Natural Gas Slides to $2.78 as a 36 Bcf Build Widens the Storage Surplus to 198 Bcf

Natural gas fell to $2.78 per MMBtu on Thursday, August 13, down 0.86% from the previous session, after the Energy Information Administration reported a larger-than-expected storage injection for the week ending August 7.

Energy firms added 36 billion cubic feet into underground storage against market expectations of 31 Bcf. The build compared with an injection of 49 Bcf during the same week a year ago and a five-year average increase of 33 Bcf for that calendar week. Stockpiles rose to 3.153 trillion cubic feet.

That is the second consecutive weekly beat. The prior report covering the week ending July 31 delivered 33 Bcf against a 31 Bcf consensus, against 13 Bcf in the year-ago week and a 23 Bcf five-year average. The week before that, ending July 24, produced 28 Bcf against a 35 Bcf expectation — the last time the market got a bullish surprise.

Over the past month the front-month contract has fallen 4.27% and it sits 2.15% below the same point last year. Prices hovered below $2.70 earlier this week, near the lowest level since April, pressured by the cumulative weight of consecutive above-consensus builds.

The recent trading history frames how narrow this market has become. The NYMEX August contract closed at $2.92 in mid-July, traded at $2.85 the following session, and averaged $2.91 across that week against a $3.13 average the week prior. The physical benchmark managed to nudge 8 cents higher to $3.220 on the weekly index in late July, but futures have been unable to follow.

The pattern across the summer is a market that absorbs heat, absorbs record power burn, absorbs LNG maintenance completions, and still cannot clear $3.00.

Crude is telling a different story. Brent fell 1.19% to $87.92 and WTI 1.39% to $82.11 on demand downgrades from both major agencies, but both remain up more than 28% year over year. Natural gas is down 2.15% over the same twelve months. The energy complex has split, and gas is on the wrong side of it.

Storage at 3,153 Bcf and 198 Bcf Above the Five-Year Average

The absolute inventory position is the single most important number in this market and it has been building all summer.

Working gas in storage stood at 3,153 Bcf as of Friday, August 7. That is 25 Bcf below the same week last year — a 0.8% deficit — but 198 Bcf above the five-year average of 2,955 Bcf, a surplus of approximately 6.7%. Total working gas remains within the five-year historical range.

The trajectory through the injection season has been remarkably consistent. Storage stood at 3,024 Bcf on July 10 after a 41 Bcf build, 181 Bcf above the five-year average. It reached 3,056 Bcf on July 17 after 32 Bcf, a 183 Bcf surplus. It hit 3,084 Bcf on July 24 after 28 Bcf, at 185 Bcf above average and 6.4% higher than the norm. It reached 3,117 Bcf on July 31 after 33 Bcf, holding 6.7% above the five-year mark. And now 3,153 Bcf with the surplus widening to 198 Bcf.

That progression matters because it shows the surplus expanding through the hottest weeks of the year. Analysts had expected the cushion to narrow slightly to around 6.6% for the week ending August 7. It came in at 6.7% instead, and the absolute gap widened by 13 Bcf.

Earlier in the season the surplus had been projected to push as high as 171 Bcf on one weekly print, described at the time as the largest excess of 2026. It is now 198 Bcf, comfortably beyond that.

The year-over-year comparison is the only genuinely constructive element. At 25 Bcf below last year, inventories are not building faster than 2025 in absolute terms. The problem is that 2025 was itself a heavy year, and the five-year average includes tighter periods. A market 0.8% below last year and 6.7% above the five-year norm is not tight by any working definition.

Total commercial inventory heading into the withdrawal season is what sets winter price formation, and on the current trajectory the market is going to arrive at the end of October with the largest cushion in a decade.

Record Production at 111.2 Bcf/d Is the Binding Constraint

Supply is the reason storage keeps building through record cooling demand, and the number has kept climbing.

Natural gas production in the Lower 48 states averaged a record 111.2 Bcf/d in August, up from 110.7 Bcf/d in July. That is the highest monthly output in the history of the U.S. market, and it is arriving on top of a base that was already elevated.

For context, dry gas production averaged just above 98 Bcf/d in mid-2022 and forecasters at the time projected growth to 99.7 Bcf/d. The market has added more than 11 Bcf/d of daily supply since — roughly the equivalent of adding an entire European import market's worth of production in four years.

