Israel votes, China's AI rout deepens, and UK prices break cover

Three years of war, a hostage crisis, and a genocide inquiry at the UN have all fed into one ballot. The political science concept that fits here is what scholars call a valence election, one where voters judge a leader not on ideology but on competence and events on their watch. Netanyahu has governed Israel longer than anyone, but the October 7 attacks happened under his watch, and polling suggests many Israelis hold him responsible for the intelligence failure that allowed them.

The economic read is less obvious but real. Israel's economy contracted sharply in late 2023 and has recovered unevenly, with defence spending crowding out, taking the place of, investment in other areas. A change of government would not immediately fix that, but it could change the calculus on Gaza, where a faster end to the conflict would reduce the fiscal drag, the weight of war costs on the budget, and make a normalisation deal with Saudi Arabia more plausible. That deal, if it ever happens, would unlock investment flows and regional trade links that neither side has been able to access. Whether a new coalition could actually hold together long enough to pursue it is the question I genuinely cannot answer from today's result alone.

The number that does the explaining here is 40%. A stock index falling by that much in four months means investors have collectively decided that the expected future profits of China's technology sector are worth far less than they thought in June. In economics, a share price is essentially a bet on future earnings discounted back to today. When that price collapses, it usually means one of two things: either the earnings forecast has been cut, or the rate used to discount them has risen, or both. Here, both are happening. Profits are not materialising fast enough, and the risk of investing in Chinese tech has risen because of US chip restrictions.

The US export controls on advanced semiconductors, the chips that power large AI models, are acting as a supply-side constraint, a limit on what Chinese firms can produce regardless of how much money they spend. That is different from a demand problem. Chinese AI companies are not short of customers or government backing. They are short of the specific hardware needed to train the most powerful models. The parallel with the Soviet Union's technology gap in the 1980s is tempting, but it breaks down quickly: China's private sector is far more dynamic, and some Chinese firms are finding ways to work around the restrictions with older chip designs. Whether that workaround is enough to stay competitive is the question the market is currently pricing as: probably not.

Retailers have been sitting on a cost increase for months without passing it on. That is unusual behaviour, and it has a specific explanation: in a competitive market, the first firm to raise prices loses customers to rivals who have not yet moved. So everyone waits. When enough firms decide they cannot wait any longer, prices jump across the sector at roughly the same time. That is what the British Retail Consortium is describing when it says the dam is breaking. In economics, this delayed pass-through is one reason inflation can look contained and then accelerate quickly.

The timing is the part that makes this genuinely awkward for the government. A cost-push inflation episode, where prices rise because firms' costs have gone up rather than because consumers are spending more, is hard for a Budget to fix. Spending more public money into an economy where prices are already rising risks making things worse. Cutting taxes might help households in the short run but adds to borrowing, which pushes up government bond yields, the interest rate the government pays on its debt. The Bank of England is caught in the same bind: the case for cutting rates to support growth runs directly against the case for holding them to keep inflation down. One month of retail data is not proof of a sustained surge, but the direction is clear enough that the Chancellor cannot ignore it tomorrow.

India's situation right now is a good illustration of the gap between real-economy fundamentals and financial-market sentiment. GDP growth is solid, domestic manufacturing is expanding, and the long-run story has not changed. But none of that stops foreign investors from pulling money out when global risk appetite falls. The mechanism is straightforward: when investors worldwide become more cautious, they move money from higher-risk emerging markets into safer assets like US Treasury bonds. India gets caught in that broad move even if nothing specific has gone wrong in India.

The margin call dynamic makes the sell-off sharper than the fundamentals alone would justify. A margin call happens when an investor who borrowed money to buy shares sees those shares fall in value, and the lender demands they put up more cash or sell the shares to repay the loan. Forced selling pushes prices down further, which triggers more margin calls, which causes more selling. This feedback loop is why small-cap stocks, which are less liquid and more often bought with borrowed money, fall faster than large-caps in a risk-off episode. For the RBI, the dilemma is real: cutting rates would support growth but might accelerate the rupee's fall; holding rates steady protects the currency but slows the economy. That trade-off is exactly what the October policy meeting will have to navigate.