METANASDAQThe short version
Meta Platforms, Inc.
Meta Platforms runs the Facebook, Instagram, WhatsApp and Messenger advertising engine — about 98% of revenue — while its founder directs tens of billions of dollars a year into AI infrastructure and the Reality Labs metaverse bet.
From a 2021 peak the shares fell 77% to an $89 close in November 2022, then re-rated past the old high to $790 by August 2025, and trade near $669 today.
$669
Share price
$1.7T
Market cap
$201B
FY2025 revenue
60.8%
Founder voting power
SwipeScroll▾
The statements
Revenue reaccelerated to $201B; a one-time tax charge, not the business, dipped 2025 earnings
FY2021 → FY2025as reported · $
Revenue$201.0B+22%
Operating margin41.4%−0.7pp
Net income$60.5B−3%
EPS$23.49−2%
Free cash flow$46.1B−15%
Open the full statements →As reported, FY2021–FY2025. Free cash flow is operating cash flow less capital expenditure.
- Top line. Revenue grew 22% in both 2024 and 2025 to $201.0B, and the operating margin climbed from 35% in 2023 to 41% as costs grew more slowly than sales.
- The earnings dip is tax, not trade. Reported net income slipped to $60.5B from $62.4B only because a one-time $15.9B charge lifted the tax rate to 30%; pretax profit rose 22%.
- Cash tells the fuller story. Operating cash flow rose 27% to $115.8B, but free cash flow fell to $46.1B as capital spending nearly doubled to $69.7B.
The signature finding
Meta commits more to compute than either rival — and alone earns nothing renting it out
The three hyperscalers, FY2025
| Company | Capex/rev | Ext. cloud rev | P/E |
|---|---|---|---|
| Meta | 35% | $0 | 28x |
| Alphabet | 23% | $58.7B | 33x |
| Amazon | 18% | $128.7B | 34x |
Alphabet and Amazon rent the same class of compute to outside customers under contract; Meta does not.
- The commitment. Meta has locked in the largest and fastest-growing non-cancelable infrastructure commitment of the three hyperscalers — $237.7 billion by March 2026, up from $32.1 billion in September 2024, plus $182.9 billion of leases not yet commenced —
- yet at 34.7% capex intensity it is the only one of the three with no third-party cloud business, so unlike AWS ($128.7B revenue) and Google Cloud ($58.7B revenue) every server it commits to must earn its return entirely from Meta's own advertising engine plus an unproven metaverse and superintelligence bet.
- It can afford it. The build is self-funded from $115.8B of operating cash flow against a net-cash balance sheet, so the concentration is a question about the quality of the return, not the ability to pay.
What the price implies
At $669 the price already assumes only about 5–6% long-run earnings growth
~5–6%
Implied 10-yr earnings growthreverse DCF at 9%
20.4x
Forward P/E (2026E)
26%
Consensus 2026 revenue growth
A reverse discounted-cash-flow on consensus 2026 earnings at a 9% cost of equity and a 15–18x exit multiple; price $669.21.
- The bar. The price implies only about 5-6% long-run earnings growth, a modest bar on an earnings basis, but one that holds only if the AI capex build lets free cash flow re-converge with earnings.
- The counter. On earnings the bar is easy — consensus sees ~26% revenue growth in 2026, and management guides 2026 operating income above 2025's $83.3B; the catch is that free cash flow was $46.1B, under 40% of operating cash flow.
- Why it is conditional. The reverse-DCF treats reported earnings as distributable cash — they only become cash once capital intensity normalizes and the build stops consuming the difference.
One engine, one bet
One advertising engine funds a metaverse bet that has lost about $88B since 2019
FY2025 revenue vs operating income by segment ($B)
The two segments net to $83.3B of operating income at a 41% margin.
- The engine. Family of Apps — Facebook, Instagram, WhatsApp, Messenger — turned 3.58 billion daily users into $198.8B of revenue and $102.5B of operating income, a 52% margin.
- The bet. Reality Labs earned $2.2B and lost $19.2B in 2025; its operating losses total roughly $88B since 2019, more than the whole company's 2025 operating profit.
- Discretionary and unchecked. The metaverse and AI spend is an owner's wager the founder funds at will — no governance mechanism stops it, and no filing discloses a return milestone.
The moat
The ad margin fell to 37% under Apple's privacy shock, then rebuilt to 52%
Family of Apps operating margin
Apple's App Tracking Transparency and a soft ad market hit in 2022; AI-driven targeting rebuilt pricing power by 2025.
