Meta Beat Revenue Estimates and Still Fell 20% in Two Weeks. Here's the Number That Actually Broke the Stock

Meta just delivered a quarter that beat on the number investors usually care about most, revenue, and got punished anyway because of a number that's increasingly overshadowing everything else this earnings season: free cash flow. Shares are down roughly 20% over two weeks, with the sharpest single move coming after the July 29 report, even as advertising revenue grew 28% and ad impressions climbed 14% year over year.

The timing made the reaction sting even more. This happened during what's being called the most important week for AI stocks in 2026, with Microsoft surging 16% and Amazon jumping 10% on their own earnings in the same stretch. Meta and Apple both fell that week while their megacap peers rallied hard, splitting what had been a unified AI trade into stocks the market believes are showing returns on their AI spending and stocks it doesn't.

Technically, this decline has been a slow bleed rather than a single crash, which is arguably more damaging to sentiment than one sharp drop would have been. The stock had already been sliding since mid-July before earnings even hit, then gapped down further on the report itself, falling as much as 10% in after-hours trading before settling around a 7 to 9% decline. That combination, a weakening trend into the print followed by a confirming drop on the news, is the technical signature of a market that had growing doubts validated rather than surprised by new information.

Support at $521.50 is now the level in focus. A clean break below that zone with volume would suggest sellers remain in control; holding it, even after this much damage, would at least establish a base for the stock to work from.

The number that actually moved this stock sits a few lines below revenue on the income statement. Meta generated $31.9 billion in operating cash flow for the quarter and spent $31.1 billion of it on capital expenditures, servers, data centers, and networking infrastructure, leaving free cash flow of just $784 million once other obligations were factored in. That's down from $8.5 billion in the same quarter a year earlier, a collapse of more than 90% in the metric investors use to judge whether a company's spending is self-funding or eating into its cash reserves.

Compounding the concern, management raised full-year capex guidance for the third time since spring, with CFO commentary offering no clear timeline for when this spending translates into proportionate returns. The one detail that briefly excited some investors, Zuckerberg hinting at a potential cloud business, was notable specifically because it was vague, a signal of a new revenue angle rather than a confirmed plan, in an earnings call otherwise dominated by cost questions.

The bull case treats this as classic overreaction to a metric that will normalize once the current capex cycle matures. Meta's core advertising business remains genuinely strong, 28% revenue growth and 14% higher ad impressions aren't signs of a struggling company, and a $3.58 billion one-time charge for legal proceedings and severance from workforce reductions inflated the EPS miss beyond what recurring operations would suggest. Some analysts explicitly framed the selloff as a "don't panic, accumulate" moment, arguing that Amazon and Microsoft's rallies the same week prove the market still rewards AI infrastructure investment, just not from every company equally, and that Meta could still earn that reward once its own returns become visible.

The bear case is that raising capex guidance three separate times in one year, while free cash flow craters over 90%, is exactly the pattern that should worry long-term holders regardless of how strong the underlying ad business looks quarter to quarter. Unlike Microsoft, which showed Azure growth translating directly into results that justified its spending, Meta offered no comparable proof point this quarter, just a hinted-at cloud ambition and a bigger number attached to the same open-ended AI infrastructure bet. The market's split reaction across the Magnificent Seven that week suggests investors are starting to demand evidence of return on AI capex company by company, and Meta didn't provide it.

Buying this dip means betting that Meta's core advertising engine eventually produces the cash flow to justify a capex figure that's grown from around $100 billion to as high as $145 billion in guidance within a single year, without a clear catalyst yet identifying when that payoff arrives. The stock's 20% decline has already priced in real disappointment, but nothing in the earnings call gave a specific timeline that would let a buyer here time an entry with confidence.

Avoiding or shorting Meta here means missing a business still growing revenue 28% with expanding ad impressions and pricing, fundamentals that haven't deteriorated even as the cash flow story has. If Meta's cloud hint develops into an actual disclosed strategy, or if capex guidance stabilizes rather than rising a fourth time, the stock could reverse quickly given how much bad news is already reflected in the price.

Meta's quarter is a reminder that beating revenue estimates doesn't protect a stock anymore when the market has shifted its attention to a different line entirely. Free cash flow collapsing over 90% while capex guidance rises for a third straight time is the story investors are trading on, not the 28% revenue growth sitting above it. Microsoft proved in the same week that the market will still reward AI spending, just only when the returns show up alongside it. Meta hasn't shown that yet.

Does 28% revenue growth outweigh a free cash flow collapse this severe for you, or has Meta lost the benefit of the doubt on its AI spending until the cash flow numbers actually turn around? Let me know how you're weighing it.