The Great Divergence: $2.70 vs. $15
The arithmetic is stark. A US gas producer selling at Henry Hub receives roughly $2.70 per MMBtu. After liquefaction, shipping to Europe, and regasification — approximately $4.00-5.00/MMBtu in total cost — that same molecule is worth $14.80 at TTF. The net margin for LNG exporters sending US gas to Europe is roughly $7-8/MMBtu. For Asia, the margin widens further: JKM at $15.80-18.00 minus total delivered cost of ~$7.50-8.50 yields a netback of $7-10/MMBtu above Henry Hub. (Source: Rystad Energy LNG netback models, May 2026; S&P Global Commodity Insights)
This is the widest the LNG arbitrage has been since the 2022 energy crisis — and unlike 2022, this time the spread is not a function of Russian supply panic. It is structural. European gas storage is full, but the continent has structurally reduced its pipeline dependence on Russian gas. Asia's demand is growing steadily, led by China's coal-to-gas switching policy and India's industrial expansion. The US has the resource base to satisfy both — and the price signal to build the infrastructure is screaming.
The Henry Hub-to-TTF spread of roughly $12/MMBtu represents the largest inter-basin gas arbitrage in the world. For every 1 Bcf/d of LNG exported, the value uplift from Henry Hub to global prices is approximately $4.4 billion annually at current spreads.
The LNG Export Wave: 2.4 Bcf/d of New Capacity
Two major projects are poised to reshape US LNG export capacity in 2026-2027. Golden Pass LNG — a joint venture between ExxonMobil and QatarEnergy at Sabine Pass, Texas — is on track to begin commissioning its first train in H2 2026, adding approximately 2.0 Bcf/d of nominal liquefaction capacity. Corpus Christi Stage 3, Cheniere's midscale expansion, is adding another approximately 0.4 Bcf/d. Combined, these two projects will boost total US LNG export capacity by roughly 15%. (Source: U.S. Department of Energy LNG export authorizations; FERC project filings; Cheniere Q1 2026 earnings call)
When fully operational, the US will have roughly 16 Bcf/d of liquefaction capacity — more than any other country by a wide margin. The implications are profound. Every incremental Bcf/d of LNG exports removes approximately 1 Bcf/d of gas from the domestic market, tightening the US supply-demand balance. At 16 Bcf/d, US LNG exports will consume roughly 17-18% of total US dry gas production — up from roughly 11% in 2023. (Source: EIA Monthly Natural Gas Gross Withdrawals; EIA LNG Data)
The timing matters. Golden Pass and Corpus Christi Stage 3 are coming online just as the US domestic market is already absorbing a surge in power demand from data centers and as storage surpluses are narrowing. This is not a demand shock that the market can absorb idly — it will require either a significant increase in production or higher Henry Hub prices to ration demand.
US LNG Export Capacity Timeline
| Project | Capacity | Expected Online | Operator |
|---|---|---|---|
| Golden Pass LNG (TX) | ~2.0 Bcf/d (Train 1) | H2 2026 | ExxonMobil / QatarEnergy |
| Corpus Christi Stage 3 | ~0.4 Bcf/d | H2 2026 | Cheniere Energy |
| Existing Total (2025) | ~13.6 Bcf/d | Operational | Various |
| Projected Total (2027) | ~16 Bcf/d | Full ramp | — |
AI Data Centers: The New 15 GW Demand Engine
Perhaps the most transformative domestic demand development for US natural gas is the explosive growth of AI data centers. These facilities are among the most energy-intensive installations ever built. A single large-scale AI training cluster can draw 200-500 MW of power — comparable to a mid-sized industrial plant — and operate at near-100% utilization rates. (Source: McKinsey Global Institute, Data Center Energy Demand 2026; Goldman Sachs Research)
Industry estimates peg incremental power demand from AI data centers at roughly 15 GW through 2026-2027, with the majority concentrated in regions served by gas-heavy power grids (PJM, ERCOT, MISO, and the Southeast). Assuming a 70% gas share of incremental generation (the historical average for new US capacity) and a 50% capacity factor, 15 GW of data center demand translates to roughly 4-5 Bcf/d of incremental natural gas demand. (Source: Goldman Sachs, Data Centers & Power Demand, January 2026; S&P Global Commodity Insights)
This is not a hypothetical projection. Utility-scale interconnection queues in PJM, Virginia (the world's largest data center market), and Texas are dominated by data center requests. Dominion Energy has flagged that its 2026 load forecast increased by 25% versus 2024, almost entirely due to data center growth. (Source: Dominion Energy IR Day Presentation, 2026) The practical effect is that US gas demand is growing at a pace not seen since the shale boom — and the supply side is struggling to keep up.
