Opinion — Henry Hub Isn't a Weather Trade Anymore — Weekend Edition

The US withdrew 52 billion cubic feet from gas storage last week. The five-year average for the same week is 168 bcf. Weather-adjusted demand ran 69 percent below normal. And yet NYMEX Henry Hub front-month settled Friday at $2.91, holding comfortably above the $2.75 floor that should have cracked weeks ago. Something fundamental has changed about what sets the price of American natural gas, and most of the models traders are using haven't caught up. For two decades, Henry Hub was a weather derivative with a basis component. Cold winters drove demand. Hot summers drove power burn. Storage levels set the floor. That model broke sometime in the past six months. The number that broke it: 141 billion cubic feet in weekly LNG vessel departures, up 26 bcf from the prior week, even with maintenance underway at several export terminals. US LNG feedgas demand has quietly become the marginal price-setter for Henry Hub. Not weather. Not storage. Exports. The positioning data confirms the market hasn't adjusted. Managed money holds a net short of 96,289 contracts in NYMEX Henry Hub gas front-month, though it covered 23,581 contracts last week. That net short made sense when Henry Hub was a domestic weather play and storage sat 8 percent above year-ago levels. It makes considerably less sense when the price floor is being set by international LNG buyers. The demand destruction everyone expects from mild weather is real. It's also irrelevant to the price. The 52 bcf withdrawal versus the 168 bcf five-year average tells you domestic heating demand has collapsed. But Henry Hub didn't follow it down because the gas isn't staying in the country. The risk is asymmetric. If the Hormuz crisis resolves and LNG demand normalises, Henry Hub falls back to $2.50 and the shorts were right. But if the crisis extends, the US gas market tightens into summer cooling season with export terminals running flat out and storage injections lagging. For anyone still running weather-based gas models, the question isn't whether to update the framework. It's how much money the old framework has already cost you.