Let’s start with the number that matters: $547.54. That’s where Meta closed today — down 9.75% in a single session, erasing roughly $120 billion in market cap before the closing bell. In the same afternoon, Microsoft bled 5.93% to $365.97, and Nvidia slid 4.1% to $171.24 on volume of 182 million shares — nearly double its average daily float. The NASDAQ fell 1.63% to 21,408. The S&P 500 dropped 1.21% to 6,477.

This wasn’t a random Tuesday. This was a collision of three distinct catalysts — earnings disappointment, AI spending anxiety, and geopolitical noise from the Iran nuclear negotiation headlines — compressing into one toxic afternoon. And while the headlines blared ‘tech rout,’ the reasons behind each stock’s drop are completely different. Meta’s story is about spending. Microsoft’s is about cloud growth deceleration. Nvidia’s is about forward demand signals. Treating them as one story is how investors make the wrong call.

Here’s the forensic breakdown — what actually happened, why, what the numbers tell us about where each stock goes from here, and the verdict you can act on today.

Contents

• Meta’s 9.8% Collapse: Is Zuckerberg’s AI Spending Bet Blowing Up?
• Microsoft’s 5.9% Drop: When ‘Beat’ Isn’t Good Enough
• Nvidia’s 4.1% Slide: Demand Signal or Just Collateral Damage?
• Valuation Smackdown: Are These Three Actually Cheap Now?
• Iran, the Fed, and the Market Mood: Why Timing Made It Worse
• Three Investors, Three Different Outcomes
• Buy, Hold, or Sell: The Clear Verdict on Each Stock
• FAQ

Meta’s single-session drop wasn’t driven by a revenue miss. It was driven by something scarier for a market running on AI euphoria: the revelation that the AI spending cycle has no ceiling. Meta guided for 2026 capital expenditures in the range of $60–65 billion — up from roughly $38 billion in 2025. Wall Street had penciled in around $50 billion. The gap is enormous.

Here’s the thing: Meta’s Q4 2025 revenue was actually strong — approximately $48.4 billion, up roughly 21% year-over-year. Ad revenue held firm. Daily active users across the Family of Apps crossed 3.35 billion. On the surface? Stellar. But the market doesn’t price stocks on what just happened — it prices them on what’s coming. And what’s coming is a $65 billion CapEx year that will compress free cash flow and pressure margins.

Zuckerberg’s AI pivot — building out Llama models, AI-driven ad targeting, smart glasses, and eventually the metaverse compute layer — requires data center infrastructure at a scale that would make Amazon Web Services blush. The problem isn’t whether the strategy is right (it probably is, long-term). The problem is the payoff timeline. Investors are being asked to fund a multi-year infrastructure buildout with returns that won’t materialize until 2027 at the earliest. At a forward P/E that was sitting around 25x before today’s drop, that’s a tough sell.

The sell-off is also a sector-wide anxiety signal: if Meta is spending $65B on AI infrastructure, Microsoft and Google are presumably spending similarly. That raises a pointed question for the whole group — who’s paying for all this compute, and when does the ROI show up?

Microsoft’s situation is more nuanced — and in some ways more worrying from a structural standpoint. The company did beat on headline earnings. But Azure cloud growth came in at approximately 31% year-over-year for the most recent quarter, and management guided for the same rate in the next quarter. Wall Street wanted 33–35%. A two-percentage-point miss on cloud growth guidance from the world’s second-largest company by market cap is worth about $100 billion in market cap apparently — because that’s roughly what today’s session took off the table.

The deeper issue is that Microsoft’s AI monetization story — specifically Copilot — hasn’t visibly broken into Azure growth numbers yet. Microsoft has embedded Copilot into Microsoft 365 at $30/user/month premium pricing, has GitHub Copilot generating real developer revenue, and has the OpenAI partnership as both an asset and a liability (the compute costs are staggering). But the market expected all of this spending to show up as accelerating Azure growth, not flat growth.

To be clear: 31% cloud growth is not bad. AWS grew at roughly 19% over the same period. Google Cloud grew at about 30%. Microsoft is competitive. But at a forward P/E of approximately 32–33x — even after today’s drop — you’re pricing in acceleration, not deceleration. And deceleration is what management just guided for.

Here’s where it gets complicated: Microsoft is simultaneously a victim and a beneficiary of the AI spending cycle. They spend billions on OpenAI compute and Azure infrastructure — but they also sell that compute to other enterprises. The question is whether the revenue from selling AI services will outrun the cost of building them. Today, the market voted: not yet.

Nvidia’s 4.1% drop to $171.24 is the most complex of the three — because Nvidia didn’t report earnings today. It moved purely on collateral damage from the Meta CapEx shock, combined with the AMD demand signal that arrived separately. Reuters reported that AMD predicted weaker first-quarter sales and shares plunged on Nvidia comparisons — meaning the AI chip demand story got its first real crack in the narrative.