Natural Gas Holds $2.766 With Production at a Record 110.6 Bcfd — EU Gas Runs 68.62% Higher Year Over Year as December Trades Above $4

Natural gas futures rose early Monday as a hot August pattern squared off against comfortable supplies and soft LNG demand, with the front contract trading near $2.766 per MMBtu after basing at $2.682. The September NYMEX contract closed at $2.72 last Wednesday, added a nickel to $2.77 by Thursday, and averaged $2.74 across the week against $2.88 the week prior. That five-day average decline of 14 cents describes a market losing altitude even on the sessions it manages to close green.

The breakdown that produced this level was clean and violent. Gas broke sharply from the late-July swing high near $3.00, slicing through a rising trend line and its short-term moving averages to confirm resumption of the broader downtrend. Price fell below $2.70 at one point, the lowest level in three months, weighed down by strong production, comfortable inventory levels, and weak LNG feedgas demand. The selloff dragged the contract to $2.682 before buyers stepped in.

The recovery since has the shape of a corrective bounce rather than a reversal. Gas has been carving higher lows in a shallow grind, and price is now testing a descending trend line drawn off the July highs that lines up with the 61.8% Fibonacci retracement at $2.872. That confluence acts as a hard ceiling if the move down is simply pausing before another leg. Below it, the 50% retracement sits at $2.836 and the 38.2% level at $2.799, both directly overhead from spot.

The wider tape delivered a cross-current rather than a push. President Trump called off planned strikes against Iran and announced negotiations opening Monday afternoon, collapsing West Texas Intermediate 6.21% to $79.41 and Brent 5.11% to $83.24. That headline pressured European gas hard and did almost nothing to American gas, which is the defining feature of this market right now.

Weather has been the swing factor in both directions. Updated forecasts calling for cooler temperatures across the central and eastern US weighed on prices through late July by reducing expected air-conditioning demand, and that shift is what carried the contract from $3.00 down to $2.682. Monday reversed part of it as the August heat pattern reasserted itself in the models.

The immediate question is whether $2.872 caps the bounce or gets taken out. Every other variable in this market resolves through that level.

July Delivered a 12.5% Monthly Decline

The July arc was one of steady erosion punctuated by a single sharp break. Gas began the month near $3.256, printed a high at $3.355, fell to a low of $2.688, averaged roughly $3.037 across the period, and finished around $2.850. That is a 12.5% monthly decline in the middle of cooling season, when the balance is supposed to be at its tightest.

The year-over-year comparison is where the deterioration becomes structural rather than seasonal. Henry Hub spot averaged $2.77 per MMBtu in April 2026, $2.94 in May, and $3.14 in June. The same months in 2025 delivered $3.42, $3.12, and $3.02. April was 19% cheaper this year, May was 5.8% cheaper, and June flipped positive by 4.0%. That June crossover was the only month of the second quarter where 2026 gas commanded a premium, and July gave it back immediately.

The path here also inverted the seasonal norm. Gas typically firms from May through August as cooling load builds and power burn peaks. This year the front month peaked at $3.355 in the opening days of July and spent the remaining four weeks working lower, which reflects supply overwhelming demand rather than demand failing to arrive.

Forward modeling puts August between a $2.430 low and a $3.029 high with an average near $2.781 and an end-of-month level around $2.814, a further 1.3% decline. September projects $2.646 to $2.924 with an average of $2.792 and a month-end print near $2.785, down another 1.0%. October is the first month where the models turn, projecting $2.785 to $3.106 with an average of $2.909 and a 6.2% gain into the heating-season transition.

Three consecutive months of modeled decline followed by a sharp October turn is the shape of a market waiting for winter rather than trading its own fundamentals. The 2026 experience so far has been a moderate range with Henry Hub oscillating between roughly $2.70 and $3.35, punctuated by a January polar vortex event that drove the front month above $7 and reminded everyone what the tail risk looks like on a market this thinly buffered.

The EIA's July outlook had Henry Hub averaging close to $3.60 per MMBtu for full-year 2026, which requires a substantially higher fourth quarter to hit.

The Divergence With Europe Is the Trade of the Summer

European gas fell to 57.81 EUR/MWh on August 3, down 2.13% on the day, after Trump's announcement provided relief following days of escalating tensions that had driven energy prices sharply higher. Traders stayed on edge after an explosion was reported near a tanker off the coast of Oman, which is the kind of headline that has repeatedly reversed the continent's price direction inside a single session.

