It seems you can’t have a conversation about technology these days without someone mentioning AI. It’s the new gold rush, a seemingly endless frontier of innovation and potential profit. But whilst Silicon Valley is busy selling shovels in the form of GPUs and cloud services, some very sharp people on Wall Street are starting to worry about the stability of the mines themselves. And right now, all eyes are on a rather telling indicator from a tech behemoth: Oracle.
Let’s be clear, when the cost to insure against a company like Oracle defaulting on its debt triples in a few months, it’s not just a minor tremor. It’s a signal flare. The market is getting jittery, and the conversation is quietly shifting from “How high can AI go?” to “How hard could it fall?”. We need to talk about the growing spectre of AI bubble risks.
Those Who Forget History…
Deja Vu, All Over Again?
So, what exactly is an AI bubble? Imagine it’s 1999. Everyone is piling money into any company with “.com” in its name, conviction soaring that the internet would change everything. They were right about the internet, of course, but dead wrong about most of those companies. An investment bubble happens when asset prices detach from their fundamental value, driven by hype and a fear of missing out, rather than solid business plans and profits.
Sound familiar? Today, “.com” has been replaced by “AI”. The rush to build, buy, and integrate artificial intelligence is creating a dizzying level of investment. The problem is that many of these investments are based on the promise of future returns, not current ones. This is precisely where the parallels to the dot-com craze become uncomfortable, and where the risk of a painful correction grows.
The Canary in the Coal Mine: Credit Derivatives
For those who don’t spend their days glued to financial terminals, credit derivatives might sound like jargon. But they are a fascinating and crucial piece of this puzzle. Think of a specific type, a credit default swap (CDS), as an insurance policy. Investors who own a company’s bonds can buy a CDS to protect themselves if that company fails to pay back its debt.
The price of this “insurance” tells you how risky the market perceives that company to be. As reported by Yahoo Finance, Oracle’s CDS price soared to 1.28 percentage points, its highest level since the financial crisis in 2009. To put that in perspective, it was just 0.36 points in June. This surge suggests that big-money investors are suddenly very concerned about the company’s ability to handle its debts amidst this AI spending spree, contributing to significant market volatility.
The Mountain of Debt Powering the AI Dream
Building the Future on Borrowed Money
So, what’s spooking investors? It’s the sheer amount of money being borrowed to fund the AI revolution. Building the massive data centres required for today’s large language models is eye-wateringly expensive. Companies are taking on enormous amounts of tech debt to finance this construction.
Oracle is a prime example. The company’s total debt has ballooned to a staggering $105 billion. A significant portion of this is going towards building out the cloud infrastructure needed to support AI clients, including its high-profile partnership with OpenAI.
It’s a colossal gamble. Oracle is betting that the revenue from these AI services will eventually be more than enough to service this mountain of debt. But what if the returns are slower to materialise than expected? What if a competitor offers a better, cheaper service? That’s the question causing sleepless nights, not just in Oracle’s boardroom, but for its creditors too.
The Search for Investment Safeguards
This isn’t to say that investing in AI is foolish. The technology’s potential is undeniable. However, the current climate calls for robust investment safeguards. For companies, this means a ruthless focus on the path to profitability for their AI projects. For investors, it means looking beyond the hype.
– Diversification is key; don’t put all your chips on one AI play.
– Scrutinise the balance sheet. How much debt is a company taking on to fund its AI ambitions?
– Is there a clear, believable business model, or is it just a story about future domination?
The smart money isn’t just buying insurance; it’s asking the tough questions before placing its bets.
What the Experts are Saying
The rumblings aren’t just coming from obscure market signals. Financial analysts are openly discussing the sustainability of this trend. Hans Mikkelsen, a strategist at TD Securities, noted the “impressive leverage growth” in the tech sector, raising questions about whether it can continue. The sentiment was echoed by Jigar Patel at Barclays, who pointed out that the bond market is becoming more sceptical about these massive debt-fuelled expansions.
This isn’t just theory. Oracle’s recent $18 billion bond sale saw a surge in CDS trading volume, jumping from around $350 million a year prior to $8 billion over just nine weeks. Investors aren’t just saying they’re nervous; they are putting billions of dollars on the line to hedge against a potential downturn.
Looking ahead, the borrowing isn’t set to slow down. TD Securities forecasts that US investment-grade bond sales could hit $2.1 trillion in 2026. A huge slice of that will be for tech companies funding their AI arms race. This suggests the market volatility and anxiety we’re seeing now could become the new normal.
The AI revolution is happening, and it will reshape our world. But revolutions are messy, chaotic, and often financially ruinous for those who get swept up in the initial speculative mania. The signals from Oracle’s credit risk are a stark warning that whilst the technological promises are grand, the financial foundations are looking increasingly shaky. The biggest AI bubble risks may not come from the technology failing, but from the simple, old-fashioned problem of borrowing too much, too fast.
Are we simply witnessing the necessary, debt-fuelled construction phase of a new technological age, or are we just building prettier dot-com castles in the sky? What are your thoughts?