Additional supply from West Texas pipelines coming online is expected to keep the market well supplied through the balance of the injection season. Those takeaway additions release associated gas that had previously been constrained, and associated gas is price-insensitive: it flows because the oil economics work, regardless of what Henry Hub does.

That last point is critical to the forward outlook. With WTI at $82.11 and Brent at $87.92 — both up roughly 28% and 32% year over year — Permian oil economics are strongly supportive, which means associated gas volumes have no reason to decline. The gas market is receiving a supply stream whose production decision is made in a different commodity's price environment entirely.

Strong output combined with relatively mild weather earlier in the year has kept inventories above the five-year average since March. That is a five-month stretch during which the surplus has never closed, through a spring shoulder season and the peak of summer cooling demand.

Demand has genuinely improved. Lower 48 demand recently registered 81.8 Bcf/d, up 5.5% year over year. Natural gas-fired electricity generation has been rising this year, in contrast to the year-over-year declines that occurred in 2025 when gas prices were sharply higher. The Natural Gas Supply Association's summer outlook projected record gas-fired power burn at 40.3 Bcf/d.

Demand at record levels against supply at record levels produces a surplus of 198 Bcf. Supply is winning.

The EIA Sees $3.00 as a Ceiling Until November

The August Short-Term Energy Outlook delivered the most explicit bearish framing the agency has published on domestic gas this year.

Henry Hub spot prices are expected to remain below $3.00/MMBtu until November and to average $3.03/MMBtu over the remaining five months of the year — nearly 50 cents lower than the prior month's forecast. That is a substantial downward revision delivered in the same monthly cycle in which the agency raised its Brent forecast by $11 per barrel.

The reasoning is straightforward. LNG terminal maintenance combined with record production is building the largest storage cushion heading into winter in a decade. Prices are expected to rise gradually in the coming months but remain relatively low because inventories are well above the five-year average. Futures prices show a similar pattern, with contracts through September 2026 remaining depressed.

That forecast represents a significant deterioration from earlier in the year. The January outlook had projected the annual average Henry Hub price decreasing about 2% to just under $3.50/MMBtu in 2026 before rising sharply to just under $4.60/MMBtu in 2027. A December 2025 revision had projected $4.30/MMBtu across the November-through-March winter and $4.00 for full-year 2026.

The market is now trading roughly 30% below that December projection with three and a half months of the year remaining.

The supply-demand balance underpinning the longer-term view has not changed. Supply growth was forecast to outpace demand growth by 0.5 Bcf/d in 2026 before falling behind by 1.6 Bcf/d in 2027, driven mainly by feed gas demand from LNG export facilities. Annual average spot prices were projected to decrease 2% in 2026 and then increase 33% in 2027.

The 2027 inflection is the entire bull case, and it is eighteen months away. Between now and then the market must absorb record production into a storage complex that will arrive at winter with more gas than it has held in ten years.

A Record 3,985 Bcf Heading Into Winter

The end-of-season inventory projection is the number that determines whether this market has any upside before the fourth quarter.

The EIA forecasts natural gas inventories at 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 place stocks at their highest level heading into winter since 2016.

From the current 3,153 Bcf, reaching 3,985 Bcf requires roughly 832 Bcf of net injections across the remaining twelve weeks of the injection season, or about 69 Bcf per week. At the current run rate of 28 to 36 Bcf, the market would fall well short of that projection — which would be bullish if the comparison were valid.

It is not. Injection rates rise sharply through September and October as cooling demand collapses and heating demand has not yet begun. The five-year average injection for early October is roughly 94 Bcf per week, and builds above 80 Bcf are routine in the shoulder. The 69 Bcf weekly average required is achievable and roughly consistent with normal seasonality.

The relevance of the 3,985 Bcf figure is what it does to winter price formation. Periods with higher-than-average inventories are associated with lower prices, while lower storage corresponds with higher prices and tighter conditions. A market entering the withdrawal season 5% above the five-year average with a decade-high absolute cushion has a substantial buffer against cold weather.

That buffer is precisely what the forward curve is pricing. Henry Hub contracts for December 2026 through February 2027 have averaged $4.160, which is a meaningful premium to spot but well below the levels a genuinely tight winter setup would command.

