Global markets
Could an AI crash happen without a financial crisis?
What the article reports
Donya-e-Eqtesad summarises a new editorial in The Economist that asks how worried the world should be about a bubble in artificial intelligence. Canals, railways and the internet all set off market booms that ended in disappointment, even though the technologies themselves proved revolutionary.
The figures it cites on the size of the boom:
- One forecast puts average annual AI investment from 2025 to 2032 at about 3.6 percent of US GDP.
- The largest AI-related stocks make up about 40 percent of the S&P 500.
- Shares are 38 percent of US household net wealth, nearly a quarter more than at the peak of the dot-com bubble.
- In the 12 months to August, net foreign buying of US shares passed 1 trillion dollars for the first time.
The article also reports signs of strain: share prices have been almost flat for six months, the market fell nearly 5.5 percent in one day after the US midterm elections, and the leading AI chipmaker's share price is about 30 percent below its peak.
The pessimists' case
Pessimists point to the 1873 railway crash, the last time investment was this large relative to GDP, which was followed by deflation and the longest recession in US history. They argue that AI model developers sit at the top of the whole supply chain, so weak sales or price pressure from cheaper Chinese rivals, said to be about six months behind, could spread to data centres and chipmakers.
Why a crash may not become a crisis
- Much AI spending goes on imported chips and equipment. The direct effect on US output would be much smaller, perhaps only three-tenths of a percent of GDP. Taiwan would be hit harder.
- Unlike in 1873, a central bank could cut rates.
- The boom is financed mainly with equity. The five largest data-centre owners have debt and lease obligations of about 30 percent of book value, or only 6 percent of market value, and their older businesses remain profitable.
The editorial's conclusion, as reported: a crash would most likely bring a mild US recession, comparable to 2001. Something worse would need a second shock at the same time, such as a bond-market panic or a war.
Why it matters for Iran
Our view. Iran's markets are largely cut off from global equity flows, so the direct link is small. The indirect channels are global risk appetite, the dollar and commodity prices. A sharp fall in technology shares that turned into a global slowdown would likely weigh on oil, metals and petrochemical prices, which matter for many of Iran's export industries.
Source
- Donya-e-Eqtesad, “A possible AI crash without a financial catastrophe?”, 13 November 2026 (in Persian), summarising an editorial in The Economist, issue of 14 to 20 November 2026. donya-e-eqtesad.com
This is our own summary and commentary on a third-party article, written for readers outside Iran. Facts and figures are as reported by the source; sections marked “Our view” are DapezSolution analysis. It is not investment advice or a recommendation to buy or sell any security.
Want briefs like this every week? Subscribe to the free weekly brief.