The AI Money Trap Nobody Talks About
The AI Money Trap Nobody Talks About
Michael Burry put it plainly last week: big tech companies are spending like they're drunk on venture capital, and the market has cheered them on. Nvidia is in talks to back OpenAI's Ohio data center with $250 billion. That is a quarter-trillion dollars for one project. The pre-bell bounce on oil prices dropping and Middle East tensions easing suggests traders are still in celebration mode. But Burry's warning cuts through the noise: the S&P 500 has voted, and what it has voted for may be a mistake.
Here is the uncomfortable part that most market commentary skips over: capital intensity is not a feature of great businesses. It is a sign of desperation.
Capital Intensity Signals Weakness
High spending on assets eats into returns and signals weak competitive moats. Low-capex businesses compound wealth faster.
When a company needs to pour $250 billion into one data center just to stay competitive, it is admitting something crucial: the business has no economic moat. A moat is durable advantage. Durable advantages do not require you to spend more than your rivals every single quarter. They allow you to earn more on less capital.
Contrast this with what Vanguard and the market's boring heroes are offering right now. Five utility stocks "just keep paying." Those utilities are boring precisely because they have moats: regulated infrastructure, inelastic demand, and the ability to return cash to shareholders rather than burn it chasing the next arms race.
Why This Matters for Your Retirement
If you are holding tech because you believe in AI, you are not wrong about AI. You are just making a bet that today's leaders will remain leaders while spending ruinous amounts of capital on infrastructure that commoditizes faster every year. That is not a given.
Jamie Dimon's recent advice to avoid long-dated bonds and favor short-term bond ETFs reflects the same skepticism: uncertainty is rising, and locking in long commitments is dangerous. The parallel applies to tech stocks. If you buy them at valuations that assume AI spending continues forever without margin compression, you are essentially buying a 30-year bond at credit-card rates.
How AI Spending Erodes Returns
Drag the growth rate down as capex eats into margin expansion. Watch what happens to your IRR when you exit at today's P/E.
The headlines tell you where the crowd is. NinjaTrader offering single-stock futures on Nvidia and Tesla to retail traders, oil dropping on geopolitical relief, AstraZeneca kicking off earnings season: all of these are noise. The signal is what Burry said and what the utility payers silently demonstrate. Proven cash generation beats speculative capex every time.
Varta, a German battery maker, just filed for insolvency. It was capital-intensive from day one. Utilities do not file for insolvency because they do not need to.
Downside Risk by Model
Companies dependent on infinite capital deployment face insolvency risk when markets tighten. Moat-bearing utilities face refinancing risk, which is manageable.
The bottom line
The market has voted for AI, but voting for something does not make the unit economics work. If you want reliable long-term returns, stop asking whether your stock is cool and start asking whether it needs to spend $250 billion every few years just to stay in the game.
This is educational information, not financial advice.
One well-researched article at a time. No spam, unsubscribe in one click.
No spam, no selling your address, unsubscribe in one click. The tools stay free either way.
Keep exploring: browse the stocks we cover or see what the super investors hold.