Is an AI Stock Market Crash Coming?
For the first time in ages, you didn’t hear this here first. Rumours are swirling about an impending AI stock market crash, and they’ve already reached the pages of The Guardian and the Financial Times.
So, what’s going on?
The Price-to-Earnings Problem
The central issue is the price-to-earnings (P/E) ratio — a classic stock valuation metric. It shows how much investors are willing to pay for every dollar of a company’s earnings, by dividing its current share price by its earnings per share (EPS).
- A high P/E ratio can suggest overvaluation, or that investors expect explosive growth.
- A low P/E ratio may point to undervaluation or doubts about future profitability.
Right now, AI companies have some of the worst P/E ratios since the dot-com boom — worse, in fact, than the run-up to the crash of 2000.
Wall Street’s AI Spending Spree
Investors have poured staggering amounts of capital into AI. Around $2.9 trillion is being spent in the US alone on AI infrastructure.
These companies have no trouble raising money, because investors are desperate not to miss out. The share prices of firms with AI assets are soaring — but with almost no profits to show for it.
Without profits, P/E ratios collapse, leaving stocks looking increasingly shaky.
The CoreWeave Warning Sign
Nvidia-backed CoreWeave just lost 30% of its share value in two days — wiping out $23 billion. The fall came after earnings revealed widening losses and infrastructure bottlenecks.
For institutional investors and fund managers, this is a problem. They have to prove they’re generating the best returns for their clients. Right now, they can’t. Nervousness is spreading fast.
Déjà Vu: Remember the Dot-Com Bubble?
The comparison with the dot-com crash is hard to ignore.
Take Pets.com: backed by Amazon and respected venture capital firms, it peaked at a $410 million market value in 2000. Within 12 months it went bankrupt.
The worry is that AI could be headed the same way.
Where Are the Productivity Gains?
The bigger issue is not just stock valuations but whether AI is delivering real productivity gains.
Yes, it has transformed creative output — this blog included — and some industries are experimenting with AI-generated content. Wired and Business Insider even admitted that one of their “journalists” was actually an AI.
But beyond these anecdotes, most businesses aren’t reporting major profit growth, productivity jumps, or headcount reductions to justify their vast AI investments.
The Bureaucracy Paradox
History offers some perspective.
When PCs first arrived in offices, many assumed clerks and accountants would vanish. In reality, IT expanded what finance teams could do — and companies demanded more. Far from shrinking, departments grew.
Capitalism, deregulated and data-obsessed, created new layers of bureaucracy: accountants, lawyers, IT specialists, MBAs. Governments embraced the same culture via outsourcing and subcontracting.
AI might streamline some tasks, but experience suggests that companies will just invent more work — and hire more managers to monitor it.
Why Businesses Aren’t Seeing the Benefits
Part of the problem is overestimating AI’s short-term impact.
- Yes, ChatGPT can fix problems faster than Google — but not always that much faster.
- Workers may use AI to finish tasks more quickly, but instead of producing more, they quietly pocket the time savings.
- Working from home makes this even easier.
Eventually, managers will find ways to capture those productivity gains for businesses. But that means more management, more monitoring, and — you guessed it — more bureaucracy.
Time Is Running Out
The clock is ticking for AI firms. Fund managers work on short timeframes and need results soon. If companies can’t demonstrate that their valuations are realistic, the AI stock market bubble could burst.
Big players like Meta may survive, because they have diversified income streams. But smaller AI-focused firms are at risk.
The stakes are high. The US economy only avoided a deep recession thanks to AI spending. Strip that away, and growth collapses. An AI-driven crash would hit harder than the dot-com bubble, especially given tariffs, government debt, and a weakened tax base.
Bounded Rationality and Bubbles
Economists once peddled ideas like the efficient market hypothesis, claiming free markets allocate resources optimally. Reality tells a different story: markets are prone to bubbles, booms, and busts fuelled by irrational exuberance and equally irrational fear.
The truth lies in bounded rationality. Humans don’t always make optimal choices. Decisions are constrained by time, attention, and limited information. Sometimes, instead of maximising utility, we satisfice — making choices that are merely “good enough.”
Like when you spend hours searching eBay for a replacement part, get bored, and end up buying a T. Rex LP instead. And, to be fair, “King of the Rumbling Spires” isn’t the worst consolation prize.
Conclusion: Is the AI Crash Inevitable?
The risk of an AI stock market crash is real. Valuations are inflated, profits are scarce, and productivity gains aren’t yet materialising.
The long-term potential of AI remains huge. But in the short term, investors may have bet too heavily, too quickly. If the bubble bursts, the fallout will be bigger than dot-com — and the shockwaves will be felt far beyond Silicon Valley.
https://www.theguardian.com/technology/2025/jul/30/zuckerberg-superintelligence-meta-ai
https://www.theguardian.com/technology/2025/aug/19/openai-chatgpt-stock-sale-reports
https://www.theguardian.com/us-news/2025/aug/21/ai-author-articles-wired-business-insider