AI Bubble About to Burst? Bank of England Warning

Artificial Intelligence, Debt, and Market Risk: A Breakdown of the Bank of England’s Financial Stability Report

Key Concepts:

• Risk Premium: The return above the risk-free rate received on risky assets, currently compressed indicating high valuations.
• Magnificent 7 (MAG 7): The seven largest US tech companies (Alphabet, Apple, Amazon, Meta, Microsoft, Nvidia, and Tesla) dominating the S&P 500.
• Hyperscaling: Rapid and large-scale expansion of AI infrastructure and operations.
• Investment Grade Bonds: Bonds with a relatively low risk of default, issued by companies with strong credit ratings.
• Private Credit: Lending from non-bank financial institutions, often with less regulatory oversight.
• Wealth Effect: The tendency of consumers to spend more when their wealth increases (e.g., due to rising stock prices).
• Correlation Tightening: A phenomenon where asset classes become more correlated during market stress, leading to widespread declines.
• Contagion: The spread of financial distress from one institution or market to others.

I. AI Euphoria and Market Valuations

The Bank of England (BoE) has issued a warning regarding the current valuations in the stock market, specifically linking the AI-driven surge to the dot-com bubble. Analysis of risk premiums across the US, Europe, and the UK reveals that prices are significantly high relative to fundamentals, with many risk premiums falling below their 10th percentile. This indicates reduced future expected returns. The concentration of the S&P 500 in a small number of AI-focused tech companies – the Magnificent 7 – now comprising over a third of the index, exacerbates this risk. These high valuations are predicated on future earnings from AI that are not yet realized, creating vulnerability to a correction if AI monetization is slower than anticipated or if technological bottlenecks emerge. As stated, the market is “priced for perfection and any bad news could trigger a sharp correction.”

II. The $5 Trillion AI Debt Bomb

Beyond the stock market, the BoE highlights a significant, often overlooked, risk: the debt used to fund AI’s rapid growth. Estimates suggest this could reach $5 trillion over the next five years, as reported by JP Morgan. While organic cash flow from large tech companies and investment-grade bonds will contribute to funding, a substantial $1 trillion+ gap remains. This gap may be filled by private credit or potentially government intervention. The reliance on private credit is particularly concerning, as these opaque markets haven’t been stress-tested at this scale, especially in a high-interest rate environment. Since 2008, debt issued by five major US tech companies has surged, with Alphabet issuing $7 billion in Euros and $18 billion in US debt in a single day in November, increasing its debt by nearly 70%. JP Morgan’s report details the expected funding sources: organic cash flow and investment grade bonds will cover a significant portion, leaving a substantial gap.

III. Potential Triggers for a Market Correction

Several factors could trigger a correction related to this AI-fueled debt bubble:

• Bond Maturities: Bonds have a fixed maturity date. If AI revenue doesn’t grow sufficiently to service the debt, refinancing will be difficult and expensive.
• Slower AI Improvement/Monetization: If AI models don’t improve as predicted by scaling laws or if companies struggle to convert users into paying customers, earnings forecasts could collapse.
• Power Constraints: Connecting new data centers to the electricity grid is becoming a major obstacle. Without sufficient power, AI projects cannot generate revenue. OpenAI alone has signed infrastructure deals totaling over $1.4 trillion to increase power production by 28 gigawatts over the next 8 years. Sam Altman described the power crunch as “existential,” stating, “a certain risk is if we don’t have the compute we will not be able to generate the revenue or make the models at this kind of scale.”
• Geopolitical Disruptions: The AI supply chain is globally distributed, relying on components and materials from various countries. Trade fragmentation could disrupt this chain, hindering AI development. For example, potential US tariffs on China could lead to retaliatory export restrictions on rare earths.

IV. Implications for the UK Economy

The BoE identifies four potential “dominoes” that could impact the UK:

1. Reverse Wealth Effect: Rising US stock prices have contributed to 10% of US consumption growth since the start of 2024. A sharp correction in AI stocks could erase this wealth, leading to reduced consumer spending and impacting global demand.
2. Sentiment and Correlation: A crash in a major sector like AI could trigger widespread panic and “correlation tightening,” causing even markets with low AI exposure (like the UK’s FTSE 100) to fall.
3. Hidden Leverage in Private Credit: Losses in private credit loans to AI projects could spread to systemic banks, tightening lending standards and freezing credit markets. The BoE is actively exploring these risks through systemwide exploratory scenarios and bank capital stress tests.
4. Commodity Market Shock: A crash in AI demand could cause a significant shock to energy and commodity prices, potentially triggering liquidity crises in commodity markets, similar to the nickel crisis on the London Metal Exchange in 2022. The UK is particularly exposed due to the listing of many mining and energy companies on its exchanges.

V. Bank of England’s Perspective and Conclusion

The BoE acknowledges AI’s potential to boost productivity but warns that the financial speculation surrounding it has become dangerous. The current environment is characterized by stretched valuations, ballooning debt, and a shift towards a “show me the money” phase. If earnings don’t materialize, a painful correction is likely. The BoE is actively monitoring these risks and attempting to implement measures to prevent contagion.

In conclusion, the Bank of England’s assessment paints a picture of significant risk associated with the current AI boom, extending beyond stock market valuations to encompass a substantial debt burden and potential systemic vulnerabilities. The report emphasizes the need for caution and highlights the potential for a sharp correction if expectations regarding AI’s future performance are not met.