The context around that 2.13% decline is what matters. EU gas has risen 30.38% over the past month and sits 68.62% higher than the same time last year. US Henry Hub futures fell 12.5% over the same month and now trade below year-ago levels for the equivalent period. Two benchmarks for the same molecule moving in opposite directions by roughly 43 percentage points across four weeks is the cleanest expression of what the Middle East conflict has done to global energy.

The explanation is straightforward on the surface. Europe imports its gas and has been paying a war premium since the conflict began in late February. The United States produces more of it than at any point in history and has been paying a glut discount. Bank of England Governor Andrew Bailey specifically flagged lower than usual European gas stock levels as an upside inflation risk when the MPC held rates on July 30, which is the policy-side confirmation that the continent's supply position is genuinely stressed.

The transmission channel between the two markets runs through LNG, and it is not working the way the arbitrage would suggest. US Gulf Coast feedgas demand has been struggling for weeks, a trend occurring simultaneously in Canada and Mexico. Flows to major export terminals averaged 17.2 bcfd in July, down from 17.4 bcfd in June, partly because of scheduled maintenance at Freeport LNG's Texas facility.

That is the failure point. European prices running 68.62% above year-ago levels should be pulling every available cargo across the Atlantic and tightening the American balance mechanically. Terminal maintenance and liquefaction throughput constraints mean the price signal cannot convert into physical flow fast enough to matter this summer.

The structural argument is that the growing LNG export base steadily links US domestic gas to global demand and places a firmer floor under Henry Hub than existed two years ago. A floor is not the same thing as a removed ceiling. Until liquefaction capacity clears the bottleneck, European scarcity and American abundance coexist on the same planet.

Production at 110.6 Bcfd Matches the All-Time Record

Average gas output in the US Lower 48 states rose to 110.6 bcfd so far in July from 110.0 bcfd in June, matching the monthly record high set in December 2025. Producers are pumping at the fastest rate ever recorded during a month when seasonal demand should be at its peak, and that single figure explains most of the price action.

Adding roughly 0.6 bcfd of month-over-month supply into a cooling season that delivered mixed weather kept the market oversupplied through the period when it should have been tightest. The increase in production has added directly to concerns over an oversupplied market, with inventories remaining 6.4% above the five-year seasonal average as of July 17 and projected to rise to 6.6% above normal for the week ending July 24.

The Permian is the driver, and the mechanism there is the one that gas bulls cannot argue their way around. ExxonMobil chief executive Darren Woods said Friday that Permian Basin natural gas takeaway is reaching the point where the market can clear, potentially unlocking more oil production and the associated gas that comes with it. Associated gas is the most price-insensitive supply on the continent because it arrives as a byproduct of oil economics rather than as a response to gas economics. When Permian takeaway constraints ease, molecules show up regardless of what Henry Hub is paying.

The EIA attributes above-average inventories directly to record natural gas production, led by growth in the Permian region, helping meet rising demand. That is the agency putting the surplus squarely on the supply side rather than on weak consumption.

The rig response has not turned and there is no reason to expect it to. With the front month at $2.77 and the December contract trading above $4, producers have every incentive to keep drilling into the winter curve rather than curtail into summer weakness. That behavior is rational at the individual company level and self-defeating at the market level, which is the recurring dynamic that has capped natural gas rallies for a decade.

Not every basin is participating. Canada's largest gas producer, Tourmaline, is scaling back upstream activity as Western Canadian prices remain suppressed. That is the supply discipline the US shale complex has declined to apply.

Storage at 3,084 Bcf and 185 Above the Five-Year Average

The EIA reported a net injection of 28 Bcf for the week ending July 24, lifting Lower 48 working gas to 3,084 Bcf. Stocks sit 32 Bcf below the same week last year and 185 Bcf above the five-year average of 2,899 Bcf, which puts inventories 6.4% above normal and 1% under the year-ago level. Total working gas remains within the five-year historical range.

The build was a bullish surprise against expectations and the market barely blinked. Consensus ran 33 to 38 Bcf with survey estimates clustering at 34 to 35 Bcf, so the 28 Bcf print landed well below the range and extended a trend of injections coming in under forecast. The prior week delivered 32 Bcf against a 35 Bcf expectation, and the week before that 41 Bcf against 43 Bcf expected.

