Permian relief, Henry Hub risk: when the marginal molecule moves locations.

Today, the marginal Permian molecule prices at Waha. After the takeaway wave, it prices at Henry Hub. That migration is the structural story Permian analysis is missing.

The bottleneck is upstream. Permian gas production runs around 29 Bcf/d in 2026, growing roughly 6% on associated-gas and GOR-creep mechanics that don't shut off when WTI weakens. Egress is constrained — Waha basis has averaged around -$2.40 to HH in 2026 YTD, with 21 sub-zero day-ahead prints already this year and 49 in the 2024 record. When all pipes are full, the marginal molecule clears at whatever Waha price keeps producers from shutting in. Henry Hub stays partially insulated; the constraint sits upstream of HH demand.

The relief wave is sequenced into a narrow window. Matterhorn Express added 2.5 Bcf/d in November 2024 — the basin ate the slack by Q1 2025. Blackcomb brings 2.5 Bcf/d to Agua Dulce in 2H 2026. Hugh Brinson adds 1.5 Bcf/d to Maypearl late 2026, plus up to 0.7 Bcf/d of compression in 2027 if Phase II proceeds. Roughly 5–6 Bcf/d firm through 2027, with another 2–3 Bcf/d if Apex and Saguaro land in the window. Permian gas growth runs roughly 1.5 Bcf/d per year; in this window, planned capacity additions briefly outpace the production tail.

When the constraint migrates, the marginal molecule does too. Relief volumes don't disappear — they clear at the Gulf Coast hubs that price against Henry Hub. The risk shifts from Waha basis to Henry Hub flat price. Whether HH softens depends on demand absorption. The LNG ramp through 2027 (Plaquemines Phase 2, Corpus Stage 3, Golden Pass Trains 2 and 3, Rio Grande Phase 1, Port Arthur Phase 1) adds roughly 6–8 Bcf/d of incremental feedgas. ERCOT power burn adds another structural sponge. EIA's STEO has supply and demand roughly matching through 2027 — in any regime where LNG slips, weather is mild, or storage is comfortable, the system runs supply-led, and the incremental Permian molecule clears through Henry Hub–linked markets.

The Marcellus precedent ran this way. From 2017 to 2019, roughly 10 Bcf/d of Northeast takeaway commissioned, Dom South basis tightened by about a dollar, and Henry Hub drifted from the low-$3s into the mid-$2s even as LNG ramped. Symmetric repricing.

As of late April, the Cal 2027 Waha strip has tightened materially. The Cal 2027 HH strip has not. One of those two markets is going to be wrong.