Bank of England Governor Andrew Bailey said on Monday that the risk of an AI asset-price bubble should be recognised. “Bubble spotting is an imprecise science. But, let’s recognise that such a risk exists, which we should do”, Bailey said in a speech in Zurich published by the Bank on 9 November. The Governor then set out four ways such a bubble could burst, including “everyone is priced to be a winner”.
The Bank’s own Financial Policy Committee record of 30 September puts the debt behind that build-out in numbers. It cites a JP Morgan estimate of about $4.1 trillion of debt-financed AI capital expenditure between 2026 and 2030. That is about $820 billion a year, by our arithmetic, or 1.8 times the roughly $450 billion of AI-related debt Morgan Stanley estimated had been issued by early September.
Business Pill · Financial Conditions
An evergreen Business Engineer explainer on financial conditions. It teaches the general idea only and does not discuss the Bank of England or this speech.
The key insight: As we read it, Bailey’s speech moves the AI-bubble question from market chatter into a Governor’s published text: the risk “exists”, and there are four ways it could burst. The Bank’s own FPC record supplies the scale, citing a $4.1 trillion estimate of debt-financed AI capex for 2026-2030, about $820 billion a year by our arithmetic.
Bailey opened with the speculation: “There is a lot of speculation on the subject of whether there is an asset pricebubble in AI, and the financial stability risks that go with that”. Bailey then asked “How might the bubble burst and potentially cause a problem?” and listed four arguments.
First, “The returns from AI may be good, but still markets may overvalue them”. Second, returns may match expectations but take longer to arrive than predicted, as with “Edison and electricity”. Third, “everyone is priced to be a winner, but history tends to tell us that this is not the case. Remember Netscape”.
Fourth, public trust. Bailey said the risk of a loss of public trust in AI “could affect asset valuations” until it is solved. Bailey concluded: “So, there are clear financial stability risks here, and that’s without talking explicitly about cyber risks”.
Posts on X summarised the speech as a warning of a market crash. The published text does not use the word “crash”. Bailey also said “I am an optimist on AI” and that AI investment explains some of the UK economy’s resilience.
AI-related debt estimates by Morgan Stanley and JP Morgan, as cited in the Bank of England’s September 2026 FPC record. The $820 billion yearly average is our arithmetic ($4.1 trillion over 2026-2030).
The Debt Figures in the Bank’s Record
The September FPC record says an increasing volume of AI-related investment is being financed through debt issuance, with global issuance in 2026 expected to exceed that of countries such as the UK. It cites Morgan Stanley’s estimate of around $450 billion of global AI-related debt issuance as of early September, “more than double the total issuance in all of 2025”.
It also cites JP Morgan analysts’ estimate of around $4.1 trillion of AI-related capital expenditure financed through debt between 2026 and 2030, and Morgan Stanley analysts’ estimate that $700 billion of data centre capital expenditure between 2026 and 2028 would be financed by private credit. In sterling credit markets, AI hyperscalers accounted for 47% of corporate bond issuance in 2026 up to the time of the record, it says.
The record warns that “increasing leverage, opacity and, at times, ‘circular arrangements’” associated with this financing could complicate the assessment of risks and could amplify losses if expectations disappointed.
What the Committee Said About Valuations
The same record says AI-related and semiconductor stocks fell sharply in July, and that some leveraged investors were forced to unwind positions, with no spillovers to core markets. “The risk of a sharper correction with spillovers to core markets persisted”, the committee said.
It adds that growth prospects and fiscal outlooks depend in part on expected AI productivity gains, so a reassessment “could therefore affect not only AI-related asset valuations but also sovereign debt markets”.
As we read it, two of Bailey’s four arguments are about timing and dispersion rather than the size of AI returns: returns may arrive later than markets expect, and not every company priced as a winner will be one. The September record adds the channel: debt, with leverage and opacity that it says could amplify losses if expectations disappointed.
Andrew Bailey, Governor, Bank of England, speech in Zurich, 9 November 2026
“Third, everyone is priced to be a winner, but history tends to tell us that this is not the case.”
Three Implications
FOR AI BORROWERS The September FPC record says AI-related debt issuance of around $450 billion by early September was more than double all of 2025, citing Morgan Stanley.
FOR UK CREDIT MARKETS The record says AI hyperscalers accounted for 47% of sterling corporate bond issuance in 2026 up to that point, though their sterling issuance remained significantly smaller than in the US and euro area.
FOR SOVEREIGN DEBT The record says a reassessment of expected AI productivity gains could affect not only AI-related asset valuations but also sovereign debt markets.
The framework describes AI’s dual character as bubble and supercycle at once and argues: “The mistake is treating them as competing claims.”
As we read it, Bailey’s speech holds both sides too: it says “I am an optimist on AI” and also that the bubble risk exists. The framework predates this speech and does not discuss the Bank of England.
What Is Not Established
Bailey’s remarks were opening comments to prompt discussion, according to the speech text, not a policy decision. The speech does not say whether Bailey thinks AI stocks are in a bubble, and gives no valuation measure. The debt figures are bank analysts’ estimates cited by the FPC, not Bank of England forecasts.
This Week In The AI Supercycle
The Great Convergence of Capital, Power, and Geopolitics
This week’s issue: for two years the AI build-out was priced as a supply story, and last week as a credit story. This week it was priced as a permission story, and the answer arrived as a cost, not a ban.
Bank of England Governor Andrew Bailey said on 9 November that the risk of an AI bubble exists and named four ways it could burst. The estimates in the Bank’s September FPC record put debt-financed AI capital spending at about $820 billion a year to 2030, by our arithmetic, 1.8 times the AI debt issued by early September 2026.
A note on sourcing. Bailey’s words are quoted from the speech text published on bankofengland.co.uk on 9 November. Debt and market figures come from the Bank’s Financial Policy Committee record of 30 September, which cites Morgan Stanley and JP Morgan estimates. The $820 billion yearly average and the 1.8x comparison are our arithmetic.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
