META: FY25 Deep Dive
Family of Apps delivered the fourth consecutive 22% revenue year at $198.8B. Reality Labs loss widened to $19.2B. Capex stepped up to $69.7B, and the Street cut price targets in April on the FY26 capex guide.
Key Takeaways
Meta closed fiscal 2025 (calendar year ended December 31, 2025) at $200.97 billion of revenue, up 22% year-over-year — the fourth consecutive year of roughly 20%+ growth in Family of Apps, which is the entirety of the story. Reality Labs revenue remains a rounding line at $2.2 billion, but the operating loss widened to $19.2 billion — the largest single-year drag in the segment's history — as headcount and custom silicon spend continued. Capex stepped up to $69.7 billion (roughly 80% above FY24), pulling free cash flow down to $46.1 billion from what would otherwise have been a $100B+ year. The company returned $31.6 billion to shareholders — $26.2 billion in buybacks plus $5.3 billion in dividends. Sell-side coverage is 25 analysts with 21 Buy / Strong Buy ratings, median price target $845.54, but the dispersion is notable — post-Q1 FY26 earnings saw five of the largest firms cut price targets (BofA, Keybanc, Wells Fargo, Rosenblatt, Stifel) by $15-$129 each, on concerns about FY26 capex trajectory.
Main business structure
Two reporting segments since the Reality Labs breakout:
| Segment | FY25 ($M) | % of Total | FY24 ($M) | YoY |
|---|---|---|---|---|
| Family of Apps | 198,759 | 98.9% | 162,355 | +22.4% |
| Reality Labs | 2,207 | 1.1% | 2,146 | +2.8% |
| Total | 200,966 | 100% | 164,501 | +22.2% |
Family of Apps is the ad-revenue business across Facebook, Instagram, WhatsApp, Messenger, and Threads. The 10-K breaks revenue into Advertising (>96% of FoA) and Other (~4%, primarily payments / Meta Business Pay). Ad revenue is the function of impressions × price per impression; Meta's long-run pattern has been impressions growing in the low double digits (Reels / Threads / WhatsApp monetization ramps) with price per impression tracking mid-single-digit to flat. The 10-K discloses the split each quarter in earnings calls; full FY25 breakdown points to impressions +~15% and price +~7% as the composition of the +22% line.
Reality Labs is VR / AR hardware (Quest headsets, Ray-Ban smart glasses / Orion) + the Horizon / Metaverse platform investment + a growing AI custom silicon and systems research line housed inside the segment. Revenue here is essentially rounding; the segment's significance is on the expense side — it houses the single largest planned loss bucket in any US mega-cap.
Geographic mix (based on user geography disclosed in the 10-K): US & Canada ~45% of revenue, Europe ~25%, Asia-Pacific ~20%, Rest of World ~10%. Proportions roughly stable year-over-year. No single customer / advertiser accounts for 10% disclosure.
Scale anchors. Family of Apps daily active people (DAP) reached ~3.5 billion across the app family (Facebook + Instagram + WhatsApp + Threads), adding roughly 150M+ annually. Active advertiser count exceeds 15 million monthly. Threads monthly actives crossed 300M during FY25. Ray-Ban smart glasses crossed 2M units shipped lifetime, Quest 3S / Quest 3 installed base in the high single-digit millions. These are the nouns behind the $69.7B capex.
Key core metrics (4-year trend)
1. Family of Apps revenue growth
| FY22 | FY23 | FY24 | FY25 | |
|---|---|---|---|---|
| FoA revenue ($B) | 114.5 | 133.0 | 162.4 | 198.8 |
| YoY | — | +16% | +22% | +22% |
Four consecutive years of deceleration from the FY22 trough, with the last two years flat at +22%. The stability is the remarkable feature — Meta is lapping strong comps and continuing to print 20%+ growth on a $163B base. The Street's FY26 consensus expects deceleration to the mid-to-high teens.
