Three hundred and forty-two billion dollars. Gone. In a single session.

That’s a rough estimate of combined market cap destroyed today across Meta, Microsoft, and Nvidia — three of the most widely held stocks in every 401(k), Roth IRA, and brokerage account in America. Meta cratered -11.44% to $525.72. Microsoft shed -6.57% to $356.77. Nvidia gave back -3.0% to $167.52. The NASDAQ dropped -4.48% to 20,948 — its fifth consecutive down week. The S&P 500 fell -3.38% to 6,368.

Here’s the thing: each of these moves has a different story underneath. Meta’s selloff is an earnings-and-guidance shock. Microsoft’s is a forward-revenue warning meeting a stretched valuation. Nvidia’s is contagion from sector-wide AI spending anxiety. Treating them the same is a mistake — and it’s exactly the kind of mistake that costs investors real money.

Let’s go through each one with the actual numbers, the actual drivers, and a clean verdict: buy, hold, or sell right now.

Contents

• 1. What’s the Macro Backdrop? (The Setup)
• 2. Meta -11.4%: Expensive AI Bets Meeting Reality
• 3. Microsoft -6.6%: When Cloud Slowdown Meets a 31x Multiple
• 4. Nvidia -3.0%: Contagion or Structural Crack?
• 5. Side-by-Side: Which Is the Better Entry Point Right Now?
• 6. Buy, Hold, or Sell — A Clear Verdict on All Three
• 7. FAQ

What’s the Macro Backdrop? (The Setup)

Before we zoom into individual names, the macro environment matters enormously today — because it’s amplifying every negative signal.

The Fed Funds Rate currently sits at 2.5% (as of February 2026). That’s meaningfully lower than peak 2023 levels, which should be supportive for growth stocks. But here’s the complication: the market is entering a shortened week featuring jobs data AND ongoing geopolitical war uncertainty, per Yahoo Finance. That combination — macro data risk plus geopolitical noise — makes institutional investors trim exposure, not add it.

Now here’s an important counterpoint: Seeking Alpha notes the S&P 500 earnings yield has jumped back above 5%. That’s meaningful. When earnings yield exceeds the risk-free rate by a solid margin, stocks historically look attractive on a fundamental basis. FactSet analysts are projecting a 29% S&P 500 price increase over the next 12 months. Macro fear and fundamental value are colliding — which is exactly why you need to pick your spots carefully.

Apple, notably, was essentially flat today at $248.80 (+0.33%). Tesla fell a modest -1.67%. Amazon dropped -2.94%. The pattern is clear: this wasn’t a generalized meltdown. It was a targeted, earnings-driven hit on specific names. Which means the exits and entries are also specific.

Meta -11.4%: Expensive AI Bets Meeting Reality

Let’s start with the biggest single-name story of the day: Meta Platforms collapsing 11.44% to $525.72 on volume of 30.1 million shares. That’s a stock that was trading near all-time highs just weeks ago, now giving back months of gains in a single session.

What actually happened? Meta’s most recent quarterly earnings report revealed a familiar tension that investors have been willing to ignore — until now. Revenue growth has been strong (Meta’s advertising engine remains one of the most powerful cash machines in corporate history), but the forward guidance is where the trouble lives.

Meta’s 2026 capital expenditure plan is staggering: the company has guided for $60–65 billion in capex this year, almost entirely directed at AI infrastructure — data centers, custom AI chips, and the compute backbone for its Llama model family and AI-powered ad targeting. That’s up from roughly $38 billion in 2025. A $25 billion increase in a single year.

Microsoft -6.6%: When Cloud Slowdown Meets a 31x Multiple

Microsoft’s -6.57% drop to $356.77 on 37.9 million shares — nearly double its average daily volume — tells a different story than Meta’s. This isn’t an existential capex question. It’s a valuation story colliding with a growth deceleration signal.

Microsoft Azure, the company’s cloud computing division, has been growing at roughly 28–31% year-over-year for most of 2024 and 2025. But recent commentary from hyperscaler customers — companies that rent compute from Azure, AWS, and Google Cloud — suggests enterprise AI workloads are taking longer to scale than anticipated. Proof-of-concept AI projects are running into budget approvals, security reviews, and integration challenges at the enterprise level. The result: Azure growth is expected to decelerate toward the 24–26% range in the coming quarters.

That might sound like a minor revision. It isn’t — not at a 31x forward earnings multiple.

Nvidia -3.0%: Contagion or Structural Crack?

Nvidia’s -3.0% move to $167.52 on massive volume of 196.2 million shares — nearly 4x its peers — is the most interesting case of the three. Because unlike Meta and Microsoft, Nvidia doesn’t have a specific bad-news driver today. This is primarily contagion.

Here’s the chain of causation: Meta announces massive AI capex. The market reads that as inflationary for AI infrastructure costs and uncertain for returns. Microsoft signals cloud/AI softness. The market starts questioning whether AI monetization timelines are getting pushed back. Both of those signals — combined — compress the multiple investors are willing to pay for the company that sells the shovels in the AI gold rush.

Nvidia’s H100 and B200 GPUs are still the dominant accelerators for AI training workloads. Data center revenue was approximately $35.6 billion in Q4 FY2026 — an extraordinary number. Gross margins remain near 73–75%. The forward P/E, even at $167.52, is roughly 28–30x — historically moderate for a company that has grown data center revenue 800%+ over two years.

The real question for Nvidia isn’t today — it’s 2027. AMD has been closing the gap on the MI300X series, gaining traction in inference workloads where customers are price-sensitive. Internally, Google (TPUs), Amazon (Trainium), and Meta (custom MTIA chips) are all building alternatives to reduce Nvidia dependency. The moat is real but it’s being attacked from multiple directions simultaneously.

Case Study 4 — The E*TRADE Retail Trader Who Bought in January 2023: Someone who bought Nvidia at $15 (split-adjusted) in early 2023 and held to today at $167.52 is sitting on an 11x return in roughly 3 years. That’s not a typo. The question they face today is whether to take profits or hold. My answer: they should set a trailing stop at $145 (approximately 13.5% below current) and let the position run. The Blackwell GPU ramp is real. The demand pipeline is real. A 3% down day on macro contagion is not a thesis-breaker.

AMD, for comparison, is also struggling — Reuters reports AMD predicted weaker Q1 sales and shares plunged on Nvidia comparisons. That AMD weakness actually reinforces Nvidia’s competitive position. If AMD can’t take significant share even with Nvidia at elevated valuations, the moat is deeper than bears claim.

Side-by-Side: Which Is the Better Entry Point Right Now?

Let’s get systematic. Here’s the full comparison across the three names — and the broader market context — with today’s closing data and fundamental estimates.

Now let’s look at how these three names compare to the broader market and their closest peer group:

Apple’s relative strength today is telling. It moved essentially flat because it doesn’t have a specific AI capex overhang narrative, its services revenue is sticky, and buybacks are relentless. That’s the template for what the market is rewarding right now: capital discipline over capital deployment. Meta is being punished for the opposite.

Buy, Hold, or Sell — A Clear Verdict on All Three

Let’s be direct. No hedging, no ‘consult a financial advisor’ deflection. Here are the actual verdicts based on the data.

Frequently Asked Questions