The Oil Surplus Nobody Is Modeling

The global oil market was already entering structural surplus before the Iran war began. Five demand-destruction forces — each underestimated by consensus models — were bending the demand curve below supply on a timeline the IEA, OPEC, and Rapidan Energy Group have not priced. The war accelerates the timeline. It does not reverse it.

Act 1: The IEA was already wrong

Before Hormuz the market was already loose. Supply ran at 106 million barrels per day (mbpd) against 104 mbpd of demand — a 2 mbpd surplus the International Energy Agency (IEA) expected to widen to 3.7 mbpd in 2026.

None of the three big forecasters read that surplus as a turning point. Their demand lines are, at bottom, bets on transport: nearly 60% of every barrel goes to moving something — cars, trucks, ships, planes — so a long-range oil forecast is really a forecast of what the world's vehicles will run on. The IEA still had demand rising gently to 106 mbpd by 2030. OPEC has the global fleet of cars, trucks, and buses growing from 1.7 to 2.9 billion by 2050, modeling more than 70% of them will be burning gasoline and diesel. Rapidan's bull case runs on the same motorization math. All three share a common assumption: the developing world will motorize on oil the way the OECD did. It won't.

Every oil shock from 1973 through 2022 — OAPEC embargo, Iranian revolution, Gulf War, Libya, Russia — was a supply story, and the market was right to treat it as bullish, because oil had no substitute. That premise has changed. Every supply disruption from here pushes demand lower, not higher, because there is finally a way out. The Hormuz closure is not bullish for oil. It is the largest electrostate recruitment event in history.

Five forces the consensus is underestimating

China is flattening. EV sales share passed 50% in 2025, domestic oil demand growth has stalled, and coal-to-olefins (CTO) is already substituting for oil-based petrochemicals. All observed, not projected.

India is not the next China for oil. It is electrifying at solar and battery prices China never had when it motorized, and its two- and three-wheelers, the largest vehicle category, are going electric fastest. India adds roughly 1 mbpd to global demand by 2030, not the 1.5–2 mbpd the bulls assume.

Southeast Asia is leapfrogging. Vietnam, Thailand, and Indonesia already have higher EV sales shares than the United States. New vehicle purchases will be disproportionately electric because that is what is cheapest. OPEC's fleet arithmetic — 1.2 billion more vehicles by 2050, most of them assumed internal-combustion — requires those buyers to choose the technology that costs more.

Trucking electrification is ahead of schedule. CATL's Shenxing batteries and BYD's electric trucks are already operational on sub-800km routes in China, and at war-era oil prices the economics flip faster still.

Coal-to-olefins displaces the one sector where oil demand was supposed to grow. Petrochemicals represent essentially all of the IEA's remaining demand growth projection (+2.1 mbpd). CTO strips 1–2 mbpd of that structural demand. This variable has no line item in any Western energy model.

Act 2: Coal-to-olefins — the variable nobody has a line item for

I covered CTO in detail in Post 4. China's coal-based petrochemical industry displaces oil-based naphtha cracking directly, and it does so in petrochemicals, the sector carrying essentially all of the IEA's remaining demand growth. CTO margins run $112–126/ton at $80 Brent. Naphtha crackers are losing money. The higher oil goes, the wider the gap.

Hydrogen-CTO — Baofeng's architecture of solar → electrolyzer → CTO — makes the economics even more dramatic. At $0.015–0.02/kWh off-grid desert solar, the break-even drops to roughly $40 Brent. This makes CTO displacement structurally permanent regardless of oil price.

Act 3: Then the war came

The Hormuz closure crimped supply by roughly 12 mbpd. Brent spiked above $120. SPR releases, pipeline rerouting, and emergency shale ramp partially compensated, but the supply shock was real and will take 2–3 years to fully unwind as infrastructure is rebuilt.

Every week the strait stayed closed, another energy ministry re-ran the arithmetic Post 1 described — and the arithmetic now ends at a Chinese factory gate. The demand destruction the war triggers is permanent; the supply shortage is temporary.

The supply response doesn't save the market

Theoretical capacity keeps rising — shale, deepwater Brazil and Guyana, OPEC spare — on a path that gets producers to roughly 112 mbpd by 2030 if they choose to pump it. They don't, because they read the demand signal. Shale breaks first; the 2014–2016 precedent saw roughly 200 rigs shut down. Deepwater projects deliver on their own multi-year timelines regardless. OPEC defends a floor by withholding.

Actual production therefore holds near 104. Actual demand plateaus near 103. The observed surplus stays in the 1–2 mbpd range. But the gap that sets prices is wider — the unexpressed oversupply between what could be produced (~112 mbpd) and what the market can absorb (~103 mbpd). Every producer knows what their competitor could bring online. That knowledge keeps the bid soft.

SPR refilling absorbs 1–1.5 mbpd temporarily in 2028–2029 as China, Japan, and IEA nations replenish depleted reserves — a two-year inventory rebuild, not structural demand. When it ends, the gap doesn't close.

No demand recovery cycle. 2016 snapped back because the destruction was cyclical; this time, the destruction is technological.

The model that breaks — and the one that replaces it

We are watching the petrostate model break in real time. The countries whose fiscal systems depend on oil revenue above $70 — Saudi Arabia, Russia, Iraq, Nigeria — face a structural revenue crisis that no amount of OPEC coordination can solve. You cannot cut your way to prosperity when demand is falling permanently.

There will not be a vacuum. Another model is already here — one that exports electrons, batteries, and infrastructure instead of petrochemicals and prices off learning curves rather than depletion curves. One that gets cheaper every year instead of more expensive.