Is the AI boom heading for a 2008-style crash?
AI has pushed the S&P 500 to record highs, but the gap between what tech giants are spending and what they're actually earning is starting to look familiar.
Is the AI boom heading for a 2008-style crash?
AI has pushed the S&P 500 to record highs, but the gap between what tech giants are spending and what they're actually earning is starting to look familiar.
Artificial intelligence (AI) has driven the S&P 500 to historic heights and pushed equity valuations to their most stretched levels since the 2008 global financial crisis. This overwhelming market domination is undoubtedly concerning, particularly when viewed in parallel to the late-1990s dot-com bubble and 2008 crash. The similarities are striking, and bring into question whether AI's unprecedented growth is here to stay, or just a bubble about to burst.
The problem lies in wild optimism, echoing the blind faith of 2008 when big banks assumed house prices would never fall. Today, the stock market operates under a similarly unproven assumption that individuals and businesses will eventually buy enough AI software to justify the immense up-front costs. This assumption is so strong that tech giants are predicted to have spent an estimated $5.5 trillion on AI, yet a glaring $600 billion gap has emerged between what these companies spend on tech hardware and what the AI market actually makes in sales. This massive divide between spending and real revenue is even wider than the gap seen during the 2001 dot-com crash, signalling that expensive data centres are being built way faster than the actual customers are showing up to pay for them.
Just as Wall Street used confusing financial tricks to hide the risks of bad mortgages in 2008, the modern tech world has built its own shaky foundation by essentially recycling investment money in a closed circle. Rather than waiting for real customers to buy their products, the industry's biggest players are trading the same cash back and forth to make themselves look successful. For example, Microsoft has funnelled billions into OpenAI, which OpenAI then turns around and spends directly on Microsoft's cloud services. To handle that workload, Microsoft buys billions of dollars' worth of computer chips from Nvidia, and Nvidia then takes those profits and invests billions right back into OpenAI. This circular money machine creates a misleading illusion of booming demand. Ignoring the recycled money being passed around amongst the tech elites, there seems to be comparatively less investment from ordinary organic sources.
This insular growth pattern has created a market in which a tiny handful of companies control everything, mirroring the "too big to fail" setup that triggered the 2008 banking collapse. The stock market is no longer a diversified mix of the whole economy; just seven massive tech companies now command over a third of the S&P 500's total value. If you take away these pro-AI-related stocks, the rest of the market has not exponentially grown at all, leaving the entire financial world dependent on a single trend. If just one prominent AI startup runs out of cash, or a tech giant decides to cut back on chip spending, it could trigger a massive domino effect that drags down the whole market, just like how the collapse of a few interconnected banks paralysed global finance.
If this AI bubble bursts, the damage will ripple past Wall Street and into the real economy through corporate debt and everyday utility bills. To keep funding these massive data centres and keeping the cash circle spinning, tech companies are taking on historic amounts of debt, leading central banks to warn of a major threat to financial stability if tech stocks crash. On top of that, building these massive data centres requires an immense amount of electricity, forcing local power companies to borrow money and expand their grids. If the AI craze suddenly cools down, these utility companies will be left with expensive, unused power equipment, and they will likely pass those losses directly to consumers through much higher home energy bills. Ultimately, 2008 proved that pouring cash into a trend cannot save it if the underlying business lacks an organic foundation, and if real customers do not show up soon, the resulting correction will hurt far more than just tech investors.