Meta’s first-quarter numbers looked strong on paper: revenue rose 33% to $56.3 billion, and the company kept its operating margin at 41%.
But the stock still fell roughly 7% in extended trading after Meta lifted its 2026 capex plan to $125 billion to $145 billion.
The reason was simple: Wall Street likes growth, but it does not love an even bigger AI bill.
Meta said the higher spending reflects heavier component pricing and extra data center costs, while Zuckerberg struck a confident tone, saying the company sees enough evidence in its own work and across the industry to justify the investment.
What Meta’s $145 billion really means
The cleanest way to read Meta’s new guidance is that this is not a one-quarter overshoot. It is a much larger commitment to AI infrastructure.
Meta said the money is going toward higher component costs and future data center capacity, and it has already spent $19.8 billion in capex in the first quarter alone.
That fits with its February agreement with Nvidia, which calls for large-scale deployment of Nvidia CPUs, plus millions of Blackwell and Rubin GPUs and Spectrum-X networking gear.
In simple words, Meta is not just buying more servers; it is locking in a multi-year buildout of the AI stack.
Zuckerberg put it bluntly on the call: “Every sign” the company is seeing gives it “confidence in this investment.”
Why Nvidia and Micron sit at the center of this
Nvidia is the obvious first beneficiary as the chip giant has close to 90% share in training and inference markets.
That kind of position matters when hyperscalers are still pouring money into AI factories.
Nvidia traded at 24.5 times forward earnings at the time, even as analysts expected more than 70% earnings growth in the current fiscal year.
In simple words, the investors are still paying a valuation that looks reasonable relative to the company’s growth and its grip on the AI chip market.
Micron is the less obvious, but in some ways more interesting.
Micron is one of only three global suppliers of high-bandwidth memory, or HBM, alongside Samsung and SK Hynix.
That matters because HBM sits next to advanced AI chips and helps feed them the data they need at high speed.
Meta’s own wording made the connection harder to miss as the company said higher component pricing was driving its capex increase.
Micron’s latest results also showed how tight the market remains.
The company forecast fiscal third-quarter adjusted EPS of $19.15, far above the $12.03 consensus cited by MarketWatch.
Smarter read on a misread call
Retail investors saw a spending problem and professional investors saw confirmation that the AI infrastructure cycle is not cooling.
Meta’s call did not weaken the bull case for chipmakers; it strengthened it.
With Meta, Amazon, Microsoft and Alphabet still expected to spend hundreds of billions of dollars on AI buildouts in 2026, demand for Nvidia’s GPUs, networking gear and CPUs, as well as Micron’s HBM chips, now looks more like a structural trend than a one-off surge.
The post Why smart investors are buying Nvidia and Micron after Meta's call appeared first on Invezz