2026-10-23

Meta Beats Q3 Revenue by 26% to $64.6B - Stock Still Falls Nearly 7% as Future AI Spending Commitments Hit $825 Billion

Written & reviewed by the TradeSmrt editorial team · 2026-10-23

In this article

What Happened
Why a Revenue Beat Turned Into a Credit-Market Scare
What to Take Away From This
FAQ
Sources

What Happened

Meta Platforms (NASDAQ: META) reported third-quarter 2026 results after the close on Wednesday, October 21, and on paper the headline numbers looked strong. Revenue came in at $64.6 billion, up 26% year over year and comfortably ahead of Wall Street's roughly $63.4 billion estimate. The company also guided fourth-quarter revenue above analyst expectations and announced an effective 2-for-1 stock split - though not the kind most investors are used to. Rather than splitting existing shares, Meta plans to issue one new non-voting Class C share for every share currently outstanding, with trading expected to begin in late January 2027. The move lowers the per-share price and widens the potential investor base without diluting Mark Zuckerberg's and other existing holders' voting control.

Shares initially rose in after-hours trading on the headline beat, then reversed hard once investors worked through the rest of the report. By the close on Thursday, October 22, META had fallen to $648.85, down $48.67, or roughly 7%, from the prior close of $697.52 - one of the stock's worst single-day declines in recent memory. The reversal happened alongside a broader tech selloff: the Nasdaq Composite dropped more than 2% the same session as investors digested a cluster of megacap earnings reports (Meta, Tesla, IBM, and chip-equipment maker Lam Research all reported in the same stretch) alongside rising oil prices and a Treasury market on edge ahead of the Federal Reserve's October 27-28 meeting. But Meta's drop stood out even against that backdrop, and the reasons were specific to the company's own numbers.

Three details in the filing explain why a 26%-revenue-growth quarter turned into a near-7% selloff. First, reported diluted earnings per share came in at $3.83, which looked like a miss against some analyst models - but the figure was distorted by a one-time $9.89 billion charge tied to Meta's August settlement with attorneys general from 52 U.S. states and territories, a deal that could total roughly $18.6 billion over time. Strip that charge out, and adjusted EPS was actually $7.40, well above what the market had priced in. Second, and more importantly for the stock reaction, capital expenditures for the quarter alone hit $34.9 billion, pushing full-quarter free cash flow to negative $2.1 billion - the first time in years investors have seen Meta's core cash generation go negative. Third, and most consequential of all, Meta disclosed that its future spending commitments - mostly multi-year, off-balance-sheet leases and contracts for AI data centers and computing capacity - have ballooned to approximately $825 billion, up from roughly $700 billion as of Meta's prior quarterly update in July. Credit markets reacted immediately: the spread on Meta's five-year credit default swaps, a market gauge of how risky bond investors consider a company's debt, jumped to a record high the same day.

Why a Revenue Beat Turned Into a Credit-Market Scare

The mechanism connecting these three details is straightforward once you separate the income statement from the balance sheet. Meta's operating business is still growing fast and profitably - a 26% revenue beat with strong ad pricing is not a company in trouble. What spooked investors instead was the pace and structure of its AI infrastructure build-out. Meta raised the floor of its full-year 2026 capital expenditure guidance to a range of $135 billion to $145 billion, lifting the bottom end by $5 billion from its prior outlook, and lifted its full-year total expense guidance to $175-179 billion to account for the legal charge. That alone is a familiar story in this earnings season - AI-spending guidance hikes have hit several megacap tech names in 2026. What made Meta's disclosure different in scale was the $825 billion commitment figure, which breaks down into roughly $454.5 billion in non-cancellable contracts already on the books, plus another $355.7 billion in leases that have been signed but haven't started yet and therefore don't even appear on the balance sheet today. A smaller slice, up to $14.7 billion, is a five-year cloud-capacity purchase obligation that could shrink if the provider resells unused capacity elsewhere. Roughly $45.9 billion of the near-term, non-cancellable piece falls due before the end of 2026, with another $121.4 billion due in 2027 - a cash outflow schedule that doesn't wait for the revenue these data centers are supposed to eventually generate.

That structure is exactly why the credit market, not just the stock market, reacted so sharply. Equity investors can at least point to Meta's 26% revenue growth as a reason to believe the spending will eventually pay off. Bondholders care about something narrower and less forgiving: whether Meta's cash flow can service a debt and lease load that now stretches out to 2036, through contracts running as long as 30 years, while the company's free cash flow has just gone negative for the first time in years. A record-high credit default swap spread means the market is pricing in a meaningfully higher probability that Meta's credit profile deteriorates before this spending translates into matching revenue. This is the same tension that has shown up across the "hyperscaler capex" story all year - compressed into one quarter's filing, with specific, disclosed dollar figures investors could point to for the first time rather than general unease about "AI spending levels."