Prices edged lower on Friday as traders looked beyond Thursday's smaller-than-expected injection and refocused on elevated production, softer LNG export demand, and a mixed weather outlook. That reaction function is the market stating its priorities directly: supply and exports matter more than the weekly storage headline.

Inventories were expected to rise to 6.6% above normal for the July 24 week, so the actual 6.4% reading came in marginally tighter than modeled. Against the July 17 comparison, the five-year surplus expanded by 2 Bcf to 185 Bcf while the year-over-year deficit widened from 16 Bcf to 32 Bcf. Both moves are small enough to be noise, which is itself the point: nothing in three consecutive below-consensus prints has changed the structural picture.

The storage position also has to be read against what the market is carrying forward. The EIA forecasts working gas inventories reaching 3,966 Bcf by the end of October, 5% above the five-year average. Entering a heating season 5% long, after a summer that ran 6.4% long, means the winter balance starts with a cushion the market has not had in every recent year.

The next weekly report releases Thursday August 6 at 10:30 a.m. ET. It covers the week ending July 31 and it is the first print that captures the early-August heat pattern that lifted prices Monday.

The Injection Cadence Collapsed and the Surplus Widened Anyway

The weekly build sequence through July is the strongest argument the bulls have, and it has produced nothing. The week ending July 3 delivered a 61 Bcf injection. July 10 came in at 41 Bcf. July 17 at 32 Bcf. July 24 at 28 Bcf. That is a 54% reduction in weekly build across three weeks, and the July 10 print was already the lowest increase since the first week of June, when the building season peaked at 108 Bcf.

Every one of those prints landed at or below consensus. July 10 came in at 41 against 43 expected. July 17 at 32 against 35. July 24 at 28 against a 33 to 38 range with surveys at 34 to 35. Three consecutive misses to the downside, each one larger than the last relative to expectations.

The surplus expanded through all of it. The five-year gap ran 181 Bcf on July 10, widened to 183 Bcf on July 17, and reached 185 Bcf on July 24. The market injected less gas than forecast for three straight weeks and finished the month further above normal than it started.

The arithmetic behind that apparent contradiction is simple. The five-year average builds for those weeks are themselves smaller than what the current market is producing. The July 17 week carried a five-year average build of 30 Bcf against an actual 32 Bcf, and the year-ago comparison was 27 Bcf. Coming in below consensus is not the same thing as coming in below the seasonal norm, and the market has repeatedly conflated the two.

The year-over-year deficit has been narrowing and widening in a tight band. Stocks ran 15 Bcf below last year on July 3, 21 Bcf below on July 10, 16 Bcf below on July 17, and 32 Bcf below on July 24. A 32 Bcf gap against a 3,084 Bcf base is 1.0%, which is a rounding error in a market that measures its inventory in trillions of cubic feet.

Higher levels of LNG exports have been offsetting record domestic production on the demand side, which is what kept the year-over-year position roughly flat despite the supply record. That offset weakened in July as feedgas slipped.

Three of Five Storage Regions Withdrew Gas Last Week

The aggregate storage number conceals a regional picture far more informative than the headline. The East region held 654 Bcf after a 23 Bcf build. The Midwest reported 789 Bcf, also up 23 Bcf. The Mountain region fell 2 Bcf to 238 Bcf. The Pacific dropped 7 Bcf to 307 Bcf. The South Central declined 9 Bcf to 1,096 Bcf, with salt storage falling 14 Bcf to 303 Bcf and nonsalt rising 5 Bcf to 793 Bcf.

Three of five regions withdrew gas during an injection week. That is the genuinely bullish detail underneath a bearish headline, and it reflects extreme cooling demand across the West and Gulf Coast pulling molecules out of storage faster than they could be replaced. Storage levels slipped across the West as heat gripped major markets and sent demand upward to meet cooling needs.

The South Central salt draw of 14 Bcf is the line to watch every Thursday. Salt dome storage across Louisiana and Texas is the highest-deliverability, fastest-cycling inventory in the country and it serves the Gulf Coast liquefaction complex directly. Pulling 14 Bcf from salt during July, with feedgas demand at 17.2 bcfd and softening, indicates power burn is doing that work rather than exports.

