China’s March 2026 oil shock: supply risks and the limits of U.S. leverage

China faced a serious oil-supply shock in March 2026 after the war that began on February 28 disrupted shipping through the Strait of Hormuz. In its March 12 assessment, the International Energy Agency estimated that Gulf producers had cut output by at least 10 million barrels a day. The shock exposed China to interrupted deliveries and higher costs across a global oil market.

The IEA's March oil-market report described crude and product flows through Hormuz falling from roughly 20 million barrels a day before the conflict to a trickle. The report captures conditions during the March disruption. China's immediate challenge was finding replacement cargoes and managing inventories while shipping remained restricted.

For refiners, the immediate pressure came through transport costs, access to discounted cargoes and industrial margins. A cargo delayed at sea creates a different problem from an oil field that cannot produce, and each may need a different response. Replacement imports can restore volumes while leaving buyers with higher delivered costs or a less suitable crude grade. The effect on a refinery depends on both the barrels it receives and what it has to pay for them.

Venezuela mattered more to particular refiners than to China's entire supply

In a January 7 analysis, Columbia University's Erica Downs and Luisa Palacios described Venezuela as a modest supplier to China overall, but an important source of discounted crude for independent refiners. Their estimate put Venezuelan supply at about 3% of China's total crude imports, while warning that official customs records understated the flow because shipments could be relabeled through third countries.

That is a different denominator from the share of Venezuelan exports bought by China. A country can sell most of its exports to one customer while supplying only a small part of that customer's needs. A refinery's replacement bill depends on the price and quality of the alternative crude it buys.

The U.S.-China Economic and Security Review Commission's January 13 fact sheet estimated Venezuelan oil at 4%–4.5% of China's seaborne imports and identified possible substitutes from Canada, Iran and Iraq. Its seaborne denominator differs from total crude imports. Seaborne imports exclude land routes, so that figure describes a narrower slice of Chinese supply than total crude imports.

Substitution can still hurt. A replacement barrel may involve a different purchase price, voyage, financing arrangement or refinery yield. Exposed refiners faced a procurement and margin problem. The cost would vary with replacement volumes, delivered prices and the duration of the disruption, rather than follow a fixed annual bill.

Iranian oil was a larger exposure, but dependence ran both ways

The Commission's March 16 China-Iran fact sheet estimated that China bought almost 1.4 million barrels a day of Iranian oil in 2025, about 12% of China's total crude imports. It also estimated that China purchased upward of 90% of Iran's crude exports. These figures describe different relationships: Iran's reliance on a buyer and China's reliance on a supplier.

These trade figures show the exposure of refiners and importers. Effects on the wider economy depend on how fuel shortages and prices pass through transport, manufacturing and consumer demand, and on how quickly buyers can secure replacement cargoes.

Sanctions and shipping affected supply through different channels. Treasury's March 13 announcement issued Venezuelan licenses addressing specified oil and petrochemical activities, supplies and contingent investment contracts. Those licenses changed which activities were authorized. Physical supply still depended on cargoes being loaded, transported and delivered, as well as buyers' access to finance and shipping services.

Underground reserves are not immediately available supply

Venezuela's resource base gives it long-term importance, while its production history shows the difficulty of bringing those barrels to market. The U.S. Energy Information Administration's February 2024 country analysis put Venezuelan proven crude reserves at roughly 303 billion barrels in 2023, about 17% of the global total. Venezuela nevertheless produced only 0.8% of global crude that year.

The same analysis described much of the resource as extra-heavy crude and documented constraints involving expertise, capital and infrastructure. Those historical figures illustrate the gap between a large stock underground and the daily flow a country can deliver. They are not a current production forecast.

Multiplying a reserve estimate by a benchmark oil price would produce a gross arithmetic figure, not an asset valuation or immediately spendable wealth. Such a calculation omits extraction costs, investment, timing, fiscal terms and the quality of the crude. It also assumes the entire reserve could be sold at one price. Turning reserves into saleable exports takes investment, operating capacity and a route to paying customers.

Inventories buy time; their definition determines the calculation

China also had substantial oil inventories. In her March 4 assessment, Columbia's Downs cited a Kayrros estimate of 1.39 billion barrels in storage as of March 2, equivalent to about 120 days of net crude imports at the 2025 rate. She also identified Iranian oil in floating and bonded storage as potential near-term supply. These were estimates of a buffer, not a guarantee that every stored barrel was immediately available to every refinery.

A reserve-duration calculation must identify the shortfall it is meant to cover. For illustration, 100 million usable barrels divided by a disruption of 2 million barrels a day equals 50 days. Dividing the same inventory by 10 million barrels a day gives 10 days. Neither hypothetical result is a forecast for China. The exercise shows why an estimate of “weeks remaining” is incomplete without both the usable stock and the daily draw assumption.

The distinction between strategic, commercial, bonded and floating inventories also matters. These stocks can have different owners, release arrangements and routes to refineries. A large national total may include barrels that are difficult to move to the plants that need them. Usable inventory and delivery capacity together determine how much protection a stockpile offers during a prolonged interruption.

Emergency releases cushion a flow shortage over time

On March 11, the IEA's member countries agreed to make 400 million barrels from emergency reserves available to the market. The agency specified that releases would follow national circumstances and schedules. An announced total is not a statement that all those barrels enter the market on the announcement day.

As a scale comparison, dividing 400 million barrels by the March report's minimum 10-million-barrel daily Gulf production reduction gives 40 days. This calculation compares a stock with a flow; it does not predict the release schedule, the duration of the conflict or China's supply position. Allocation, delivery rates and the size of the eventual shortfall determine how much relief reaches buyers.

A long interruption can exhaust buffers, disrupt access to the right grades of crude and increase costs even when aggregate stocks appear large. Sustained import shortfalls, accelerating inventory draws and constrained replacement supplies would put greater pressure on buyers. Restored shipments and slower draws would offer relief. The duration and distribution of the shortage determine how far the disruption spreads from individual refineries to the wider economy.

An oil shock does not directly measure China's AI capacity

China's artificial-intelligence industry depends directly on electricity supply. The IEA's 2025 analysis of energy supply for AI estimated that coal supplied nearly 70% of electricity consumed by Chinese data centres, with renewables near 20% and nuclear close to 10%. The report describes the physical electricity mix, rather than operators' contractual purchases. Coal, renewables and nuclear therefore played the central roles in powering those facilities.

Oil disruption can affect transport, industrial inputs and the wider economy. For computing facilities, the immediate constraints are electricity availability, power contracts, equipment delivery and project financing. Higher crude prices can increase costs elsewhere in that supply chain, while access to reliable grid power remains a separate concern for operators expanding their server capacity.

The IEA's separate AI and energy-security analysis identified grid connections, power equipment and concentrated mineral supply chains as constraints on data-centre expansion. Its scenarios suggested grid constraints could delay about 20% of planned global capacity through 2030. That is a global conditional estimate, not a China-specific outcome. For technology readers, these are more direct indicators to examine alongside oil exposure when assessing the consequences of an energy shock.

Shipping access, replacement crude and the pace of inventory drawdowns were central to China's ability to absorb the March shock. Longer interruptions would increase the pressure on refiners and industrial buyers; restored trade routes would ease it. For investors, the outcome depended on how those conditions developed, with supply costs and company performance remaining uncertain.