1st October 2026
The Bank of England has put a striking figure on one of the risks building in global financial markets: a 45% fall in US share prices.
It is important to be clear about what the Bank is saying. It is not predicting that stock markets will fall by 45%, nor is it saying that an AI crash is inevitable.
Instead, the Bank has constructed a hypothetical scenario to examine what could happen if investors suddenly decided that the enormous profits and productivity gains expected from artificial intelligence were unlikely to materialise as quickly as anticipated.
And the consequences for Britain could be considerable.
The Bank's July Financial Stability Report says AI-related companies have become an increasingly important driver of global equity markets. In the US, AI-related companies account for around half of the S&P 500's market capitalisation, while the index itself represents about half of global equity capitalisation.
Britain is less directly exposed because the FTSE 100 contains fewer AI companies. But that does not provide immunity.
The Bank's scenario assumes a 45% fall in the US equity market over six quarters, accompanied by a substantial rise in corporate borrowing costs and a weaker dollar.
Its modelling suggests that the shock could eventually knock 2.2 percentage points off UK GDP, with tighter credit conditions playing an important role. Around half of the estimated UK impact comes through higher corporate credit spreads, while roughly 36% comes through the equity-market channel.
For ordinary households, the first impact might be difficult to see.
Someone without a large portfolio of American technology shares might reasonably ask: why should a fall in Nvidia, Microsoft or other technology companies matter to them?
The answer is that modern financial markets are interconnected.
UK pension funds, investment funds and other institutions have investments in global markets. Many households also have indirect exposure through pensions and ISAs.
A major market fall could therefore reduce the value of savings and pension investments.
But the bigger concern could be what happens next.
If investors become more nervous, companies can find it more expensive to borrow. Banks and other financial institutions can face higher funding costs. Businesses may postpone investment, recruitment and expansion.
Consumers may also become more cautious if the value of their pensions, investments or homes comes under pressure.
And if the shock is severe enough, weaker investment and spending can feed into employment and economic growth.
This is where the Bank's warning becomes more interesting for Britain.
The UK is not simply vulnerable because of its stock market. It is vulnerable because Britain is part of a global financial system in which American markets have enormous influence.
There is another complication. The Bank says leverage in equity markets has increased substantially, including among hedge funds and leveraged exchange-traded funds. That matters because borrowed money can turn an ordinary market correction into something much more violent as investors are forced to sell.
The Bank also points to the rapidly increasing use of debt to finance AI infrastructure.
That creates a potential feedback loop. If investors lose confidence in future AI profits, share prices could fall. Companies could then find borrowing more difficult and expensive, while lenders and investors could become more cautious.
There is, however, an important reason not to assume another 2008-style financial crisis.
The Bank says its stress testing indicates that Britain's banks would remain resilient under shocks of this scale. It also says an equity-market shock on its own would be unlikely to threaten UK financial stability.
So the issue is not simply whether AI shares fall.
It is whether a fall in AI valuations becomes a much broader loss of confidence that spreads into credit markets, business investment, household spending and government borrowing costs.
That is why the Bank is watching the AI boom so closely.
AI could eventually deliver enormous productivity gains and make economies more efficient. But today's share prices already contain very large expectations about tomorrow's profits.
If those expectations prove too optimistic, the adjustment could be painful.
For Britain, the lesson is perhaps less about whether there will be an "AI crash" and more about how dependent the modern economy has become on financial markets behaving normally.
A 45% fall is only a scenario.
But the Bank of England is effectively asking a more uncomfortable question: what happens to Britain if the financial markets funding today's economic optimism suddenly decide tomorrow will not be quite as profitable as expected?