The asymmetry for a winter position is unattractive from here. A normal winter delivers roughly the $4.16 strip. A warm winter with 3,985 Bcf in the ground takes the entire curve lower. Only a genuinely cold December against LNG demand that materializes faster than projected produces meaningful upside, and the storage cushion absorbs the first several weeks of any cold snap.

August Heat Delivered Record Power Burn and Did Not Dent the Surplus

The most damaging data point for the bull case this summer is that the weather worked and the price did not respond.

U.S. natural gas-fired generation consumed more gas over the past week than during any week of July, despite record heat earlier in the season. Forecasts pointed to significantly hotter conditions across the central and southern United States, which should boost electricity demand as air-conditioning use rises. Above-normal temperatures were projected to persist through August 25.

Some of the strongest late-summer cooling demand in decades was expected to do little more than modestly reduce the storage surplus, with persistently high production and an uneven response from power and LNG demand limiting the impact of widespread August heat.

That assessment has proven accurate. The surplus to the five-year average widened from 185 Bcf on July 24 to 198 Bcf on August 7, across the hottest stretch of the year.

Total generation by the U.S. electric power sector was up 37 billion kilowatt hours, or 1.8%, in the first half of 2026 compared with the same period in 2025. The composition worked against gas. Solar generation was up 21% and wind up 6%, reflecting continuing capacity growth, and new additions including a 3.7 gigawatt wind project are expected to sustain those rates through the second half and into 2027. Hydropower was up 9% in the first half, though a 3% decrease is forecast for the second half on intensifying western drought.

Renewable generation growing at 21% for solar takes the marginal megawatt-hour away from gas at exactly the hours when cooling demand peaks. The load is there; gas is capturing a shrinking share of it.

Electricity output dipped in some recent weeks despite the heat, showing signs of weakening burn rates. Weather forecasts pointing to cooler conditions after August 25 remove the last near-term demand support.

Despite repeated bouts of intense heat and soaring electricity loads this summer, futures and forward physical prices have struggled to find momentum since spring while inventories coast into winter at comfortable levels.

The Seven-Times Spread: $2.78 at Henry Hub Against Roughly $19.90 in Europe

The defining anomaly of the 2026 gas market is the gap between what American gas is worth in America and what gas is worth everywhere else.

European TTF front-month has been trading between €54 and €62 per MWh. It reached €61.8 on Tuesday as prospects for a Hormuz deal faded, fell below €59 on Wednesday as talks between Iran and Oman were reported at an advanced stage, and has been quoted near €54.10. The July average was €53.48 with a range of €43.02 to €63.14, and the front-month stood at €59.44 on July 31, up 30% from end-June.

At €59/MWh and current exchange rates, European gas costs roughly $19.90/MMBtu. Henry Hub trades at $2.78. That is a ratio above seven to one.

European storage is the reason. EU inventories stood at approximately 57% full in early August, around 11 points below the prior year, against a binding target that was lowered from 90% to 80% by November 1. Storage recently fell to the lowest seasonal levels in records going back to 2009. A buffer ending the refill season in the low seventies rather than comfortably above 90% leaves materially less headroom than a normal year.

Persistent European heatwaves are boosting power demand for cooling, and analysts expect a firm floor under European prices until storage shows clearer signs of building ahead of the heating season. One bank raised its year-end European forecast to €50/MWh from €45/MWh; a major utility expects €50 to €60 as long as Hormuz remains closed.

Every European storage purchase competes with Asia for a shrinking pool of free cargoes. The JKM-TTF spread flipped from a European premium of $0.90/MMBtu in January and February to an Asian premium averaging $2.80/MMBtu.

That spread should be the most powerful bullish force imaginable for U.S. gas. It is not, because the spread cannot be arbitraged. Liquefaction capacity is the binding constraint, and it is fixed in the short run. American producers can see $19.90 gas in Rotterdam and sell it for $2.78 in Louisiana because there is no additional train to put it through.

Ras Laffan Removed 12.8 Million Tonnes and US Gas Did Not Move

The global LNG supply shock of 2026 was the largest in the industry's history, and Henry Hub is trading lower than it was a year ago.

On March 2, Qatar shut LNG production at Ras Laffan after Iranian drone and missile strikes on the complex, which houses 14 operational trains generating approximately 77 million tonnes per annum and roughly 20% of worldwide LNG supply. European TTF surged as much as 54% within 24 hours and Asian JKM rose 39%.

