Meta’s (NASDAQ:META | META Price Prediction) Chief Financial Officer, Susan Li, said in the earnings call that they anticipate 2026 capital expenditures “…to be in the range of $125-145 billion, increased from our prior range of $115-135 billion.” Wall Street punished the stock for this level of spending, arguing that Meta Platforms was leaning too heavily into data centers and AI.
Analysts prescribed that Nvidia cut back on spending and strengthen the business elsewhere. However, Mark Zuckerberg may have just revealed why the AI spending figure keeps going up, and Meta may not have as much flexibility as you think.
Why capex is set to keep increasing
Most analysts looked at the capex number instead of looking at the rationale behind it. Mark Zuckerberg gave that rationale early in the call, saying, “…we are increasing our infrastructure capex forecast for this year. Most of that is due to higher component costs, particularly memory pricing”.
Zuckerberg said most of the increase stems from higher component costs like memory pricing.
Thus, Meta Platforms isn’t just seeing an increase in capex because the company keeps building data centers to meet compute demand. The increase in capex is mostly coming from paying more to companies like Micron (NASDAQ:MU), Nvidia (NASDAQ:NVDA), and other hardware companies it needs to source from.
Meta essentially has its hands tied because it could either delay or cancel its data center projects and fall behind. Or, it can pay higher and higher for components, with manufacturers having increasing leverage due to every other hyperscaler doing the same.
And it’s not just Meta
Almost every other hyperscaler is increasing its capex forecast significantly this year, and some are pre-emptively warning that next year will include even more spending. These companies are going cash flow negative for this. You shouldn’t see more spending as a sign that their data centers are expanding.
Zuckerberg was the only CEO to admit that the increasing capex spending is in large part due to higher component costs.
But is this a money pit?
Hyperscalers are investing here before they get any return on their investment. Zuckerberg said during the same Q1 call that Meta does not have a “…very precise plan for exactly how each product is going to scale”. If you want to benefit from all the capex with the least amount of speculation, I’d again lean into AI hardware stocks.
That said, if AI works out in the longer term for these software companies, hyperscalers will be the bigger beneficiaries. They can save hundreds of billions if you take AI cost savings into account in the coming decades.
Why I wouldn’t double down on META stock
Meta Platforms is not among the premier hyperscalers, and I’d argue it’s the worst Magnificent 7 stock when it comes to AI. It simply does not have a solid consumer-facing AI offering. There’s not a single Meta AI LLM that managed to stay in the top 10 in benchmarks for a sustained period. You’re not going to see any programmer using a Meta AI model to develop an app.
You can buy it for just 19 times forward earnings, but I’d argue that’s fair value for a business whose earnings are tied to ad spending. AI-assisted ads have done well, but customers will pull back on spending if there’s a downturn. Other hyperscalers are not as exposed since their cash flows are far more diversified.
Things look more concerning when you look at free cash flow. EV-to-FCF is at over 32 times. META stock is more expensive than 80% of media companies.
I’ll argue your best move now is to buy into AI hardware companies like AMD (NASDAQ:AMD), MU, and NVDA instead. You get (very!) positive cash flow plus hypergrowth through at least 2028 if the spending holds.
Top Gaining Stocks
Top Losing Stocks