AI data center demand of ~4-5 Bcf/d is roughly equivalent to the entire residential gas consumption of California. This is demand that has emerged essentially out of nowhere in the span of 18 months — and it shows no sign of slowing.
Storage Surplus Narrowing: The Cushion Is Shrinking
Through the mild 2025-2026 winter, US natural gas storage levels entered the spring injection season at a significant surplus relative to the five-year average. But that surplus has been eroding faster than seasonal norms would suggest. The combination of steady LNG exports, expanding industrial demand, and the early onset of gas-fired power generation for air conditioning has drawn down the cushion. (Source: EIA Weekly Natural Gas Storage Report, May 2026)
As of mid-May 2026, working gas in storage stood at roughly 2.2 Tcf — above the five-year average of 2.0 Tcf but well below the peak injection trajectory that bears had anticipated in March. Injections have been consistently below the five-year average for four consecutive weeks. If this trend continues through the summer, the market enters the 2026-2027 winter with a storage surplus that has effectively evaporated. (Source: EIA Weekly Storage; Platts Gas Daily)
The narrowing storage surplus is the most direct sign that the US gas market is tightening. It means the spare capacity that kept Henry Hub pinned below $3.00 for much of 2025 and early 2026 is being absorbed. The question is whether production can ramp fast enough to refill storage before winter — or whether the market will need higher prices to do the work.
Supply Constraints: Can Production Keep Up?
US dry gas production has plateaued at roughly 103-104 Bcf/d through 2025 and early 2026, with the rig count remaining stubbornly low. The Permian Basin remains the most active region, but associated gas from oil-directed drilling has its own constraints: Permian crude output is growing modestly, and the associated gas-to-oil ratio is declining as operators shift toward shallower, less gassy targets. (Source: EIA Drilling Productivity Report; Baker Hughes Rig Count, May 2026)
The Appalachian Basin (Marcellus/Utica) — historically the largest US gas-producing region — is seeing production flat to slightly declining as operators shift capital toward liquids-rich plays. Haynesville, the third major gas basin, is constrained by weak Henry Hub prices that make dry gas drilling uneconomic at current levels. At $2.70/MMBtu, most Haynesville wells are cash-flow negative on a full-cycle basis. (Source: Rystad Energy Upstream; Haynesville breakeven analysis, Q1 2026)
The supply response to higher prices is not immediate. Even if Henry Hub rises to $3.50-4.00, it takes 6-12 months for the rig count to respond and new production to come online. Given that LNG exports and data center demand are growing on a linear trajectory, the market faces a window of vulnerability in late 2026 through 2027 where demand growth outpaces near-term supply additions.
At $2.70/MMBtu, most Haynesville dry gas wells are cash-flow negative on a full-cycle basis. Even if Henry Hub rises to $3.50-4.00, it takes 6-12 months for new production to come online. The market faces a supply-demand gap window in late 2026 through 2027.
The LNG Arbitrage: Structural or Transient?
The central question for the global gas market is whether the Henry Hub-to-TTF/JKM spread is structurally sustainable. The bull case is compelling: the US has the world's lowest-cost gas supply, the infrastructure to liquefy it, and an insatiable global demand pool willing to pay $12-15/MMBtu. The bear case rests on the fact that the US gas market is large enough that incremental supply can eventually respond — and that European and Asian buyers may balk at sustained prices above $15/MMBtu if weather remains mild and storage fills.
History suggests the spread narrows when global supply constraints ease, not when US prices rise. In 2023-2024, TTF fell back below $10/MMBtu for extended periods as European storage filled and LNG cargoes flowed freely. But the structural backdrop has changed. Europe's remaining Russian pipeline gas (via Ukraine transit and TurkStream) is at risk of further disruption. Asia's gas demand is structurally higher due to coal plant retirements and industrial growth. And the US is the only source of incremental LNG supply that can scale meaningfully before 2028. (Source: IEA World Energy Outlook 2025; Oxford Institute for Energy Studies)
The most likely path: the spread remains wide ($8-12/MMBtu) through 2026-2027, narrowing toward the mid-cycle level of $5-7/MMBtu as additional US and Qatari LNG capacity reaches the market in 2028-2029. For now, the US gas advantage is real, wide, and durable.
Price Outlook: $2.70 Is the Floor, Not the Ceiling
The Henry Hub forward curve for 2027 is already pricing in higher gas: calendar 2027 strips trade around $3.60-3.80/MMBtu, reflecting the market's expectation that the supply-demand balance tightens materially. But the real risk is that the forward curve is not high enough. If AI data center demand materializes faster than expected, or if Golden Pass commissioning is delayed, the market could face a genuine physical shortage that sends spot prices well above $4.00.