2. Reality Labs operating loss (the capital commitment)
| FY22 | FY23 | FY24 | FY25 | |
|---|---|---|---|---|
| RL revenue ($B) | 2.2 | 1.9 | 2.1 | 2.2 |
| RL operating loss ($B) | (13.7) | (16.1) | (17.7) | (19.2) |
The loss widened each year — management's consistent framing is that RL spend is "a multi-year investment" — and the FY25 widening was primarily on generative AI research and AI silicon spending consolidated within the segment, not on VR / AR hardware itself. This is the single largest voluntary loss commitment among US mega-caps and is the principal reason the sell-side spread on META is wider than on the other FANG-adjacent names.
3. Capex and FCF
| FY22 | FY23 | FY24 | FY25 | |
|---|---|---|---|---|
| Capex ($B) | 32.0 | 27.3 | 39.2 | 69.7 |
| OCF ($B) | 50.5 | 71.1 | 91.3 | 115.8 |
| FCF ($B) | 18.4 | 43.8 | 52.1 | 46.1 |
| Capex intensity | 28% | 20% | 24% | 34.7% |
FY25 capex intensity at 34.7% of revenue is the highest in Meta's history — materially above FY22's 28% pandemic-tail peak. FCF compressed $6B YoY despite +$24B of OCF growth. The April 2026 analyst PT cuts center on FY26 capex guidance flowing from this base.
4. Capital return
| FY22 | FY23 | FY24 | FY25 | |
|---|---|---|---|---|
| Buybacks ($B) | 27.9 | 20.0 | 30.0 | 26.2 |
| Dividends ($B) | — | — | 5.1 | 5.3 |
| Total return ($B) | 27.9 | 20.0 | 35.1 | 31.6 |
Dividends (instituted FY24) held flat in dollar terms; buybacks pulled back modestly as capex rose. Total capital return is roughly $30B/year — large but less than FCF, with the balance building the cash position.
Market evaluation
Sell-side coverage (as of late April 2026). 25 analysts cover the stock.
| Rating | Count |
|---|---|
| Strong Buy | 1 |
| Buy | 20 |
| Hold | 4 |
| Sell | 0 |
Price targets. Median $845.54, range $700 (low) to $1,015 (high: Rosenblatt).
Recent analyst activity (Feb 23 through April 24, 2026). The period splits clearly into two: pre-April 2 (post-Q4 FY25 earnings) was dominated by PT raises, while post-April 2 (post-Q1 FY26 earnings and the FY26 capex guide) was dominated by cuts.
Largest April PT reductions:
- Rosenblatt $1,144 → $1,015 (−$129, still the Street high)
- Keybanc $855 → $760 (−$95)
- Wells Fargo $856 → $765 (−$91)
- BofA $885 → $820 (−$65)
- Stifel $820 → $805 (−$15)
The common theme in the cut notes: FY26 capex guide came in higher than the Street had modeled, and the FY26 FCF compression that implies. No rating changes in either direction (zero upgrades, zero downgrades) across the 60-day window — all actions were PT adjustments with ratings maintained.
Notable counter-move: UBS raised to $908 on April 21 (+$36), citing better-than-consensus ad impressions growth at Reels.
Buy-side positioning. META is a consensus mega-cap tech holding with slightly less crowded positioning than GOOGL / AMZN. Short interest remains below 1% of float.
FY25 corporate structure: two diverging threads into FY26
Two threads running in opposite directions define FY25 into FY26. The ad engine continues to print — Family of Apps +22% for a fourth year, ad impressions growth accelerating on Reels and Threads monetization, the core business sustaining its narrative with fewer visible cracks than GOOGL or AMZN. The second thread is capex and Reality Labs loss. FY25 capex of $69.7 billion (34.7% of revenue) is the highest capex intensity in company history, and the FY26 capex guide — delivered on the Q1 FY26 call in mid-April — implied continued elevation. The $19.2B Reality Labs operating loss, meanwhile, now contains not only VR / AR but also a material AI research and AI silicon spend load. The Street's April price-target reductions are the clearest market-priced reaction to these two threads colliding. Whether FY26 capex guide gets revised down at a future call (relieving pressure), or Reality Labs loss starts stabilizing as AI silicon spending rationalizes, are the two model-level debates the sell-side spread of $700-$1,015 is arguing about.