Oil fears and AI wobbles hit markets as Iran tanker attacks raise the stakes

Seven tanker attacks in one day is the kind of supply-side event that moves an oil price almost instantly. The mechanism is straightforward: buyers who depend on Gulf shipments fear that supply will fall, so they bid up the price now rather than risk paying more later. That is a supply shock, a sudden reduction in the amount of a good that can reliably reach the market, and it pushes the price up even before a single barrel is actually lost. The oil price hitting its highest point since May shows the market is treating the risk as real, not theoretical.

For oil-importing economies the chain runs further. Higher oil raises the cost of producing almost everything, from fertiliser to freight, so firms face higher costs and either absorb them, cutting profits, or pass them on, raising prices for consumers. Central banks then face a harder choice: cut rates to support growth, or hold firm to stop inflation from climbing again. India is a sharp example here. Its West Asia oil imports only just returned to pre-war levels in September, and a fresh price spike now arrives at exactly the moment the RBI is already managing elevated inflation. If oil stays high, the RBI's room to ease policy shrinks further.

The question at the centre of this sell-off is whether AI companies can earn back the money they are spending. Building a large datacenter costs billions of dollars before a single customer pays a subscription. That is a fixed cost, a cost that does not change with how much output the firm produces. If revenues grow fast enough, the fixed cost gets spread across millions of users and the unit cost falls sharply. That is the bull case. The bear case is that the costs keep rising, competition drives prices down, and the revenues never catch up. Thursday's sell-off suggests a growing number of investors are moving toward the bear case.

What makes this interesting from an economics angle is the gap between private and social returns. Firms invest in AI because they expect to capture the profit. But a lot of the value from AI, faster drug discovery, cheaper logistics, better translation, spills over to people who never pay the company directly. That gap between what the firm earns and what society gains is called a positive externality, and it is one reason governments often subsidise research. The honest caveat is that we do not yet have good data on how large either the private or the social returns actually are. One quarter of high costs does not settle the question of whether the whole buildout was a mistake.

A confidence reading of minus 24 means far more people expect their finances to get worse than expect them to get better. That matters because consumer spending makes up roughly 60% of UK GDP, the total value of everything the economy produces. When households feel poorer, they spend less, firms sell less, and growth slows. This is the multiplier effect working in reverse: a fall in spending by one household reduces income for the businesses it buys from, which then reduces those businesses' spending, and so on down the chain. The 11-point drop in a single month is large enough to show up in GDP figures if it persists.

The budget timing is the part I find genuinely difficult. A chancellor facing a confidence collapse has two broad options. Spend more to support demand, which risks spooking bond markets already nervous about UK debt. Or tighten the public finances, which risks making the slowdown worse by removing money from the economy at exactly the wrong moment. The 2010 to 2012 period in the UK is the precedent that haunts this debate: the coalition government cut spending into a weak recovery and growth stalled for two years. The difference now is that inflation is still elevated, which limits how much the Bank of England can offset a fiscal squeeze by cutting rates.

A PMI is built from a survey asking purchasing managers whether conditions this month are better, the same, or worse than last month. The index is then scaled so that 50 is the dividing line between expansion and contraction. India's reading of 56.7 is comfortably in expansion territory, and the services component is doing most of the work. Services in India range from software exports to local restaurants, but the common thread is that they are labour-intensive. That makes the hiring stall the most important detail in the release. Output is rising, but firms are squeezing more from existing staff rather than hiring, which is a sign that confidence in the durability of demand is still limited.

The easing of input cost inflation, the rate at which the prices firms pay for raw materials and labour are rising, is genuinely good news for the RBI. If firms are paying less to produce each unit, they have less reason to raise prices, which takes some pressure off consumer inflation. But the external picture complicates this. Oil prices are rising again because of the Iran tanker attacks, and India imports roughly 85% of its oil. A sustained oil price rise would push input costs back up, potentially reversing October's improvement. One month of easing costs is encouraging, but it is not a trend until it survives a few more readings.

The EU's demand is a classic burden-sharing problem. When a group of countries jointly funds a public good, like the defence of Ukraine, each member has an incentive to let others pay more and free-ride on the result. The security benefit, a stable eastern Europe, flows to all of them whether or not they contribute. This is what economists call a free-rider problem, and it is the reason NATO has spent decades arguing about the 2% of GDP defence spending target. The EU is now applying the same logic to Ukraine aid: if the UK, Canada and Japan benefit from Ukraine holding its ground, they should pay a share of the cost.

The precedent worth thinking about is the post-2022 period, when the US carried the largest share of Ukraine military aid and European governments were repeatedly criticised for doing too little. That pressure eventually produced larger European commitments, but it took two years and a change in US administration to force the shift. The difference now is that the US under Trump is less willing to be the default funder, which means the pressure on second-tier contributors like the UK is structurally higher than it was in 2022. Whether Burnham can agree to more spending while also delivering a tight budget at home is the question this summit does not answer.