Summary: AI Spending & The Dot-Com Bubble
The world’s biggest companies are making huge bets on AI. Spending on AI infrastructure has been a driving force of positive sentiment throughout the current bull market. Among others, Amazon, Microsoft and Alphabet are dipping into their hordes of cash to fund logistics, data-centre construction and employee upskilling.
Concerns are emerging, most recently surrounding news of OpenAI’s cash burn, which has been revised to an expected $115 billion by the end of 2029 (they previously anticipated cash burn of $35 billion over the same time period). We know the costs that these companies are incurring, but whether this will drive future profitability is in question.
Whenever money is being spent in large quantities in ways that feel indiscriminate, the word “bubble” will inevitably be tossed around. And the greatest bubble of modern times formed in the 1990s and burst in the new millennium.
Is AI fervour the dot-com bubble repeating itself?
What’s Different? These Companies have to do something with their cash
Microsoft and Amazon each have more than $90 billion on their balance sheets. Apple has $55 billion. Their cash hoards would be the envy of many small countries. When companies are this large and lucrative, deploying cash towards profitable endeavours becomes a challenge. Is it better to see this money put back into the real economy, or used to purchase treasuries yielding 2%? When people express concern about the share of capital expenditures being directed toward AI capacity, it makes me wonder what they would prefer these companies spend money on.
Contrast this with the companies that fuelled the bubble in the late 1990s: they were generally pre-profit (often pre-revenue) and financed completely by debt.
The emergent concern is the growing use of debt by Big Tech in pursuit of AI dominance. The dollars required to meet the computing power demands of AI are immense: some estimates assume that by 2030 as much as $5.2 trillion will be needed for data-centre investment. Private lenders are stepping in to meet this demand for cash. New debt instruments, new lenders and relaxed terms is reminiscent of not only the dot-com bubble, but also the 2008 crash.
Similarity 1: It is Celebrated & Feared in Equal Measure
There is irony in progress. We complain about how things never change, but fear any potential catalysts of change. AI is no exception. Concerns range from spiking unemployment to the subjugation of humanity depending on which sci-fi materials are being used as a reference point. A quarter century ago, a meaningful portion of the population thought the world would end because our internet-dependent global economy wouldn’t be able to handle the logistical challenge of moving into a new millennium.
There is a tendency to focus on what aspects of the status quo will be uprooted by progress, with little consideration for the new frontiers that will be unlocked. It reminds me of the famous quote attributed to Henry Ford: “If I had asked people what they wanted, they would have said faster horses.”
As the 20th century dawned, British lamplighters were put out of work as electricity proliferated. It’s hard to argue this was a net negative to society. The shape of the economy is in a constant state of change and the most impactful changes are driven by technological breakthroughs.
Similarity 2: Immediate Impacts are Probably Being Overblown
I’d be happy to be proven wrong on this one. There are some pressing issues that it would be great to see addressed by technological progress. How many lives will be saved by enhanced diagnostic accuracy in cancer screenings? How many tax dollars can be more effectively deployed by reducing the administrative burden in the healthcare system?
Technological breakthroughs still operate within the constraints of physical infrastructure. The computing power of AI requires enormous amounts of energy, in addition to the aforementioned cost of data-centre investment. The invention of the steam engine was great, but it wasn’t until rail lines had been laid that its merit could be truly appreciated.
Difference: Investors of Today can Take Heed from Lessons During the Dot-Com Bubble…
I’ll caveat this by saying that I think human nature is enduring. The most valuable lessons we can draw from the early days of the stock market are those rooted in human behaviour. The stock market itself is a relatively new concept in the timeline of human history, and widespread ownership of equities dates back only a few decades. For many, the dot-com bubble was their first foray into equities, and the most painful and effective way to learn quickly in the stock market is by losing money.
While the dot-com bubble scared some people away from the stock market permanently, others learned a much more applicable lesson for successful investing: don’t buy overvalued financial assets. It was as true at the turn of the millennium as it is today.
Opinions are those of the author and may not reflect those of BMO Private Investment Counsel Inc., and are not intended to provide investment, tax, accounting or legal advice. The information and opinions contained herein have been compiled from sources believed reliable but no representation or warranty, express or implied, is made as to their accuracy or completeness and neither the author nor BMO Private Investment Counsel Inc. shall be liable for any errors, omissions or delays in content, or for any actions taken in reliance. BMO Private Investment Counsel Inc. is a wholly-owned subsidiary of Bank of Montreal.
