becomes the reference. That is roughly 10% below spot and it is where producer economics and export arbitrage should stop the decline. A break beneath $2.48 would require a demand shock rather than a weather pattern, and nothing in the current data supports one.

The disciplined posture at $2.766 is to avoid the prompt entirely and express any constructive view through the winter contracts. December above $4.00 is where the structural story — record power-sector consumption forecast for 2027, 508 gigawatts of gas-fired capacity, export capacity still ramping — actually gets paid. The front month is trading a physical surplus that the structural story does not touch for another year. Do not confuse the two.

That's TradingNEWS