AI goes boom. Or boom?
Investors should understand that correctly predicting artificial intelligence (AI) technology will change the course of humanity, although for the better, does not automatically translate into desirable investment returns.
General Purpose Technologies (GPT) of the past, such as the automobile, electricity, commercial flights, steam transportation and TV, all changed the course of human history and yet more than a thousand automakers have disappeared from the US and none exist today that are profitable and have not been bailed out by government or private equity.
The hype surrounding a new technology generally lowers the investment costs for participants, and they naturally take advantage of that cheap capital to build and scale that technology in a rapid land grab. But overcapacity usually follows and the returns disappoint. Disappointing returns come about because while the “this-technology-is-going-to-change-history” Although this theme is seen as both structural and continuous, the reality is that suppliers must meet customer demand that is not structural but cyclical.
A period of creative destruction follows, and all but the early investors lose between 60 and 95 percent. Of course, nothing has changed and investors are confused because they correctly predicted the life-changing nature of the technology.
Ultimately, after the enormous losses, buyers of distressed assets can secure the technology from the desperate investors of the early part of the hype cycle, ensuring that the technology is widely and affordably distributed and that the course of human history begins to change as the technology becomes widely adopted.
The process of invention, hype, low capital costs, overcapacity, creative destruction, and distressed purchases is the process that causes technology to change the course of human history, because it is the process that puts that technology in everyone’s hands.
Today
Today, the artificial intelligence revolution is dominating Wall Street, driving unprecedented investment by participants while sparking intense debate over whether we are witnessing sustainable growth or an impending bubble.
I bet this is a bubble. We just can’t predict when it will burst. But we can predict that this will be the case.
Earlier this week, reports of the end of the US government shutdown and geopolitical tension between China and the US gave investors a new excuse to join the rally. But beneath the optimism lie my aforementioned concerns, as well as massive debt issuance by tech giants, circular deals fueling spending commitments, and high-profile warnings from the likes of Michael Burry and Ray Dalio.
“Because of the size of Berkshire and because of market levels, ideas are few – but not zero.” – Warren Buffett, November 10, 2025
Additional tension can be drawn from evidence of tech giants’ ambitious future spending commitments and the reality of insufficient current revenues amid retail investors’ fears of missing out on high valuations. But as I’ve already put it, history reminds us that every capital spending boom ends in excesses – people inevitably get overexcited, this time leading to losses of potentially trillions. As Marko Papic, BCA Research’s chief geopolitical strategist, puts it: “There will be blood and tears, and it will be terrible.”
Debts and financing shortages
The rise of artificial intelligence (AI) has led to a massive debt wave among hyperscalers. In just seven weeks, Amazon, Google, Meta, Microsoft and Oracle raised a whopping $120 billion in bonds. This reflects cash flows insufficient to cover exploding capital expenditure, which is expected to reach $534 billion next year and consume 80 percent of the group’s forecast cash flow.
OpenAI has signed deals worth more than $1.4 trillion in computing power over the next eight years, despite only projecting $20 billion in annual revenue by the end of the year. Recent announcements include a $38 billion, multi-year pact with Amazon and Microsoft’s $9.7 billion deal with Australian data center operator Iren. The circular agreements – often between OpenAI, Nvidia and other technology players – emphasize interdependence and rightly raise questions about sustainability.
Perhaps most importantly, bond traders should be wary. Debt spreads for hyperscalers have risen from 50 basis points in September to almost 80 basis points. Bank of America’s Michael Hartnett calls this an important metric to watch. Why? Because tech bonds fell eight percent in the year before the dotcom peak in March 2000.
Government backstop
Another question mark was raised this past week by OpenAI executives who raised the idea of government involvement in AI funding. CFO Sarah Friar initially suggested that the US government could “backstop” the sector to facilitate massive investment, but later clarified that this meant broad support rather than direct guarantees. CEO Sam Altman followed suit in a social media post.
Altman explicitly stated: “We do not have and do not want government guarantees for OpenAI data centers.” He argued that taxpayers should bail out failing companies and opposed picking winners. Instead, he suggested that governments build and own their own AI infrastructure, possibly with cheaper capital, to create a “strategic national reserve of computing power” that benefits the public sector.
A cynic might say that Altman and Friar were testing the waters looking for indirect help that could lower capital costs across the industry – which would benefit leaders like OpenAI. Should we speculate that the Federal Reserve’s renewed quantitative easing (QE) could lead the central bank to buy hyperscaler bonds?
Burry’s warning
Michael Burry is the famous subprime short seller portrayed in the major and subsequent film, The big short one. Burry has built up a short position of more than $1 billion on Nvidia and Palantir. He points to slowing cloud revenue growth at Amazon, Alphabet and Microsoft, along with investment levels that rival those of the dot-com era.
Burry goes further and argues that the hyperscalers overestimate a chip’s lifespan, underestimating its depreciation and therefore inflating profits. Specifically, Burry estimates an overestimation of 26.9 percent for Oracle and 20.8 percent for Meta in 2028.
A bull argument
According to some strategists, bubbles will only burst at higher interest rates, and because “the Fed isn’t going to raise and rates aren’t going to rise,” these experts believe the boom won’t burst anytime soon. They note that US interest rates are falling, allowing households to gain leverage; bond markets remain calm despite debt concerns; and the US-China trade detente has reduced the risk of a spreading trade war.
As an aside, the trade truce allows China to grow within certain limits, benefiting allies like Australia – possibly the third biggest gainer behind Mexico and Canada. With low tariffs of 10 percent and US approval for raw material exports (“selling China our rocks” to finance submarines), Australia benefits from stabilized trade.
The fate of the AI sector will be determined by investors recognizing that extreme valuations and blue sky TAMs (Total Addressable Markets) may not yet be supported by customers.
Warning signs, including concentration risks, rapid debt growth, depreciation problems and unrealistic revenue targets, may be offset by supportive fiscal and monetary policies, and economic resilience could extend the runway. It may be worth tracking debt spreads, Fed actions, AI adoption rates and geopolitical shifts, but in reality the Nasdaq index will tell you everything you need to know.
He is also the author of the best-selling investing guide to the stock market, Value.able – how to value and buy the best stocks for less than they are worth.
Roger regularly appears on television and radio, and in the press, including ABC radio and TV, The Australian and Ausbiz. View upcoming media appearances.
This post was contributed by a representative of Montgomery Investment Management Pty Limited (AFSL No. 354564). The main purpose of this message is to provide factual information and not advice about financial products. Furthermore, the information provided is not intended as a recommendation or opinion about any financial product. However, any comments and statements of opinion should contain general advice only, prepared without taking into account your personal objectives, financial circumstances or needs. Therefore, before acting on any information provided, you should always consider its suitability in the light of your personal objectives, financial circumstances and needs and, if necessary, seek independent advice from a financial advisor before making any decision. Personal advice is expressly excluded in this message.
#boom #boom