Scenario mapping:
- Base case ($3.00-3.50): Production slowly responds to LNG/data center demand. Storage exits summer at roughly 3.4 Tcf — adequate but not comfortable. Henry Hub averages $3.00-3.50 for 2027.
- Bull case ($4.00-5.00): Golden Pass delays push LNG startups into 2027. AI demand accelerates. A hot summer drives storage below 3.0 Tcf. The market needs $4.00+ to ration demand.
- Bear case ($2.50-3.00): A mild summer and weak industrial demand keep storage bloated. Global LNG oversupply depresses the arbitrage. Henry Hub stays range-bound below $3.00 through 2027.
Of these three scenarios, the base case and bull case carry higher probability given the structural demand drivers already in motion. The bear case requires a coordinated misfire on LNG timelines, weather, and economic activity that the balance of risks does not currently favor.
What This Means for Buyers
US industrial buyers with Henry Hub exposure: Lock in 12-24 month physical baseload hedges at current forward curves. The sub-$3.00 Henry Hub window is closing. Consider collars with floors at $2.50 and upside participation up to $4.00. If you operate in PJM or ERCOT, add weather-dependent swing coverage — data center-driven power demand is creating new volatility patterns in gas-constrained regions.
LNG buyers with global exposure: If you have long-term contracts linked to Henry Hub + liquefaction tolling fees, your delivered cost to Europe/Asia is roughly $7-8/MMBtu — a margin of $7-10 over spot TTF/JKM. This is an enormous competitive advantage. If you are a spot buyer in Asia or Europe, consider switching to US-linked indexation for new term contracts. The US is the marginal supplier for global gas, and paying JKM-linked pricing when Henry Hub is at $2.70 is a premium with no structural justification.
Multi-region buyers: The US manufacturing cost advantage driven by cheap gas is widening. Evaluate reshoring or expanding US-based production for energy-intensive operations (fertilizers, steel, petrochemicals, aluminum). The gap between US and European gas costs alone — roughly $12/MMBtu — translates to a per-unit cost advantage that no tariff or subsidy can fully offset.
Key takeaways:
- Henry Hub at $2.60-2.70/MMBtu while global benchmarks trade at $14.80-18.00 — a $12+ arbitrage
- 2.4 Bcf/d of new LNG export capacity from Golden Pass and Corpus Christi Stage 3 coming online in H2 2026
- AI data centers adding ~15 GW of power demand, translating to 4-5 Bcf/d of incremental gas consumption
- Storage surplus narrowing as exports and power demand absorb excess inventory
- US production plateaued at 103-104 Bcf/d — supply response to higher prices takes 6-12 months
- LNG arbitrage is structural — the US advantage as the world's lowest-cost gas supplier is durable through at least 2028
- Forward curves point higher — 2027 strips at $3.60-3.80, with tail risk above $4.00 if demand surprises to the upside
The Bottom Line
The US natural gas market is at an inflection point. The combination of surging LNG exports, unprecedented AI-driven power demand, and plateaued production is creating a tightening market that the current price of $2.70/MMBtu does not fully reflect. The global arbitrage — where US gas is $12/MMBtu cheaper than European or Asian alternatives — is the defining feature of today's gas market. For US producers, it represents a massive opportunity. For US industrial consumers, it is a structural cost advantage that should be hedged and defended. For the rest of the world, it is a reminder that energy security and affordability increasingly flow through the Gulf Coast.
Rystad Energy LNG netback models, May 2026 • S&P Global Commodity Insights • U.S. Department of Energy LNG export authorizations • FERC project filings • Cheniere Energy Q1 2026 earnings call • EIA Monthly Natural Gas Gross Withdrawals • EIA LNG Data • EIA Weekly Natural Gas Storage Report, May 2026 • EIA Drilling Productivity Report • Baker Hughes Rig Count, May 2026 • McKinsey Global Institute, Data Center Energy Demand 2026 • Goldman Sachs Research, Data Centers & Power Demand, Jan 2026 • Dominion Energy IR Day Presentation 2026 • Platts Gas Daily • Rystad Energy Upstream, Haynesville breakeven analysis Q1 2026 • IEA World Energy Outlook 2025 • Oxford Institute for Energy Studies
Disclaimer: This analysis is for informational and educational purposes only and does not constitute investment advice, a solicitation, or a recommendation to buy or sell any commodity, security, or financial instrument. Past performance is not indicative of future results. Data points are sourced from publicly available industry reports including the EIA, FERC, Rystad Energy, S&P Global, Goldman Sachs, and Cheniere corporate filings, and are believed to be reliable but are not guaranteed for accuracy or completeness. RZZRO may hold positions in the commodities discussed. Consult a qualified financial advisor before making investment decisions.