Skip to content 
On January 28, 2026, Meta and Microsoft opened their books. Both beat expectations. Both promised to spend even more money on AI. And the market’s response? A split decision that tells you everything about where we are in the AI gold rush.
Meta’s stock jumped 10% in after-hours trading. Microsoft fell 4%. Same trend, same pressure, completely different investor verdict.
Let’s break down what actually happened.
The Numbers Game
Here’s what the report cards looked like:
Meta’s Q4 2025 Performance
| Metric | Actual | Expected | Verdict |
| Revenue | $59.89B | $58.42B | Beat ✓ |
| Earnings Per Share | $8.88 | $8.16 | Beat ✓ |
| Ad Impressions Growth | +18% YoY | Lower | Beat ✓ |
| Ad Price Growth | +6% YoY | Higher | Miss ✗ |
Microsoft’s Q2 FY2026 Performance
| Metric | Actual | Expected | Verdict |
| Revenue | $81.27B | $80.31B | Beat ✓ |
| Earnings Per Share | $4.14 | $3.92 | Beat ✓ |
| Azure Cloud Growth | 38-39% | 38% | Met |
| Cloud Gross Margin | 67% | 70% (prev) | Declining ✗ |
Both companies delivered solid quarters. So why did the market react so differently?
The Real Story: It’s About the Bill, Not the Beat
Here’s the thing about AI in 2026: the profit line is the snapshot, the capital plan is the plot.
Meta announced it plans to spend between $115 billion and $135 billion in capital expenditures this year. That’s up from $72 billion in 2025, almost doubling its spending in a single year.
Operating expenses? Between $162 billion and $169 billion, versus the $151 billion analysts were expecting.
Microsoft’s capital expenditure hit $37.5 billion for the quarter (including finance leases), above the $36.26 billion expected. And here’s the kicker: Microsoft Cloud’s gross margin slipped from 70% to 67%.
Translation: Both companies are spending more to make each dollar of revenue.
In normal times, that would be a red flag. But these aren’t normal times—this is an infrastructure race, and everyone’s building at once.
Why Meta Got a Standing Ovation
The market loved Meta’s earnings for three reasons:
- The ad machine is humming
Meta sold more ads than expected (18% more impressions). Yes, the price per ad didn’t grow as fast as hoped, but here’s the nuance: growth can come from volume or pricing. Meta’s delivering volume, and that’s fine for now.
- Forward guidance looked strong
Q1 2026 revenue guidance: $53.5-56.5 billion versus $51.27 billion expected. That’s not “in line”—that’s genuinely ahead.
- Zuckerberg’s message was clear
No hedging, no apologizing for the spend. His quote: “We had strong business performance in 2025. I’m looking forward to advancing personal superintelligence for people around the world in 2026.”
The market read this as: “We can afford this. Our ad business prints money, and we’re using it to build the future.”
Meta’s fundamental pitch is simple: the ad engine pays for the AI engine. As long as that equation holds, investors are willing to fund the experiment.
Why Microsoft Got the Side-Eye
Microsoft’s quarter was technically fine. Revenue beat. Azure grew 38%. Operating income came in strong at $38.28 billion versus $36.55 billion expected.
So what’s the problem?
The market is now pricing the cost of delivering growth as sharply as the growth itself.
Here’s what spooked investors:
- Margins are compressing. Cloud gross margin fell from 70% to 67%. That’s not catastrophic, but it’s the wrong direction.
- Capex keeps climbing. $37.5 billion in one quarter. Annualized, that’s $150 billion a year—just for Microsoft alone.
- Azure “met” expectations. In a crowded “AI winners” trade, meeting expectations feels like losing. The bar is higher for Microsoft because it’s supposed to be the AI infrastructure leader.
But here’s the counterpoint: Microsoft’s backlog is absolutely massive.
Commercial remaining performance obligation (basically, contracted future revenue) rose 110% to $625 billion. And 45% of that growth came from new commitments with OpenAI.
That’s $281 billion in contracted revenue tied to AI partnerships. The demand is real—the question is whether margins can stabilize while serving it.
The Bigger Picture: Four Companies, Half a Trillion Dollars
Let’s zoom out for a second.
Microsoft, Alphabet, Amazon, and Meta are expected to spend $505 billion on AI infrastructure in 2026 alone. That’s up from roughly $366 billion in 2025.
Read that again: half a trillion dollars in a single year, from just four companies.
This is not a feature rollout. This is an infrastructure arms race.
And the market is now asking a very reasonable question: When do we see returns that justify this level of spending?
What Investors Are Really Watching
Forget the headline numbers for a second. Here’s what smart investors are tracking:
- Monetization proof points
- Can Meta convert AI features into higher ad prices?
- Can Microsoft show that Azure AI customers are spending more per workload?
- Are attach rates improving, or are discounts still driving growth?
- Capital efficiency
- Is spending per dollar of revenue growth improving or worsening?
- Are margins stabilizing or continuing to compress?
- Language, not just numbers
- Did management sound confident or cautious about 2026?
- Are timelines for AI product rollouts getting firmer or softer?
Microsoft’s Satya Nadella tried to frame the narrative positively: “We are only at the beginning phases of AI diffusion and already Microsoft has built an AI business that is larger than some of our biggest franchises.”
But the market’s response said: “Prove it. Show us the path to margin recovery.”
The Verdict: What This All Means
Here’s the cleanest takeaway from this earnings night:
AI is no longer a growth story. It’s a return-on-capital story.
Meta got rewarded because it showed a clear funding model: ads pay for AI, and the ad business is strong enough to absorb the investment.
Microsoft got punished because the market wants to see margin stability alongside growth, and right now, they’re seeing the opposite.
Both companies are spending massive amounts. Both are betting on AI. But only one convinced investors that the spending will turn into repeatable cash flow in the near term.
The Bottom Line
We’re at an inflection point. The market has moved from “build at all costs” to “build, but show me the payoff timeline.”
Meta’s 10% jump says: “We believe in your plan.”
Microsoft’s 4% drop says: “We believe in your demand, but not your margins.”
The winners in 2026 won’t be the biggest spenders. They’ll be the companies that turn infrastructure into revenue, and revenue into profit—quarter after quarter, without sacrificing margin.
Because in an AI arms race, the bill comes first. But eventually, someone has to show the receipts.
