‘It is impossible to know when a bubble will burst, but it is possible to know when a bubble exists.’ – Robert Shiller
Cracks in Global Foundation
There is something increasingly unsettling about the global economy. The problem is no longer simply that markets are overvalued, currencies are under pressure, or governments are intervening to prevent financial instability. The deeper problem is that the architecture holding the global financial system together increasingly depends on confidence in assets whose underlying economic foundations are becoming harder to defend.
AI companies have become the newest and most powerful expression of this phenomenon. Companies with enormous valuations are raising extraordinary amounts of capital on the expectation that future AI revenues will justify today’s prices. The valuation creates the capacity to raise more money; the new money validates the valuation; and the higher valuation reinforces the belief in future earnings. It is a financial circle. And history has shown us how dangerous such circles can become.
Before examining the AI sector, it is useful to consider the tectonic shifts occurring elsewhere in the global economy. Recent weeks have seen several notable fractures. Japan’s currency crisis and the ongoing devaluation of the Yen have prompted aggressive intervention by the United States government by selling the Euro. Simultaneously, South Korea’s stock market experienced a sharp correction, and over $1 trillion was temporarily wiped from U.S. chip stocks. While these events are concerning, they may also reflect normal volatility in a post-pandemic environment characterised by shifting interest rates and supply chain realignments.
The response to these crises, however, has raised eyebrows. The United States’ intervention to support the Yen represents exactly the kind of currency management that Washington has historically criticised Beijing for. The Treasury’s ‘signalling’ tactics, including the widely discussed notes associated with Treasury Secretary Scott Bessent, underscore a heavy-handed approach to managing exchange rates. Bessent, a veteran who made his fortune alongside George Soros, is recognised as one of the world’s most skilled practitioners in currency markets. While his actions may be viewed by some as pragmatic crisis management, others argue they highlight a broader trend of bending the established rules of global economic engagement.
For decades, the United States has accused China of manipulating its currency to gain an unfair economic advantage. Yet when a major economy finds its currency under severe pressure, direct intervention, interest-rate policy and carefully calibrated official signals can become tools for influencing exchange rates. The principle is suddenly less abstract when the stability of the global financial system is at stake.
The same contradiction can be seen in global energy markets. Oil, arguably the world’s most important strategic commodity, is not determined by supply and demand alone. It is shaped by geopolitics, sanctions, production agreements, strategic reserves, shipping routes and financial markets. Governments intervene. Producers intervene. Traders speculate. Great powers influence the rules. And increasingly, the line between markets and political power is becoming impossible to ignore.
The Last Un-Rotten Apple?
The hyper-valuation of companies like OpenAI and their peers is playing an outsized role in propping up the US market, which in turn holds a significant portion of the world’s foreign savings. These savings are critical for financing U.S. Treasury auctions. In essence, the AI sector has become a key structural pillar of the American financial system. The concern is that this pillar may be built on a foundation that requires careful scrutiny.
Many of these companies are not yet generating substantial operational profits; instead, they are raising capital through successive fundraising rounds, each built upon optimistic valuation narratives. OpenAI, for instance, relies heavily on new investment to sustain its operations and massive infrastructure costs. The market is currently pricing in over $2.1 trillion in projected future earnings from these firms-an extraordinary figure for companies that are, for the most part, still operating at a loss.
For decades, the dominant economic philosophy was that the private sector could do almost everything better than the state. Public enterprises were privatised. Public utilities were commercialised. Infrastructure became an investment class. Education became a market. Healthcare became an industry. Housing became a financial asset.
The argument was that private capital would create efficiency. But what happens when almost everything becomes an asset? Eventually, there are fewer new places for capital to go. And that may explain part of the extraordinary enthusiasm surrounding artificial intelligence. AI has become more than a technology. It has become a financial destination for global capital searching for the next great source of growth. This is where the danger begins.
The Catalyst for a Broader Slowdown?
The systemic risk, if it materialises, lies in the interconnectedness of these markets. If the AI sector experiences a significant correction, it could trigger a sell-off in equities. If foreign investors, seeing their savings impacted, were to pull back from U.S. bonds, yields could spike, increasing the cost of servicing the national debt. Unlike 2008, there is no guarantee that the bond market would serve as a safe harbour this time around, given the current levels of global debt and geopolitical fragmentation.
Scott Bessent and his peers are experts at navigating financial crises, and they are likely to continue using aggressive signalling and market intervention to manage volatility. Their goal is to avert a severe economic contraction-a recession that could be more significant than recent downturns. The manipulation of the AI narrative, combined with currency and oil market interventions, can be seen as calculated attempts to buy time and maintain stability.
This is the uncomfortable paradox of our age. We have built a global economy in which governments are increasingly dependent on markets, markets are increasingly dependent on expectations, expectations are increasingly dependent on technology, and technology companies are increasingly dependent on capital continuing to believe in their future.
The greatest mistake would be to assume that because AI is genuinely revolutionary, every valuation attached to AI must therefore be justified. History does not work that way. Railways changed the world. The internet changed the world. Electricity changed the world. Housing remains indispensable. Yet the existence of a transformative technology or essential asset does not prevent financial bubbles from forming around it.
Financial crises rarely begin when people discover that the future is impossible. They begin when people suddenly realise that the future they have already priced into everything was never guaranteed.
The world may therefore be entering a period in which the greatest economic battle is no longer between capitalism and socialism, or between markets and governments. It is between financial expectations and economic reality. And if reality eventually wins, the adjustment could be global. What is unsustainable cannot be sustained.