The damage assessment was severe and durable. Two of the 14 trains — S4 and S6 — plus one of two gas-to-liquids facilities were damaged and taken offline. Qatar expects to lose 12.8 million tonnes per year of capacity, roughly 17% of its export volumes, with repairs likely to take three to five years. Force majeure was declared on long-term contracts to destinations including Italy, Belgium, South Korea and China.

The medium-term supply consequences extend far beyond the immediate outage. Damage to Qatar's liquefaction infrastructure has reduced the outlook for global LNG supply growth and is expected to delay the unfolding LNG wave by at least two years. The attacks could reduce Qatari output by nearly 70 billion cubic metres by 2030 assuming a four-year repair period, with delays to the North Field East expansion cutting a further 20 bcm across 2026 to 2030.

Layered on top, approximately 4.2 Bcf/d of LNG normally transits the Strait of Hormuz, which has been largely shut since the conflict began. Observed traffic dropped to roughly four vessels per day. Global LNG exports fell to a six-month low at one point, with shipments dropping roughly 20% to 1.1 million tonnes.

The world lost 17% of its largest supplier's capacity for half a decade, saw the primary export chokepoint close, and watched European prices nearly double from pre-war levels. American natural gas is at $2.78 and down 4.27% over the past month.

That disconnect is the entire structural story: the U.S. market is physically isolated from the crisis it is best positioned to solve.

LNG Exports at 15.9 Bcf/d Are the Only Bridge and They Are Full

Feed gas demand from export terminals is the mechanism through which global scarcity would reach Henry Hub, and it is running at capacity.

LNG exports have been steady around 15.9 Bcf/d, with flows to Gulf Coast terminals climbing to their highest level in more than a month as some facilities completed seasonal maintenance. That maintenance completion is the near-term bullish development the market has been waiting for, and it has produced no sustained price response.

The comparison over time shows how the ceiling binds. Exports averaged 16.1 Bcf/d in October 2025, up from 15.7 Bcf/d in September. Nearly a year later the run rate is essentially unchanged. Additional Mexican demand has been absorbing volumes as new capacity comes online, driven by stronger demand and gas-fired power.

The EIA's longer-term thesis rests entirely on this variable. Demand growth is forecast to outrun supply growth by 1.6 Bcf/d in 2027, driven mainly by more feed gas demand from LNG export facilities, reducing storage and pushing annual average spot prices up 33%.

That is a 2027 story requiring new liquefaction trains to reach commercial operation. Until they do, the domestic market clears domestically, and domestically it has 111.2 Bcf/d of production against roughly 82 Bcf/d of demand plus 15.9 Bcf/d of exports.

The arithmetic explains the entire summer. Every incremental Bcf of production has nowhere to go but storage, and storage has been absorbing it at a pace that widened the surplus to 198 Bcf through the hottest weeks of the year.

For traders, the practical implication is that LNG feed gas is a demand floor rather than a demand catalyst. A terminal outage removes 1 to 2 Bcf/d and is immediately bearish. A maintenance completion restores it and is marginally bullish. Neither changes the balance, because the ceiling on total export capacity does not move until new trains start.

Rigs and Frac Spreads Jumped 18 Units on a $4.16 Winter Strip

The supply response to the winter curve is already underway, and it is the clearest signal that the market expects the current weakness to persist into higher prices only briefly.

Lower 48 rigs and frac spreads increased by a combined 18 units in a single week — the highest combined weekly increase in those two indicators since the week ended September 13, 2024. That activity surge came while spot prices were grinding toward four-month lows.

The economics driving it sit further out the curve. Henry Hub contracts for December 2026 through February 2027 have averaged $4.160, roughly 50% above current spot. It takes anywhere from four to six months for new wells to progress from spud to sales, which places production from this activity increase directly into the December-through-February window.

That is the classic shale feedback loop compressed into a single observation. A forward curve at $4.16 signals producers to add rigs in August; those wells begin flowing in December through February; the incremental supply arrives precisely when the winter premium was supposed to be realized.

For a market already projected to hold a record 3,985 Bcf at the end of October, adding supply into the withdrawal season is the mechanism that caps the winter rally before it starts.

