원문 보기
Investors are stuck between two difficult places. Artificial intelligence stocks are in a bubble that's about to burst, which will wipe out trillions of dollars of wealth, versus the demand for more AI compute is insatiable, and the industry has years to go before we see any sign of overcapacity, making artificial intelligence stocks a bargain with years of capital gains ahead.
But the eventual outcome probably lies somewhere in the middle. Quick note, I'm AI Mitch. Mitch's video clone. Mitch is the human in the loop. He wrote this script. Every idea, every lesson I share is based on Mitch's 35 plus years of experience as a financial adviser. He uses AI to scale his content and share it with more people while he focuses on the important work of helping his clients.
Enjoy the knowledge. The most well-known bear in this space is Michael Bur. He's famous for seeing the US housing and mortgage crisis coming before the 2008 financial crisis and making an enormous contrarian bet against subprime mortgages. He looked underneath the hood of a wildly popular investment theme, found financial assumptions that didn't make sense, and bet against it before everyone else saw the problem.
He became a household name because his story was featured in Michael Lewis's The Big Short, which was later made into the movie. His bearish thesis is based on four key points. Overbuilding. Bur believes companies may be spending far too much on AI data centers, chips, and infrastructure based on expectations of enormous future demand.
If that demand falls short of expectations, companies could be left with very expensive excess capacity. Understating the real cost, Bur argues that companies may be depreciating AI chips and equipment too slowly, even though newer technology can make that equipment obsolete very quickly. That could make today's AI profits look better than the underlying economics really are.
Too much financial leverage. The AI boom increasingly depends on debt, leases, purchase commitments, and other financial obligations to fund its enormous expansion. Bur's concern is that investors may be underestimating just how much financial risk is building underneath the AI boom. Circular financing.
Bur is concerned that some AI companies are investing in or financing their own customers who then use that money to purchase AI chips and computing capacity. His concern is that this can make underlying demand and revenue growth look stronger than they really are. CNBC just highlighted three more warning signs in the AI boom.
First, the cost of borrowing is going up. The 10-year Treasury yield is around 5.17%. Which means companies borrowing enormous amounts of money to build AI infrastructure are going to have to pay more for that money. Second, Oracle has relied heavily on the debt markets to fund its AI expansion, and CNBC reported concerns about rising costs surrounding one of its massive data center.
As of now, its debt is rated just one notch above junk. Third, the borrowing isn't slowing down. JP Morgan estimates that $4.1 trillion of AI related debt could be issued through 2030 and SoftBank just borrowed 11.1 billion at yields as high as 9.75%. Hyperscalers have to raise their assumptions of future profitability to justify climbing borrowing rates.
The question is how high is too high? Not knowing is a risk that investors should consider. On the other hand, Mark Zuckerberg, Sam Alman, Dario Amade, Sander Pitchi, Satcha Nadella, and Jensen Wang are the most successful corporate leaders of our generation. They are the ones committing hundreds of billions of dollars to develop and build AI models and infrastructure.
It's hard to bet against them. Then there are two highly regarded analysts who are outspoken AI bulls, Dan Ies and Gene Monster. Their bull case is that we are still in the early part of the AI buildout. User demand for tokens, the measuring unit for AI input and output is insatiable, which justifies the enormous spending.
With such strongly opposed views, what could an ordinary investor do? Take a little profit off the table as AI related stocks climb. Make sure you aren't overconentrated in technology stocks so if they tumble, they won't drag your entire net worth down with them. Understand both sides of the year 2000 tech wreck 1995 to 1999 and 2000 to 2003.
Use that history as a guide to make sure you're prepared for whatever happens next.
Comments 0