Meta Beats Q3 2026 Revenue, but the Stock Falls 8% on AI Spending
Meta Platforms released its third-quarter 2026 results after the market close on October 21, and the market reacted with another sharp sell-off in the stock — even though revenue beat consensus comfortably. The reason, once again in 2026, wasn't the current business but the company's artificial-intelligence infrastructure spending plan: Meta raised its 2026 capex guidance and warned it will keep investing "aggressively" in 2027, without yet giving an official figure. The stock, which closed at $705.44 on October 19, ended Friday, October 23 at $648.85 — a cumulative drop of roughly 8% in just two trading sessions.
Context: the third straight quarter where capex outweighs the result
This isn't an isolated episode. Throughout 2026, Meta has repeated the same pattern: results that beat on the metric that matters most to the business (revenue), followed by a negative stock reaction driven by the pace of AI investment.
- April 2026 (Q1 results): Meta raised its 2026 capex guidance to a range of $125-145 billion. The stock fell more than 6% in after-hours trading that same night.
- July 2026 (Q2 results): the company narrowed that same guidance to $130-145 billion (from $125-145 billion previously), and free cash flow dropped sharply from the prior quarter. The stock fell about 8% the day after the report.
- October 2026 (Q3 results, the one covered here): guidance rose again, this time to $135-145 billion, and the stock fell once more.
In other words, the low end of Meta's own spending guidance for this very year has risen in each of the three quarters reported so far — from $125 billion to $130 billion to $135 billion — while the high end has stayed fixed at $145 billion all three quarters. Here's how that guidance has evolved quarter by quarter:
Meta's 2026 capex guidance: the low end keeps rising every quarter
Meta Platforms earnings releases, Q1-Q3 2026 (Yahoo Finance, Silicon UK)
The chart shows something relevant: the market isn't reacting to a one-off surprise, but to a sustained trend of upward spending revisions, quarter after quarter, without the company having had to touch the $145 billion ceiling yet — which suggests that ceiling could be the next level to move if the trend continues in upcoming reports.
The hard data, verified: what Meta said on October 21
The quarter's business metrics were solid:
- Revenue: $64.6 billion, up 26% year over year, above the $63.31 billion consensus (per Newsquawk) and above the company's own guidance range of $61-64 billion.
- GAAP earnings per share: $3.83, well below the $6.67 consensus expected (per Newsquawk) — an apparently severe miss that doesn't reflect a deterioration of the underlying business.
- The cause of the EPS miss: a one-time legal charge tied to the large settlement Meta reached on August 26, 2026 with most U.S. state attorneys general (48 states and several territories) to resolve lawsuits over the impact of its platforms on minors' mental health — a settlement whose total cost has been estimated at up to $18 billion. Excluding that one-time charge, several trackers place adjusted EPS in a range of roughly $7.25 to $7.40, above consensus.
- Quarterly capex: $34.91 billion, a pace that pushed free cash flow into negative territory (around -$2.12 billion) for the quarter, according to MarketBeat.
- Fourth-quarter guidance: revenue of $73-76 billion — which, if met, would mean another quarter of accelerating growth versus the third quarter's 26%.
The stock's reaction was immediate: it fell sharply in after-hours trading as soon as the results came out (several sources cite drops of between 6% and 11% depending on exactly when the reference price was taken), and selling pressure continued over the following sessions until the stock closed on October 23 at $648.85 — 8% below the October 19 close. A move of this size in a company with a market capitalization of more than a trillion dollars wipes out tens of billions of dollars in market value within a few sessions.
Meta's stock reaction after each 2026 quarterly earnings report
MarketBeat, InvestingLive, Yahoo Finance, TheFly
As the chart shows, the magnitude of the drop has been similar across all three quarters of 2026 (between 6% and 8%) — this isn't a one-off episode, but a pattern the market repeats every time Meta reports earnings, regardless of whether the underlying business is improving or not.
Sector/asset analysis: winners and losers from the same news
Big Tech and large-cap growth stocks: the reaction to Meta is the first major test of the "Magnificent Seven" October earnings season, with Microsoft and Alphabet reporting on October 28 and Amazon on October 29. If the market punishes those companies too for raising their own AI capex, it would confirm the concern is sector-wide rather than Meta-specific — which would pressure the entire large-cap tech segment regardless of each company's individual results.
AI infrastructure suppliers (semiconductors, memory, data centers): paradoxically, higher capex from Meta is good news for its suppliers — AI chipmakers, memory producers (such as Micron, which we've already covered on this blog over the AI memory supercycle) and data-center operators. Meta's drop shouldn't be mistaken for a negative signal for the entire AI supply chain; in fact, it's usually the direct counterpart: the spending that scares Meta's investors is revenue for its suppliers.
Data-center REITs and electric utilities: the pace of spending from Meta (and the rest of the hyperscalers) keeps supporting demand for data-center space and the electricity to power it — a tailwind for specialized REITs and for utilities exposed to Big Tech demand, even as Meta's own stock falls.
Fixed income and growth-stock valuations: with the 10-year Treasury yield still near 2007 highs (around 5.3%, as we covered on this blog on October 2), any sign that a large tech company will burn through more cash than expected is punished more harshly than in a low-rate environment — the opportunity cost of holding a growth stock without positive free cash flow rises the higher the risk-free yield available as an alternative.
Value investors: Meta trades at a rich multiple relative to the broader market, and episodes like this feed the value-investor argument that the large-cap growth segment is more exposed to short-term disappointments than companies with more predictable cash flows and lower capital intensity.
What to watch next
- October 28, 2026: Microsoft and Alphabet report after the close — the key question is whether they too raise their own AI capex and how the market reacts, which will confirm or disprove whether the concern is sector-wide.
- October 28, 2026: Fed rate decision — the market is mostly pricing in a hold at the 3.75%-4.00% range, but any surprise in the tone regarding December would directly affect the valuation of growth stocks like Meta.
- October 29, 2026: Amazon reports results, with AWS capex as another reference point for the sector's AI spending.
- 2027 capex guidance: Meta has not yet given an official figure for next year — it has only said it will invest "aggressively." Unofficial analyst estimates circulating before this report pointed to a range of $225-250 billion, a jump of 60%-80% versus 2026. The official figure will likely come with the fourth-quarter results, in January 2027.
- January 19, 2027: the date on which Meta's new non-voting Class C share structure — announced to fund its AI investment without diluting Mark Zuckerberg's voting control — would take effect, a move worth watching for its impact on free float and governance.
Historical parallel: the same script as in October 2025
What happened this week isn't new for Meta at this time of year. In October 2025, after reporting third-quarter results, Meta's stock also fell sharply for the same reason: a jump in capex guidance for the following year that unnerved investors, building to a decline of roughly 11% for the month. In other words, this is the second consecutive year the market has reacted negatively to the same type of announcement at the same point in the earnings calendar — a sign that, for a meaningful share of investors, patience with Meta's AI investment cycle has a limit that gets tested every year in this same reporting window, regardless of the fact that the underlying advertising business keeps growing at a high double-digit pace.
