Amazon Starts To Run Low On Cash

The question is not whether Amazon (NASDAQ: AMZN | AMZN Price Prediction) will run low on cash. The question is how low. Some of the other mega-cap tech companies have similar problems, but only one has a deeper problem. Oracle (NASDAQ: ORCL) has raised money “off-balance-sheet’, and its financial obligations are said to be as high as $300 billion. Its stock was down from $207 at the start of the year to $135 in April. It has bounced back since then. Off-balance-sheet funding is used by many companies, but Oracle’s stands out for its risk level.

Amazon’s gamble on AI data centers will be $200 billion this year, according to the FT. Andy Jassey, Amazon’s CEO, compares it to the money it invested in AWS a decade ago. He has said that it has paid off handsomely. He is correct about investing in AWS. However, there is no evidence that the money invested in AI data centers is comparable. It is too early to tell. Too early to tell is the same as too early to assess risk.

One other thing that is telling is a recent FT analysis of the AI spending by the big tech companies, which is driving up prices for everything from chips to electricity. The more these companies invest, the more the price of what they invest in rises. Additionally, data center construction has slowed amid local resistance. Electricity availability has also become a challenge. These factors have pushed the pace at which AI pays for itself.

Another open question is how money will be invested in data centers in the future as the horse race for AI dominance becomes increasingly competitive. According to CNBC, “Global spending on data centers could reach $7 trillion by 2030, according to McKinsey, and much of that spending can no longer come solely from hyperscalers. Instead, Big Tech is increasingly tapping private equity, private credit, and using debt to finance the capital-intensive build-out of the facilities.”

The real fear among those rational enough to look at the likely future of AI is that not every company that invests tens, if not hundreds of billions of dollars, will be a winner. Winners will fall into two categories. The first is sales to enterprises. The largest of which may be the national government. The other is use by individuals. Individual use is already starting to shake out. A number of research firms say that they have accurate data. Almost certainly, OpenAI’s ChatGPT is in the lead, followed by Google’s Gemini and Microsoft’s (NASDAQ: MSFT) CoPilot. Other research firms say CoPilot is doing poorly, and Anthropic’s Claude is doing well. Virtually every research firm shows that xAI’s Grok is so far behind in market share that it will never reach a point where it is financially viable.

Additionally, it is not clear at all what individual downloads do for these companies. No one has figured out a good way to monetize them.

The finances of AI data centers have become a Wild West of money. Amazon’s spending compared to its balance sheet is among the most aggressive, and may be the most aggressive of all. Jassey’s argument that this is similar to the investment in AWS does not have support. If he is wrong, Amazon will have spent a massive amount of the money it has in the bank.