Year-over-year positioning differs sharply by region. The East sits 0.8% above last year, the Midwest 3.5% above, the Pacific 2.0% above, the Mountain region 1.7% below, and the South Central 5.7% below. Against the five-year average the ranking inverts entirely: Mountain runs 17.2% above, Pacific 16.3% above, Midwest 6.6% above, East 3.6% above, and South Central just 3.1% above.

The South Central being tightest against both benchmarks matters because it is the region that supplies liquefaction and the region where any winter demand shock lands first. The national 6.4% surplus does not exist in the place where the export demand sits, and South Central nonsalt remained 5.4% below last year as of the July 17 report.

Gulf Coast, western, and LNG-linked buyers need regional basis, heat, production, feedgas, and power-sector demand in every procurement decision. A national surplus concentrated in the Mountain and Pacific regions, at 17.2% and 16.3% above normal, is the least useful gas in the country when Henry Hub tightens.

LNG Feedgas Slipped and the Export Bottleneck Is Real

Estimated flows to major US export terminals averaged 17.2 bcfd in July, down from 17.4 bcfd in June. The decline traces partly to scheduled maintenance at Freeport LNG's Texas facility, and struggling feedgas demand on the US Gulf Coast has captured headlines across recent weeks. The same softness is occurring simultaneously in Canada and Mexico, which rules out a purely US explanation.

A 0.2 bcfd month-over-month decline sounds small against 110.6 bcfd of production. It is not. LNG feedgas is the marginal demand source that determines whether the domestic balance tightens or loosens, because power burn is weather-driven and residential-commercial demand is dormant in July. Every incremental bcfd of feedgas removes roughly 7 Bcf per week from the storage build.

The structural trajectory runs the other direction, which is the part the front month is not pricing. Record feedgas flows above 16.5 bcfd are expected from new projects including Plaquemines and Golden Pass, and those volumes combine with AI and data center power burns adding 2 to 3 bcfd cumulatively to produce storage deficits re-emerging over winter 2026-27.

Strong LNG feedgas activity reinforces the position that export demand is now a structural pillar of the US gas balance, and that each incremental capacity upgrade on the Gulf Coast hardens the floor under winter pricing. The forward capacity additions are contracted and under construction rather than speculative.

Cheniere anchors the complex with combined production capacity of approximately 55 mtpa across the Sabine Pass liquefaction facility in Cameron Parish, Louisiana and the Corpus Christi facility in Texas. The company has more than 6 mtpa of expected capacity currently under construction and is pursuing further liquefaction expansion. It reports second-quarter results Thursday August 6 before the open with a conference call at 11:00 a.m. ET, and consensus looks for $2.80 per share against a significant year-over-year decline.

The seasonal pattern for feedgas favors recovery from here. Maintenance windows cluster in the shoulder and summer months precisely because winter demand cannot tolerate downtime. Freeport returning to full rates lifts the national feedgas number back above 17.4 bcfd without any new capacity entering service.

That recovery is the single highest-probability bullish catalyst available before winter.

Power Burn Is the Demand Engine and It Is Growing

Power sector consumption is the mechanism converting weather into price, and the trajectory is one-directional. The EIA forecasts US natural gas consumption in the electric power sector increasing in both 2026 and 2027, reaching a record next year. Average sector consumption rises 2% in 2026 and another 4% in 2027 to 38.1 bcfd. On a monthly basis, consumption is projected to hit 50.6 bcfd in July 2027, which would be the most in any month on record.

That growth is driven by rising overall electricity demand, additions to the natural gas generation fleet, and relatively low gas prices. The last item is the self-correcting mechanism: cheap gas increases the coal-to-gas switching economics in dispatch, which increases burn, which has repeatedly established floors in the $2.50 to $2.70 zone.

The data center story is no longer speculative on the demand side. Chevron expects its 20-year power deal with Microsoft to unlock a broader natural gas-fired power business as artificial intelligence data center demand outpaces grid expansion. A 20-year offtake commitment from a hyperscaler to a major integrated is the kind of contract that changes long-dated gas demand assumptions rather than quarterly ones.

The weather setup for August is the near-term variable. Updated forecasts calling for cooler temperatures across the central and eastern US weighed on prices through late July by reducing expected air-conditioning demand, and that shift produced the break from $3.00 to $2.682. Monday reversed part of it as a hot August pattern reasserted in the models and squared off against comfortable supplies and soft LNG demand.

