# Meta Q4 on deck as AI spending tests investor patience

> Meta reports Q4 Wednesday with AI capex front and center. Investors want proof that bigger 2025-26 spend lifts revenue while preserving margins, free cash flow, and buybacks.

- Type: AI news
- Page: https://completeaitraining.com/news/meta-q4-on-deck-as-ai-spending-tests-investor-patience/
- Published: 2026-01-28
- Updated: 2026-02-19
- Author: Jeroen Erne
- Jobs: Finance
- Reading time: 2 min

### Meta Q4 earnings preview: AI capex in focus

Meta reports after the bell Wednesday, and the core question is simple: does the AI buildout earn its keep? Shares are down more than 12% since October as the market digests a bigger, longer investment cycle in data centers and infrastructure.

Management lifted 2025 capital expenditures to $70-$72 billion (from $66-$72 billion) and flagged 2026 capex growth as "notably larger." That trajectory is the entire story for margins, free cash flow, and buyback capacity over the next 24 months.

#### Street setup

- Q4 EPS: $8.16 (consensus)
- Q4 revenue: $58.4 billion (consensus), up from $48.4 billion in Q4 2024
- Q4 spend: $21.9 billion, versus $14.4 billion in the year-ago quarter
- Reality Labs: ~$959 million revenue, ~$5.9 billion operating loss expected

#### What matters for finance

The scale of capex will push depreciation higher into 2025-2026. Expect pressure on gross margins and a tug-of-war between investment pace and free cash flow optics.

- Capex shape: Any color on phasing, useful life assumptions, and data center cohorts entering service.
- AI monetization: Timelines for model deployment and revenue lift in ads, messaging, and enterprise tooling.
- Opex discipline: Headcount, compute costs, and unit economics for training vs. inference.
- Capital returns: Buybacks relative to capex ramp; balance sheet flexibility.
- Reality Labs drag: Loss run-rate and how much is fixed versus variable.

#### Strategy and competitive context

Meta is spending aggressively alongside Amazon, Google, and Microsoft as AI workloads swell. The company increased its bets with a $14.3 billion purchase of 49% of Scale AI and named Alexandr Wang chief AI officer to lead Superintelligence Labs.

Execution risk is real: delays around its Llama 4 Behemoth model, and a reported consideration to shift the next flagship model to a proprietary approach from open weights. Meta also cut roles in its metaverse group, re-allocating spend into wearables like AI smart glasses.

Meanwhile, Google's Gemini 3 has seized the narrative. Over the last 12 months, Google's stock is up 66%, Meta is up 4%, and Amazon is up 1.8%-a clear read on investor patience for long-cycle AI capex versus near-term product traction.

#### Policy and legal overhangs

Calls to restrict [Social Media](https://completeaitraining.com/tag/social-media/) access for users under 16 are rising-Australia has enacted a ban, and France is weighing similar steps. In the US, the [FTC](https://www.ftc.gov/?ref=completeaitraining.com) is appealing its loss in an antitrust case that argued Meta bought Instagram and WhatsApp to protect its position in personal social networking.

#### Pre-earnings checklist

- Guide: Does 2025 capex stay anchored at $70-$72B, and how "notably larger" is 2026 in dollars?
- Margins and D&A: Any update on server GPU useful lives and depreciation schedules.
- Ads demand: Signal on pricing and conversion lift from AI-driven targeting and creative.
- Model roadmap: Clear milestones for Llama 4 and whether a proprietary pivot changes partner ecosystems.
- Reality Labs: Path to narrowing losses and the role of smart glasses in near-term adoption.
- Capital returns: Buyback cadence relative to the capex ramp and free cash flow seasonality.

#### Bottom line

The print will be judged on two axes: revenue quality and capex credibility. If Meta shows accelerating monetization alongside a tight, phased capex plan, the market will give them time. If spend swells without clearer payoff timelines, expect the multiple to stay capped.

If you're building AI capability inside a finance org, here's a practical starting point: [AI tools for finance](https://completeaitraining.com/ai-tools/10b-ai-tools-for-finance/).

---

This is the Markdown form of https://completeaitraining.com/news/meta-q4-on-deck-as-ai-spending-tests-investor-patience/. The whole site is mapped at https://completeaitraining.com/llms.txt.