The Class C non-voting stock plan added a separate layer of investor discomfort on the same day. While the mechanics (one new non-voting share issued per existing share, diluting nobody's voting power) are designed to be shareholder-neutral, some investors read the timing - landing in the same earnings release as the capex jump and the legal charge - as evidence that management is trying to make the stock more accessible to a broader retail base precisely when a lower nominal share price might help offset the headline sting of the selloff. Whether or not that reading is fair, the optics of announcing a structural share change alongside disappointing cash-flow news didn't help sentiment that day.

It's also worth noting Wall Street wasn't uniformly bearish. Several sell-side analysts reaffirmed bullish ratings on Meta in the hours after the report, pointing to the 26% revenue growth, the strength of the core advertising business, and management's argument that current AI investment is already showing returns inside Meta's existing products (ad-targeting improvements, Reels engagement, and early monetization of AI features). That split - a near-7% stock decline alongside continued analyst optimism on the underlying business - is itself a signal that the market's reaction was concentrated specifically on the financing and spending-pace questions, not a referendum on whether Meta's AI strategy will ultimately work.

What to Take Away From This

Always check what's driving an EPS "miss" before reacting to the headline number. Meta's reported $3.83 EPS looked disappointing next to some estimates, but a one-time $9.89 billion legal charge was the entire reason. Adjusted EPS of $7.40 tells a very different story, and conflating the two numbers leads to the wrong read on the quarter.
Revenue growth and cash generation are not the same thing, and the market increasingly prices them separately. Meta grew revenue 26% and still posted negative free cash flow because capex outpaced incoming cash. When a company's spending is front-loaded relative to its revenue recognition, a "growth beat" and a "cash flow miss" can both be true in the same quarter.
Off-balance-sheet commitments deserve as much scrutiny as the balance sheet itself. The jump from roughly $700 billion to $825 billion in future lease and contract commitments wasn't a change in Meta's current debt load on paper - it was a change in future obligations that don't show up in a standard balance-sheet snapshot. Reading the full disclosure, not just the summary numbers, is where this kind of risk becomes visible.
Credit markets and equity markets don't always agree, and when they move together, pay attention. A record-high credit default swap spread alongside a near-7% equity drop is a stronger signal than either one in isolation - it suggests both bondholders and shareholders are repricing the same underlying risk from different angles.
A structural announcement like a non-traditional stock split lands differently depending on what else is in the same report. The same disclosure that might have been read as a shareholder-friendly move in isolation got bundled into investor skepticism because it arrived alongside a capex surprise and a legal charge on the same day.

FAQ

Is Meta's AI spending actually a problem, or is this just normal growing pains for a company investing in infrastructure?

It's genuinely too early to say with confidence. The core argument for the spending - that AI investment is already improving ad targeting and engagement inside Meta's existing apps - has some support in the revenue numbers, which grew 26%. The counterargument is that $825 billion in future commitments is an enormous bet relative to any company's current cash flow, and the negative free cash flow this quarter shows the near-term cost is real and not just a modeling exercise. Whether this is "growing pains" or "overextension" will depend on how quickly the AI investments start generating revenue that can offset the spending pace - something that will only become clear over the next several quarters.

Why did Meta's stock initially rise in after-hours trading before falling the next day?

The initial reaction reflected the simplest headline: revenue beat estimates by a wide margin, and Q4 guidance also came in above expectations. The reversal happened once investors and analysts worked through the rest of the earnings call and filing - the legal charge's effect on EPS, the negative free cash flow figure, and especially the $825 billion future commitments disclosure - which take longer to parse than a single top-line revenue number.

What exactly is Meta's new non-voting stock split, and how is it different from a normal stock split?

In a traditional stock split, every existing share is divided into multiple shares, lowering the price per share while keeping ownership percentages and voting rights identical. Meta's plan instead issues one brand-new, non-voting Class C share for every currently outstanding share, expected to begin trading in late January 2027. The practical effect on price is similar - a lower per-share price that can make the stock more accessible to smaller investors - but it does not touch the voting power of existing Class A or Class B shares, meaning Mark Zuckerberg's control over the company is unaffected.

Sources

This article is an original synthesis and analysis based on the reporting below, not a reproduction of the original articles. Please check the source articles directly for the most current figures.

⚠️ This article is for informational purposes only and is not investment advice. Market conditions change constantly - always verify the latest data before making investment decisions.