Near-term summer prices are expected to stay subdued below $3 per MMBtu as injection-season builds offset LNG demand. A further drop below $2.50 in the second half of 2026 remains possible in a warm-weather scenario but is regarded as unsustainable given the structural export floor.

The honest framing is that August heat can lift gas from $2.77 toward $3.00 and cannot lift it to $4. The key catalyst for a meaningful rally is winter weather rather than summer cooling load, and the front month has been trading exactly that way since the July high.

The Curve Is Pricing a Winter the Spot Market Refuses

The December 2026 futures contract already trades above $4 per MMBtu against a September contract at $2.77. That spread exceeds $1.23, or 44%, across three months, and it is the single most important structural fact in this market.

The steepness creates a self-reinforcing loop that caps the front. Producers hedge the December strip above $4 while spot sits at $2.77, which funds continued drilling and guarantees the supply that keeps spot suppressed. The contango is simultaneously the market's forecast of winter tightness and the mechanism preventing that tightness from arriving.

Storage economics work the same way. A $1.23 spread between September and December covers carry costs several times over, which incentivizes injection into every available cavern regardless of the spot price. That is why builds continue at a pace holding the surplus at 185 Bcf even as weekly injections shrink from 61 Bcf to 28 Bcf.

Henry Hub is expected to hold roughly $2.80 to $3.00 through the remainder of summer 2026 before firming into the fourth quarter as heating season approaches and feedgas demand peaks. The timing and magnitude of that recovery depend almost entirely on weather. A cold fourth quarter drives prices toward $4 to $5 in the base case. A polar vortex repeat revisits the $7 range printed in January 2026.

The official view is more restrained. The EIA expects the Henry Hub spot price in the fourth quarter of 2026 to average $3.57 per MMBtu, which is 5% less than the same quarter last year, with above-average inventories heading into winter explicitly cited as the reason for limiting upward price pressure. That $3.57 estimate sits roughly 50 cents below where the December contract is trading, which is the agency and the futures market disagreeing by 12%.

The five-year outlook points modestly lower, shaped by expanding LNG export facilities, fluctuating demand, and evolving supply dynamics. Against that, storage deficits re-emerging in winter 2026-27 on record feedgas and data center burn is the counter-thesis, and both cannot be right.

The next Short-Term Energy Outlook releases August 11.

Producers Are Generating Cash at $2.77

EQT reported second-quarter revenue of $1.81 billion against a $1.814 billion estimate, with non-GAAP earnings of $0.39 per share lagging the $0.418 consensus by 6.7%. Revenue grew 13.2% year over year. Net income attributable to shareholders fell to $211 million from $784 million a year earlier, driven mainly by lower derivative gains and realized prices.

The operating performance was stronger than the headline. Sales volume reached 634 Bcfe, exceeding guidance on strong well productivity and system optimization, and the company drilled the longest shale lateral in its history. Free cash flow ran $330 million for the quarter and $2.16 billion year to date, with capital expenditures 9% below guidance and operating expenses at the low end of the range. Strategic moves included a 10-year gas supply deal with CPV, a five-year LNG offtake agreement, and the $77 million Blackline Midstream acquisition expanding propane storage and vertical integration.

The stock treated it as a non-event, gaining 0.4% after hours, and has declined roughly 3.9% over the past month on sector pressure rather than company issues. EQT entered 2026 largely unhedged, which paid off during the strong first quarter and worked against it through the summer.

Expand Energy delivered adjusted earnings of $1.33 per share against a $1.22 consensus, up from $1.10 a year earlier on strong production and lower operating expenses. Natural gas, oil, and NGL revenues of $1.8 billion missed the $2 billion estimate and fell below the year-ago figure, with total revenue at $2.96 billion. Production averaged 7,482 MMcfe per day at 92% gas, up 3.9% from 7,202 MMcfe per day. Total operating expenses fell to $2.3 billion from $2.4 billion, with marketing down to $649 million from $791 million and depreciation, depletion and amortization down to $722 million from $769 million.

