America’s AI Bubble Is Colliding With Its Debt Crisis

Welcome to The Geopolity’s What We’re Watching (3W), our daily look at the interconnected worlds of Geopolitics, Economics and Energy. Curated from the world’s leading sources of information, our analysis and commentary is designed to help you make sense of the events driving the major developments in the world.

The US – Israel Alliance War on Iran has driven government bond yields up globally, writes The Financial Times. As a result, the cost of borrowing for the world’s biggest developed economies is now significantly higher than before the war, by tens of billions of dollars. Just the G7 nations have paid an additional $16 billion in sovereign debt financing costs. If the rise in yields persists then they could pay an estimated additional $34 billion in extra financing costs by the end of the first quarter of next year. The US is paying the lion’s share of the increase in costs to date, an estimated $10.6 billion. It is set to pay a further $21.7 billion in additional interest costs if the rise in yields persists until the end of the first quarter of 2027. Countries such as the UK, Italy, Germany and Japan that are big energy importers also affected, driven by rising inflation expectations in the wake of the energy supply crisis triggered by the closure of the Strait of Hormuz. While the increases are relatively small in comparison to overall public spending commitments, they threaten to pile additional pressure on already stretched government balance sheets, FT notes.

As to the corporate side of the economy, more and more people are beginning to worry that AI has created a bubble. US cumulative AI-related investment is estimated to reach  $5,500 billion by 2030, but “hyperscaler” operating cash flow and general equity issuance are expected to cover only around a quarter of this investment, writes investment bank JP Morgan. The hyperscalers are the US technology companies such as Amazon, Microsoft, Oracle, NVIDIA etc. The “unusually large” gap will have to be financed by borrowing, for which reason the interest rate on hyperscaler debt has trended upward. While JP Morgan is generally optimistic this will not cause any problems, and instead presents the hyperscalers financing needs an important investment opportunity, others are not so convinced.

more and more people are beginning to worry that AI has created a bubble

The Bank of International Settlement (BIS) in its Annual Economic Report warned that the if in the future the returns on AI investment were to disappoint, the debt load that the hyperscalers are taking on now could become highly problematic, writes Reuters. This would not only create financial problems for the hyperscalers themselves, but due to the size of these companies, for the entire US financial market, BIS says.

Again, other are more pessimistic about AI, arguing that the risks are not just associated with possible future developments, but also the result of past decisions. Michael Burry, the man who famously became rich by seeing the US financial crisis of 2007 and 2008 coming, believes the profitability of the hyperscalers is already inflated because they “circulate” money among themselves, writes Yahoo Finance. The hyperscalers are borrowing money, they provide this money as loans or investment to other companies in the AI ecosystem, who then turn around and use this money to buy equipment or services from the hyperscalers. Both then report “earnings” from the deals, which enables the hyperscalers to borrow even more, and do the same thing again. This, Burry notes, is not a sustainable business model. In fact, it is a model that in the past always ends up in a crash, being essentially what is referred to as a “Ponzi Scheme” – once the lenders stop lending the business model implodes. The companies at the center of the AI ecosystem, in particular NVIDIA, reject Burry’s accusation, arguing that future revenues will increase rapidly and enable servicing of the debt.

But, 3W notes, the fact remains that the hyperscalers officially report debts that are significantly lower than the actual debts they have incurred over the past few years. According to the above mentioned BIS report, the five largest hyperscalers are carrying roughly $1.65 trillion in off-balance-sheet debt through special purpose vehicles and off-balance sheet arrangements, in addition to the $1.35 trillion in debts they officially report.

What this all means is that the BIS fears an “AI bubble” will be created in the economy, while Burry warns this bubble has already been created.

The European Central Bank has looked at what would happen if indeed there was an AI bubble, and this bubble would “pop”. A sharp stock market correction would result, it warns, and thus would have a significant negative effect on the entire economy. And, it notes that although the hyperscalers are all US companies, because EU entities are heavily invested in these companies, “the effects of a US correction could extend beyond financial markets to euro area sentiment, financing conditions and hiring. A US AI fallout would not remain a US problem”. In other words, just as happened 2007 and 2008, a crisis on the US financial market will quickly become a global financial crisis.

Other risks associated with the AI boom have also been identified, in particular a so-called “crowding out effect”. There is a only a limited amount of money available for investment, which means that as the world invests more in AI, it will inevitably invest less in other sectors of the economy, notes The Financial Times.

people who under normal circumstances would buy US Treasuries, and thereby finance the US government budget deficit, are now more and more buying hyperscaler bonds

And this brings 3W back to the starting point of this analysis of the global economy, as Fortune writes that the debt issuances by hyperscalers is drawing money away from US Treasuries. In other words, people who under normal circumstances would buy US Treasuries, and thereby finance the US government budget deficit, are now more and more buying hyperscaler bonds. This is forcing the US government to offer higher interest rates on its US Treasury bonds – which increases the US’s cost of borrowing, reducing its ability to spend money on the other things US society needs.

In the 3W view, this is a fundamental weakness in the US geopolitical position, and by extension the broader “West”. Government and corporate debt levels are unsustainably high. As a result, as we noted in our update “America’s Iran War Is Colliding With Its Debt Crisis”, any disruption to the normal functioning of the West’s financial markets can cause a major financial crisis. The failure of the War on Iran has caused a number of such disruptions, the fallout of which the US now actively needs to manage. This distracts it, and limits its ability to “project power” internationally to secure its interests (as defined by the Zionists).

This is also the opinion of Andrew Bailey, governor of the Bank of England and chairman of the Financial Stability Board international watchdog (FSB). Referring to the disruptions caused by the War on Iran, Bailey issued a warning to the G20 finance ministers that a combination of highly priced stock markets, driven by AI; increased borrowing by investors, in particular AI hyperscalers; and the growing concentration of money into a small number of major technology companies, mostly in the AI ecosystem; could “amplify any future market correction”, writes BBC. In plain language that means, “if AI collapses it will take everything else with it”, notes 3W.