For a decade the oil industry invested on the assumption that demand would fall, so when the Strait of Hormuz was blocked in 2026 there was no spare capacity to absorb the shock and prices could only climb.
In 2023 the International Energy Agency predicted that by 2028 the world's oil companies would produce more than people wanted to buy. The industry believed it, prices fell, and money for new wells and refineries mostly stopped being approved.
No one forecast that in March 2026 the Strait of Hormuz, the main route for Middle Eastern oil, would be almost completely sealed off after US and Israeli strikes on Iran.
A forecast is a stack of what-ifs
Going from exploring a new field to shipping oil out takes about 20 years. Companies, banks and governments decide where to invest by reading long-range projections of demand a decade or more ahead.
A projection is really several sets of assumptions turned into different stories: how fast economies grow, how quickly electric cars spread, how climate policy changes. Comparing major 2025 outlooks for 2035, the highest and lowest scenarios differ by 27 million barrels a day, roughly a quarter of what the world uses today.
The further out you look, the more it behaves like a long-range weather forecast, and the more wrong it can be.
More electric cars, not less oil
In the early 2000s experts feared oil was running out. Then shale drilling lifted US output from 5 million barrels a day in 2008 to 13.6 million in 2025. In 2022 the IEA said demand would peak in the mid-2030s; a year later, with EV sales booming, it moved the peak before 2030.
Yet in Norway, where about nine in ten new cars sold in 2024 were fully electric, road fuel use fell only about 10%, because buses and trucks still run on diesel. In China, over half of new cars sold in 2024 were electric, but the country still burns more petrol than before 2020.
Why cheaper prices take years
Money follows forecasts. If demand is expected to fall, companies don't drill; when supply needs topping up, drilling, plants and tankers are measured in years.
About 90% of recent production spending went into offsetting declines in ageing fields, not adding new supply. The system was built to let supply slide gently down, not to ramp up fast.
Money committed in 2026 won't produce much oil until the 2030s. For people filling their tanks and paying heating bills, prices are likely to stay high and volatile long after the fighting stops.
Why it matters
Forecasts are not an academic exercise: they decide whether a well gets drilled or a refinery gets built today, and those decisions take a decade to pay off. It means the energy security governments talk about was already written into spreadsheets years ago.
Curated from high-quality sources, with concise summaries and key takeaways.