The gas-directed rig count had been running near 122 rigs earlier in the year against 410 oil-directed rigs and a total count around 543. The distinction matters less than it once did because associated gas from oil drilling now represents a substantial share of incremental supply, and with WTI at $82.11 that drilling is not slowing.

The counterpoint is that four to six months of lead time means nothing arrives before November. If the market gets a genuinely cold start to winter, the December-to-February strip can reprice sharply before the new supply lands. That is the only clean bullish trade available in this market, and it requires weather rather than fundamentals.

PPI Energy Fell 3.1% While Natural Gas Rose in the Same Report

Thursday's Producer Price Index release contained a detail specific to this market that most coverage missed.

Final demand was unchanged at 0.0% against a 0.2% consensus, with the annual rate falling to 4.7% from 5.5%. Final demand energy dropped 3.1%, gasoline fell 5.7%, and crude petroleum collapsed 11.9% at the unprocessed stage. Diesel fell 6.7% and jet fuel declined.

Within unprocessed goods for intermediate demand, where energy materials dropped 7.4% overall, prices for natural gas advanced. In processed goods, commercial electric power moved higher alongside lubricating oil base stocks even as the broader energy category fell 3.1%.

Natural gas rising in a month when every other energy input collapsed is a function of power burn rather than of any tightness in the gas balance. Utilities bought more gas because cooling load was extreme, and that showed up in producer prices even as the futures curve stayed pinned.

The macro backdrop is otherwise neutral for this market. Wednesday's CPI eased to 3.4% headline and 2.5% core. The federal funds target remains 3.50% to 3.75% after five consecutive holds and a 9–3 July vote, with September hold odds at 60%, October hike odds above 53% and December at 73%. The dollar index sits at 99.87 to 100.03 after nine sessions inside a 50-basis-point band.

None of that moves natural gas. Unlike crude, which trades as a global macro asset with a dollar sensitivity and a geopolitical premium, Henry Hub is a domestic physical market cleared by pipeline capacity, storage economics and weather. Rate expectations affect it only through industrial demand, and the gas-weighted manufacturing index has been soft, with industrial sector consumption forecast to decrease in 2026 and 2027 on closer-to-normal weather and decreased activity.

The one macro channel that matters is the Middle East, and it reaches gas through LNG rather than through rates.

Technical Structure: $2.70 Floor, $2.92 Resistance, $3.00 Ceiling

The chart has compressed into a narrow band and the levels are defined by recent trade rather than by long-dated retracements.

Immediate support sits at $2.70, the area the market probed earlier this week and the lowest level since April. Below that, $2.65 marks the low registered in the first week of August and represents the floor of the summer range. A decisive break there opens $2.50, which has no recent test and would represent the weakest print in more than a year.

Resistance begins at $2.85, where the front-month traded in mid-July. Above that, $2.92 was the August contract close on July 15 and marks the top of the recent range. The psychological $3.00 level is the operative ceiling and carries institutional weight given the EIA's explicit forecast that prices remain below it until November. Beyond $3.00, the physical weekly index at $3.220 in late July is the next reference, followed by $3.50.

At $2.78 the market sits almost exactly in the middle of a $2.65 to $2.92 band that has contained it for six weeks.

The forward curve provides the structural context. December 2026 through February 2027 contracts average $4.160, which is a $1.38 premium to spot. That contango is unusually wide for this point in the injection season and reflects a market that expects the current glut to clear by winter rather than one pricing distress.

Wide contango has a mechanical consequence: it pays storage operators to inject and hold, which reinforces the build pattern and suppresses prompt prices further. Every week the front stays cheap relative to winter, the incentive to fill storage strengthens, and the projected 3,985 Bcf end-October figure becomes more likely.

The clean decision levels: holding $2.70 keeps the range intact and allows a retest of $2.92. Losing $2.65 opens $2.50 and confirms the storage surplus is overwhelming seasonal demand. Clearing $3.00 on a weekly close would invalidate the EIA's near-term framing and require either a production disruption or a weather event.

Scenarios and Targets: $2.50 Downside, $3.00 Base, $4.16 Winter Strip

The base case, carrying the highest probability, is continued range trade between $2.65 and $3.00 through October. This assumes production holds near 111.2 Bcf/d, LNG exports stay near 15.9 Bcf/d, injections run at or above the five-year average, and weather turns seasonal after August 25. Storage arrives near the projected record 3,985 Bcf at end-October, and Henry Hub averages close to the forecast $3.03 across the remaining five months. Target: $3.00.

