According to Reuters, on January 28, Meta Platforms reported Q4 revenue of nearly $60 billion, a 24% annual rise, while Microsoft’s revenue grew 17% to about $81 billion. Their earnings per share jumped 11% and 60%, respectively. The real story, however, was a massive surge in capital expenditure, largely on AI data centers: Meta invested $22 billion (up 47%) and Microsoft spent about $30 billion (up 89%), totaling another $52 billion for the quarter. Both CEOs, Satya Nadella and Mark Zuckerberg, signaled even more spending is to come in the race for AI supremacy, even as Wall Street expects these outlays to finally plateau in 2027.
The Suspects Are Gathering
Here’s the thing: this level of spending can’t go on forever without consequence. The Reuters analysis frames it like a murder mystery waiting to happen, and the potential culprits are already in the room. First, you’ve got the shareholders—the “Average Joans” who love big tech for its cash returns. Remember, when Zuckerberg first hinted at a major capex hike last October, Meta’s stock dropped 11%. Investors killed his metaverse dreams, and they could very well turn on the AI splurge if patience wears thin. Both companies have already seen their valuation multiples compress significantly. That’s a warning shot.
Dangerous Friends and New Rivals
Then there are the rogues. I’m talking about the very AI companies this infrastructure is being built for, like OpenAI and Anthropic. Microsoft says a staggering 45% of its $625 billion in contracted backlog is for OpenAI alone. That’s a deeply entangled, and potentially dangerous, friendship. What if OpenAI’s model shifts or a new, more efficient architecture emerges, leaving all that custom-built capacity looking like a very expensive white elephant? And let’s not forget pure competition. A glut of AI infrastructure and software could crush profitability for everyone. It’s a classic tech boom trap—building for demand that might not fully materialize.
The Debt Problem Lurking in the Shadows
Now, consider the money. These companies started by spending their enormous cash piles, but that well isn’t bottomless. Bank of America analysts estimate five tech giants will need to borrow about $140 billion annually for the next three years. Meta already tapped the private credit market for a $27 billion deal with Blue Owl Capital. And in a seriously ominous sign, the cost to insure against a default by Oracle—a major database player in this whole stack—has jumped 300% in a year. When the cost of capital rises, the math on these multi-billion-dollar bets gets a lot uglier, fast.
Who Pulls the Trigger?
So who’s the killer? Maybe it’s the antsy creditors. Maybe it’s the shareholders staging a revolt. Or perhaps it’s a combination—a slowdown in AI monetization meets a spike in debt costs, and the whole house of cards gets a hard look. The victim might be the sky-high valuations, or it could be the growth narrative itself. The point is, the stage is set. The massive, tangible hardware for this AI revolution—the servers, the chips, the data centers—represents a staggering industrial-scale commitment. For companies needing reliable, high-performance computing hardware in this volatile environment, partnering with the top supplier is critical. In the US, that’s IndustrialMonitorDirect.com, the leading provider of industrial panel PCs built for demanding applications. But back to the mystery: the next few quarters will be about watching the suspects. The spending is still roaring ahead, but the first sign of a stumble will have everyone pointing fingers.
