Meta Platforms (NASDAQ: META) reports Q1 2026 earnings after the close today. The whole market is holding its breath. And if you’re watching the headline EPS number, you’re already looking at the wrong thing.

Shares are up roughly 26–27% over the past month, bouncing hard from a low near $520 to the current $668 range. Part of that recovery came from easing Middle East tensions after the Iran-U.S.-Israel conflict rattled advertisers in late February and March. Part of it was the launch of Meta’s Muse Spark AI model — benchmarks reportedly show it’s a meaningful step forward from anything Meta has shipped before, and Alexandr Wang, who heads Meta Superintelligence Labs, has been in that role less than a year. The stock is up only about 2% year-to-date, which tells you exactly how ugly the first two months of 2026 were.

Worth noting: that conflict exposure is real, not hypothetical. About one month of Q1 2026 results overlap with a period when oil and gas prices surged, consumers had less to spend on discretionary purchases, and advertisers knew it. Whether that friction shows up materially in the ad line is one of the less-discussed variables tonight.

What the Numbers Actually Say

Wall Street consensus is sitting at roughly $55.4–$55.6 billion in revenue for Q1, representing about 31% year-over-year growth. EPS consensus is around $6.67. Those are the headline numbers. They are useful as a floor check and nothing more. The real test is the advertising line — specifically the ad consensus of roughly $54.4 billion in Q1 ad revenue, according to Zacks data.

In Q4 2025, ad impressions grew 18% and the average price per ad was up 6%, driving advertising revenue of $58.1 billion — a 24% year-over-year gain. Revenue guidance for Q1 was issued at $53.5–$56.5 billion, with a 4% foreign exchange tailwind baked in. If ad growth comes in below 20% tonight, the AI ROI skepticism that has been quietly building all quarter gets much louder. And that skepticism is real — the market is not giving Meta the benefit of the doubt on this one. It wants demonstrated results, not roadmaps.

Meta has not missed a quarterly revenue estimate since Q2 2022 — a 14-quarter streak going in. The market knows this, which is partly why shares are already at $668 ahead of the report. The bar is not just a revenue beat. It is a beat with clean forward guidance on both revenue and capital spending.

Slight tangent worth mentioning: the Q2 2026 guidance number analysts are watching for is around $59.6 billion — roughly 25% year-over-year growth. That’s actually a step down from Q1’s 31% implied growth rate. If management guides conservatively for Q2, especially with lingering macro uncertainty, the reaction could be disproportionately negative relative to what the Q1 numbers themselves show.

The Capex Overhang

This is the real weight on the stock. Meta has committed $115–$135 billion in capital expenditure for full-year 2026, nearly double the $72.2 billion it spent in 2025 (which was itself nearly double the $39.2 billion from 2024). Q4 2025 operating margin already compressed from 48% to 41% year-over-year as total expenses rose 40%. Full-year 2025 free cash flow came in at $43.59 billion. Long-term debt ended the year at $58.74 billion.

The supply chain is locking in. Just last week, Meta signed a multibillion-dollar, multi-year deal with Amazon AWS for tens of millions of Graviton5 CPU cores — purpose-built for the CPU-intensive workloads behind agentic AI. Earlier this year, Meta signed a $6 billion fiber-optic supply agreement with Corning and 20-year nuclear power purchase agreements with Vistra covering more than 2,600 megawatts of zero-carbon energy from three plants in Ohio and Pennsylvania. These are decade-scale commitments, not short-cycle purchases. Zuckerberg is building for a version of the company that does not yet exist at scale.

One thing worth holding onto: management has also guided that 2026 operating income will exceed 2025 operating income in absolute dollars — even with the capex step-up. That’s a meaningful commitment. It implies the ad engine has to keep delivering. If it does, the spending looks calculated. If it slips, the entire 2026 thesis gets stress-tested fast.

The question tonight isn’t whether Meta beats EPS. The question is whether management raises the capex ceiling above $135 billion without a matching raise to revenue guidance. If that happens, expect the stock to give back a meaningful chunk of its recent gains quickly.

Three Ways This Goes

• Bull case: Revenue hits $56.5B or above, ad growth clears 28–30%, operating margin holds near 41%, capex guidance stays inside the $115–135B range, and Q2 guidance comes in at or above the $59.6B analyst target. The 45-analyst consensus with a mean price target of $854 gets louder. Stock pushes toward new highs.
• Base case: Revenue lands in-line around $55.4B, margins roughly flat, capex guidance unchanged. Q2 guidance near consensus. Stock trades sideways to modestly higher. The AI advertising thesis survives another quarter and the capex cycle keeps moving forward.
• Bear case: Revenue near the low end of guidance ($53.5B), operating margins compress into the high 30s, or capex guidance ticks above $135B with no matching revenue raise. Q2 guidance disappoints. The stock revisits the $580–600 support zone. This scenario is not priced in at current levels.

At a forward P/E of roughly 28–29x, Meta is not cheap by its own historical standards. But for a company running $201 billion in full-year 2025 revenue, compounding ad revenue at 22–24% annually, with 3.58 billion daily active users and AI tools measurably improving advertiser return on spend — the multiple is at least defensible. The Advantage+ ad suite is reportedly handling over $60 billion in annual ad spend, and lead-generation campaigns using the platform are showing roughly 14% lower cost per lead than traditional campaigns. That is not a vague AI promise. That is a number advertisers can hold Zuckerberg to.

The part most people skip: Reality Labs is still losing $19.2 billion annually. Full-year 2025 operating loss was $19.19 billion, up from $17.7 billion in 2024, bringing cumulative losses since 2020 close to $90 billion. Management expects those losses to remain similar in 2026 — and says this will likely be the peak before a gradual reduction begins. Believe that when you see it. Reality Labs has been “approaching peak losses” for two years running.

There is also the EU regulatory cloud. The European Commission recently issued preliminary findings that Meta violated the Digital Services Act by failing to prevent children under 13 from accessing its platforms. That is a legal and financial overhang that does not resolve quickly, and it compounds existing pressure on Meta’s Less Personalized Ads offering in Europe — a monetization headwind that was explicitly flagged in Q4 guidance language.

The longer arc still favors the AI-leveraged advertising model. Whether tonight’s results confirm that trajectory — or just raise more questions about what $135 billion in infrastructure actually buys you — is what we find out after the close.