The bearish case begins with a break of $2.65. Triggers are straightforward: cooler-than-normal September temperatures, an LNG terminal outage removing 1 to 2 Bcf/d of feed gas, or continued weekly builds above the five-year average widening the surplus past 220 Bcf. Under that scenario the market tests $2.50, a level that would put the projected end-October surplus above 5% and force storage economics to do the balancing. Target: $2.50.

The bullish case does not come from fundamentals; it comes from weather and from the calendar. A cold start to the withdrawal season with 3,985 Bcf in the ground would still support a rally, because the December-through-February strip at $4.160 is priced for normal winter rather than for a cold one. Hurricane season disruption to Gulf production or Gulf Coast LNG infrastructure would work through the same channel. A Hormuz escalation that further damages global LNG supply would lift the international arbitrage and pull U.S. feed gas demand to the maximum the terminals can process. Target: $3.50 in the prompt, with the winter strip toward $4.60.

The longer-dated structural case remains intact and is unchanged by this summer. Supply growth is projected to fall behind demand growth by 1.6 Bcf/d in 2027 on LNG feed gas, with annual average spot prices forecast to rise 33% to just under $4.60/MMBtu. Storage inventories are expected to move gradually below the rolling five-year average across that forecast horizon.

That is an eighteen-month trade. Positioning for it at $2.78 with contango at $1.38 costs roughly 50% in roll yield across the interval, which is why the curve rather than the spot price is the correct expression.

Verdict: A Global Shortage the American Market Cannot Reach

Natural gas at $2.78 has fallen 4.27% over a month and sits 2.15% below the same point last year, after a 36 Bcf injection beat a 31 Bcf consensus and pushed working gas to 3,153 Bcf — 198 Bcf and 6.7% above the five-year average of 2,955 Bcf.

The surplus widened through the hottest weeks of the year. It stood at 185 Bcf on July 24 and 198 Bcf on August 7, expanding across a stretch that delivered the strongest weekly power burn of the summer with above-normal temperatures forecast through August 25. When record cooling demand cannot close a storage gap, the gap is a supply problem.

Production is the answer. Lower 48 output reached a record 111.2 Bcf/d in August against 110.7 Bcf/d in July, with West Texas pipeline additions still to come and Permian associated gas flowing on oil economics at $82.11 WTI rather than on gas economics at $2.78. Renewables are taking the marginal generation hour, with solar up 21% and wind up 6% in the first half.

The EIA has responded by cutting its five-month forecast by nearly 50 cents to $3.03 and stating that Henry Hub stays below $3.00 until November, with inventories projected at a record 3,985 Bcf at end-October — the largest cushion heading into winter since 2016.

Against that domestic glut sits the most severe global LNG supply shock in the industry's history. Ras Laffan lost 12.8 million tonnes per year of capacity, 17% of Qatari exports, with repairs running three to five years and up to 70 bcm of output potentially removed by 2030. Roughly 4.2 Bcf/d of LNG normally transiting Hormuz has been largely halted. European storage sits near 57% full against an 80% November target, at the lowest seasonal level in records going back to 2009. TTF has traded between €54 and €62, or roughly $19.90/MMBtu.

Seven times the price, and American gas cannot get there. Liquefaction capacity is fixed at roughly 15.9 Bcf/d in the near term, and it is running full. The arbitrage exists on paper and cannot be executed until new trains start, which is the entire basis for the 2027 forecast of a 1.6 Bcf/d deficit and a 33% price increase.

Producers are already responding to the wrong signal. Rigs and frac spreads jumped 18 combined units in a single week — the largest increase since September 2024 — on a December-through-February strip averaging $4.160. Four to six months from spud to sales places that supply directly into the winter window it was meant to capture.

Base case $3.00 with the range holding into October. Bear case $2.50 on a break of $2.65 with cooler September weather. Bull case $3.50 prompt and $4.60 on the winter strip, and it requires cold weather, a hurricane, or another Hormuz escalation — not fundamentals. The level that matters is $2.65, and the date that matters is the first hard freeze.

That's TradingNEWS