The War in Iran and What It Means for Energy Forecasting

Where Prices Stand Today

The Strait of Hormuz has been effectively closed since March 2 — not by naval blockade, but by insurance withdrawal. Tanker traffic is down more than 70%. Roughly 15 million barrels per day of crude and refined products are stranded, representing about 20% of global seaborne oil trade. Qatar’s LNG production is offline under force majeure after strikes on Ras Laffan and Mesaieed. More than 3 million barrels per day of Gulf refining capacity is dark. Iran maintains an estimated stockpile of 2,000 to 6,000 naval mines and has begun deploying them in the strait. No war termination mechanism exists.

As of this morning (March 25):

Brent Crude: $99/bbl, up 46% from the pre-war $68. Down 5% today on Trump’s claim that Iran is “talking sense.” Iran denies any direct negotiations. The market is trading on headline risk, not fundamentals.

WTI Crude: $88/bbl, up 38% from $64. WTI has been dragged up by the export arb, not by any physical supply disruption in the US.‍

TTF (European Gas): €17/MMBtu, up 72% from €10 pre-war. The Ras Laffan damage compounded the Qatar force majeure, and the EU Russian gas ban took effect March 18 — two supply shocks hitting simultaneously.‍

Henry Hub (US Gas): $2.90/MMBtu. Flat. Exactly where it was before the war started.

That last number is the one that matters most for this conversation.

Two Very Different Exposure Profiles

US: If you’re a US energy market participant exposed to oil, you’re feeling this. Brent is up 46%, WTI is up 38%, and anyone with crude exposure is repricing risk daily. But if your business is on the gas and power side — if you’re a utility, a retailer, a load-serving entity, a PJM market participant — Henry Hub has moved approximately zero percent since hostilities began. LNG export terminals are already running at fixed nameplate capacity — international price spikes don’t transmit back to Henry Hub. The US produces over 13 million barrels per day of crude domestically, so there’s no physical supply disruption to the gas system. ERCOT gas basis is driven by Permian associated gas and pipeline dynamics, not by what’s happening in the Strait of Hormuz. The war is on CNN and on the crude desk. It is not on the gas cost stack.

Europe: If you’re an European energy market participant, the picture is radically different. TTF gas prices have surged roughly 72%. Qatar LNG, which represents 20% of global supply, is offline. The EU Russian gas ban just went into effect. Storage sits at 29%, blowing up every refill assumption for the summer injection season. Demand pulls on the grid regardless — there are no molecules behind it.

The gap between ~0% Henry Hub impact and ~72% TTF impact is the story. On the gas and power side, US clients are insulated. European clients are exposed. The transmission mechanism between global commodity chaos and US power markets effectively doesn’t exist for gas.

On the gas and power side, US clients are insulated. European clients are exposed.

The Path to Demand Destruction

Where it gets interesting for forecasters is what happens next. The progression looks like this:

It starts with the oil shock. Brent is above $100. Every dollar over that threshold functions as a tax on the global economy. At $120 and above, you’re in recession territory. We touched $119 last week.

That drives contraction. Consumer spending declines. Manufacturing input costs spike. Capital expenditure freezes. Data center buildout timelines slip. The CapEx-intensive sectors that have been driving load growth projections start pulling back.

And that’s when load drops. Industrial demand falls first. Forecasts built on growth assumptions become overestimates. Long bias creeps in — the models assume the future looks like the recent past, and the recent past was a growth story. If recession hits, things go sideways in a way that most forecasting approaches aren’t designed to handle.

At $120 and above, you’re in recession territory. We touched $119 last week. That drives contraction.

Why This Moment Matters for Forecasting

Recession is a real possibility. And nobody has ever forecast through one with machine learning.

The last demand-driven recession was 2008. Modern ML forecasting didn’t exist. Every forecaster in the market is about to be tested on something none of them have seen. That includes Amperon — but our approach is built differently.

Legacy forecasting is curve-fitting to history. When the pattern breaks, the model breaks with it. You’re left with a human analyst re-tuning coefficients after the fact, chasing the data instead of leading it.

Legacy forecasting is curve-fitting to history. When the pattern breaks, the model breaks with it.

Our ML models learn from real-time signals — weather, grid conditions, economic inputs — and adapt as demand patterns shift. They don’t wait for a human to re-tune them. A recession is exactly the kind of regime change where the gap between ML and legacy modeling gets wide.

Our customers can trust that their forecasts will stay ahead of the curve, even when the curve changes shape. This team has been through novel events before — a global pandemic, Winter Storm Uri, Winter Storm Fern, the 2024 summer heat events. We know how to operate when the ground shifts.

Our ML models learn from real-time signals — weather, grid conditions, economic inputs — and adapt as demand patterns shift.

This is new ground for everyone. Our models and our team are ready for it.

Elliott Chorn is VP of Customer at Amperon, where he leads Customer Success, Solutions Consulting, Program Management, Market Analysis, and Meteorology. He has spent over six years at Amperon and has deep experience across demand forecasting, renewables, scheduling, and settlements for every type of energy market participant.

‍