Wednesday, July 29, 2026 packed in a Fed decision and two of the four largest US companies reporting earnings, all in one afternoon. The Fed held rates steady but with the most hawkish dissent in nearly a decade. Microsoft beat on every number that matters and its stock rose. Meta beat on revenue and still fell nearly 8%. Same week, same AI spending story, two very different verdicts.

The short answer

Microsoft's stock rose because Azure revenue accelerated to 43% growth, evidence that its roughly $40 billion-plus in quarterly AI spending is converting into faster cloud revenue, not just bigger bills. Meta's stock fell nearly 8% because its earnings miss, driven mainly by one-time legal and severance charges, arrived alongside another increase to its already-massive AI capital spending guidance, now $135 billion to $145 billion for 2026, without the same clear acceleration in revenue to show for it. The Federal Reserve, meanwhile, held rates at 3.50% to 3.75% but saw three officials dissent in favor of a hike, the most hawkish split since 2016.

What the Fed actually decided

The Federal Reserve's own statement confirmed it held its target rate at 3.50% to 3.75% on a 9 to 3 vote. What made the decision notable was not the hold itself, nearly every economist expected it, but the dissent: Cleveland's Beth Hammack, Minneapolis's Neel Kashkari, and Dallas's Lorie Logan all voted for a quarter-point hike instead, the most dissents pushing for higher rates since September 2016. The Fed's own statement cited elevated uncertainty tied partly to the Middle East conflict, alongside solid economic activity and strong productivity and capital investment. Higher rate odds matter directly for the same AI buildout at the center of today's earnings: more of that spending is financed with debt, and every dissent toward higher rates raises its cost. See what June's Fed decision meant for your money for how quickly these odds have moved this summer.

Why did Microsoft stock rise after earnings?

Microsoft reported in its 8-K filing fiscal fourth-quarter revenue of $90.01 billion, beating the $87.6 billion to $87.7 billion analysts expected, with adjusted earnings per share of $4.74 against a $4.24 estimate. The number that mattered most was Azure: cloud revenue grew 43% in constant currency, ahead of the roughly 40% expected and an acceleration from prior quarters, pushing Azure's annual revenue past $100 billion for the first time. Microsoft Cloud revenue overall reached $59.3 billion, up 27%. Shares rose in after-hours trading. This is the scenario last week's Tesla and Alphabet post flagged as the real test: does AI capital spending show up as faster revenue growth. For Microsoft this quarter, the answer was yes.

Why did Meta stock fall despite beating on revenue?

Meta reported on its earnings call revenue of $60.8 billion, ahead of the $60.2 billion expected, with advertising revenue of $59.3 billion also beating estimates. Earnings per share came in at $6.18 against a $7.14 estimate, a miss of roughly 13%. The gap was driven mainly by one-time items: a $2.4 billion legal contingency charge and a $1.2 billion severance charge. Strip those out and Meta would have beaten on earnings too. What actually moved the stock down nearly 8% was capital expenditure guidance: Meta raised its full-year 2026 capex range to $135 billion to $145 billion, tightening toward the high end of what was already a large number, without pairing it with the same clear, accelerating revenue signal Microsoft delivered the same afternoon.

Same spending story, different evidence
Microsoft and Meta are both spending enormous, still-rising sums on AI infrastructure. The market's differing verdict on the same afternoon was not about who spent more, it was about whose spending has visible proof of paying off right now. Azure's acceleration to 43% growth is that proof for Microsoft. Meta's ad revenue beat is real, but a murkier earnings picture and rising capex without an equally clear growth signal left more doubt.

The honest counterargument: one good quarter and one messy one are not final verdicts

It would be dishonest to treat today as settling the AI capex debate:

  • Meta's EPS miss was mostly one-time charges, not a core business problem. Legal contingencies and severance costs are real expenses, but they are not the same signal as declining revenue or user engagement. The market's nearly 8% reaction priced in more than just those two line items.
  • One strong Azure quarter does not guarantee the next one. Cloud growth rates can decelerate as comparisons get harder, and Microsoft's own capital spending is still rising to sustain this pace, the same bet Meta is making without yet showing Microsoft's exact result.
  • The Fed's hawkish dissent adds a real, separate risk to both companies' math. If borrowing costs rise from here, the return bar for all this AI infrastructure spending gets higher for every company doing it, not just the one that had a weaker quarter.

None of that erases the contrast. It just means today's verdict is a real data point in an ongoing test, not the final grade.

⚠
Do not treat "beat" and "miss" as the whole story. Both companies beat revenue estimates today. The stock moves came from cloud growth acceleration in one case and capex guidance plus one-time charges in the other, the exact kind of underneath-the-headline detail this site has been tracking since Tesla and Alphabet's earnings and this week's broader chip stock selloff.

What a long-term investor should actually do

  1. Do not trade on the after-hours headline alone. The first-session move reflects a fast read on a complex report, not a settled verdict.
  2. If you hold these companies through a broad index fund, you already own both outcomes together. No action is required beyond understanding what happened.
  3. Watch Azure's growth rate and Meta's capex-to-revenue ratio over the next few quarters, not one report. That is the actual multi-year test, not a single afternoon.
  4. Keep an eye on Fed rate odds alongside the earnings story. A genuinely higher-rate path raises the financing cost of every company's AI buildout, Microsoft and Meta included.
  5. Keep contributing on schedule. See time in the market beats timing the market for why reacting to a single high-news afternoon like this one has historically cost more than it saved.
💡
The actionable takeaway: Microsoft's Azure acceleration is the clearest evidence yet that AI infrastructure spending can convert into faster revenue growth. Meta's messier quarter is a reminder that spending alone does not guarantee that outcome, and a more hawkish Fed raises the cost of getting it wrong for everyone still spending.

The quick version

  • The Federal Reserve held rates at 3.50% to 3.75% on a 9 to 3 vote, with 3 officials dissenting in favor of a hike, the most hawkish dissent since September 2016
  • Microsoft beat on revenue ($90.01 billion vs. $87.7 billion expected) and EPS ($4.74 vs. $4.24 expected), with Azure growth accelerating to 43% and crossing $100 billion in annual revenue; shares rose after hours
  • Meta beat on revenue ($60.8 billion vs. $60.2 billion expected) but missed EPS ($6.18 vs. $7.14 expected) mainly due to a $2.4 billion legal charge and $1.2 billion severance charge; shares fell nearly 8%
  • Meta raised its full-year 2026 AI capex guidance to $135 billion to $145 billion, and the market reacted to that spending increase without an equally clear revenue acceleration signal
  • The contrast is the clearest evidence yet in the ongoing AI capex test: spending that shows up as accelerating revenue gets rewarded, spending that does not gets questioned
  • A more hawkish Fed raises the financing cost of AI infrastructure spending for every company doing it, not just the one with the weaker quarter
  • The response for a long-term investor is the same as any high-news day: do not trade on the first headline, and watch the multi-quarter trend, not one earnings call