The balance sheet transformation is the story. Total debt fell to $3.7 billion from $5.0 billion at year-end 2025 following an April senior note redemption, leaving net debt at $3.1 billion and a peer-leading leverage ratio near 0.5x. The company repurchased $530 million of stock in the quarter and $849 million year to date, roughly 4% of shares outstanding, and authorized a fresh $1 billion buyback. The base dividend of $0.575 per share is payable September 3 to holders of record August 13. Expand announced the Twin Eagle Holdings acquisition on July 27, positioning itself as North America's leading integrated natural gas company.

Shell is placing Canada at the center of its long-term growth strategy through LNG infrastructure and a takeover of Calgary-based ARC Resources.

Technical Structure: $2.872 Is the Gate

The confluence at $2.872 decides August. It combines the descending trend line drawn off the July highs with the 61.8% Fibonacci retracement of the move from the $3.00 area down to $2.682, and it has capped the recovery so far. That kind of overlap between a trend line and a major retracement level tends to hold on the first test and break on the third.

Below it, the 50% retracement sits at $2.836 and the 38.2% level at $2.799, both directly overhead from $2.766. Those are the first two hurdles a genuine recovery has to clear before the $2.872 gate even comes into play, and the shallow higher-low structure since the low suggests buyers are working through them slowly rather than aggressively.

Support runs first to $2.682, the session low that produced the current bounce, and that level now defines the entire near-term structure. Losing it on a closing basis confirms the corrective read and opens the $2.65 area, then the $2.50 psychological shelf where the structural LNG floor is generally located. Downside from $2.766 to $2.682 is 3.0%; to $2.50 is 9.6%. Modeled August low sits at $2.430, another 12.2% below spot.

Upside requires two sequential breaks. Clearing $2.872 invalidates the corrective interpretation and puts $3.00 back in play, worth 8.5% from current levels. Above $3.00, the modeled monthly high sits at $3.029 and the June Henry Hub spot average of $3.14 becomes the next reference, with July's $3.355 peak marking the ceiling of the entire summer range.

The moving average picture stays negative. Price broke through its short-term averages on the way down from $3.00 and has not reclaimed them, which is why the bounce reads as corrective rather than as a trend change. Reclaiming those averages and holding above $2.872 on a weekly close would be the first genuine signal that the July downtrend has ended.

The trading range that contains everything is $2.682 to $2.872, a band of just 7.1%. Volatility compression that tight resolves directionally, and the resolution usually arrives on a storage print or a weather model shift rather than gradually.

Forecast: $2.60 to $3.03 With Winter Doing the Heavy Lifting

Base case holds natural gas between $2.60 and $3.03 through August with the balance tilted toward the middle of that band. Spot at $2.766 sits 3.0% above the $2.682 low and 3.8% below the $2.872 confluence ceiling. Monthly modeling puts the period between a $2.430 low and a $3.029 high with an average near $2.781 and an end-of-month level around $2.814, a 1.3% decline from the July close. September projects $2.646 to $2.924 with an average of $2.792.

The bear path needs only production to hold at 110.6 bcfd while the cooler pattern reasserts across the central and eastern US. Storage builds re-accelerate toward 40-plus Bcf, the surplus pushes past 200 Bcf, and $2.682 gives way toward $2.50. That level is regarded as unsustainable given the structural export floor, but the path there is open and Permian associated gas provides no resistance on the way down. Downside from $2.766 to $2.50 is 9.6%.

The bull path requires three things. The hot August pattern has to verify in the actual load data rather than the models. LNG feedgas has to recover above 17.4 bcfd once Freeport maintenance completes. And the injection cadence has to keep printing below the five-year average build rather than merely below consensus. Hit all three and $2.872 breaks, opening $3.00 for an 8.5% move and $3.14 beyond it.

The structural setup for the fourth quarter is more constructive than the front month suggests. Record feedgas above 16.5 bcfd from Plaquemines and Golden Pass, AI and data center burn adding 2 to 3 bcfd, power sector consumption rising toward 38.1 bcfd in 2027, and December futures already above $4 all argue the winter tightening thesis is intact. Against it sit inventories forecast at 3,966 Bcf by end-October, 5% above the five-year average, and an EIA fourth-quarter estimate of $3.57.

Watch three things. Thursday's EIA print on August 6 for whether the injection stays under 30 Bcf and the surplus stops widening. Whether feedgas recovers from 17.2 bcfd as Freeport returns. And whether the December contract holds above $4, because the entire producer hedging complex, and therefore next winter's supply, is set against that number.

That's TradingNEWS