- Survived a real stress test. In 2022 Apple's tracking change cut Meta's targeting signal and price per ad fell 16%; advertising revenue declined 1%, the only annual fall in its history.
- Rebuilt from the inside. New AI ad-ranking models turned price per ad positive again — up 9% in 2025 — and revenue through automated ad tools reached a $60B annual run-rate.
- The catch. The recovery was bought with the same AI and infrastructure spend now compressing free cash flow and the segment margin; the moat and the bet share a budget.
The capex ramp
Capex has raced ahead of depreciation — the cost wave has barely landed
Capex vs depreciation & amortization ($B)
2026 capex shown at the $115–135B guidance midpoint. Most 2025 spend is not yet in service, so depreciation lags.
- Guided to roughly double. Capital spending nearly doubled to $69.7B in 2025 and is guided to $115–135B in 2026 — more than the company's entire 2025 operating income.
- The charge is coming. Depreciation reached $18.6B in 2025, a fraction of the assets being built; management calls it the single largest driver of 2026 expense growth.
- Now committed, not optional. A rising share of the build is locked into multi-year contracts — the capital outside holders cannot vote on is increasingly capital management itself cannot unwind.
Founder control
The founder owns about 13% of Meta and controls 61% of the vote
Zuckerberg: economic ownership vs voting power
Class B shares carry ten votes to Class A's one. An 88% outside-shareholder vote for one-share-one-vote failed in 2025.
- Skin in the game. Zuckerberg holds about $229B of stock, takes a $1 salary and no equity; his $25.1M of pay is almost all personal security — 0.04% of net income.
- But no outside check. Ten-vote Class B shares convert 13.5% of the economics into 60.8% of the vote, so no shareholder can gate the AI and Reality Labs spend.
- The check that remains is behavioral. The evidence it is used well is capital returns and discipline — $31.6B came back to owners in 2025; a cut while capex climbs would weaken the case.
Forward estimates
Consensus sees revenue past $300B by 2027, with earnings growth cooling
Revenue: reported and consensus estimate ($B)
2026–2027 are analyst consensus. Consensus diluted EPS is $32.84 (2026) and $34.99 (2027).
- Reacceleration continues. Consensus has revenue near $253B in 2026 and $302B in 2027 — growth of about 26% and 20% — extending the two prior years of 22% gains.
- Earnings growth decelerates. Headline 2026 EPS of $32.84 looks like a 40% jump, but most is the 2025 tax base normalizing; against a tax-adjusted ~$30 the step-up is closer to 10%.
- Margin compression penciled in. By 2027 EPS growth slows to about 7% even as revenue grows 20% — the depreciation from the build arriving in the income statement.
Downside floor
A net-cash balance sheet and $116B of operating cash leave little bankruptcy risk
$23B
Net cash (cash & securities less debt)
$115.8B
Operating cash flow, FY2025
$31.6B
Returned to owners in 2025
$81.6B of cash and marketable securities against $58.7B of long-term debt at year-end 2025.
- Hard to go bust. A business earning $83B of operating income with roughly $23B of net cash carries essentially no near-term bankruptcy risk; the risk here is overspending, not insolvency.
- Self-funded returns. Meta bought back $26.3B of stock and paid $5.3B of dividends in 2025 while funding the entire capital build from internal cash.
- Debt is new and worth watching. A November 2025 note issuance roughly doubled long-term debt to $58.7B as net cash shrinks to fund AI infrastructure.
Legal overhang
The big legal threats are to value, not solvency — the breakup case already receded
Largest legal demands vs FY2025 net income
| Matter | Largest demand | % of FY25 NI |
|---|---|---|
| New Mexico AG — abatement sought | $3.7B | 6.1% |
| EU Marketplace fine (on appeal) | $0.9B | 1.4% |
| Spain — publisher damages | $0.6B | 1.0% |
| New Mexico — civil penalty | $0.4B | 0.6% |
The FTC breakup suit was won at trial in Nov 2025 and is on appeal; most items are on appeal and not final.
- Fines are noise. Penalties paid or imposed to date total roughly $2B, most on appeal — trivial against $60.5B of net income and $115.8B of operating cash flow.
- The material front is Europe. Consent rules could permanently lower monetization on the ~23% of revenue Meta earns from Europe, unlike a one-time fine.
- A survivable tail. Youth-harm damages could reach the tens of billions over years of trials, and the balance sheet does not yet reserve it — a valuation risk, not an existential one.
Valuation
Six defensible multiples on one price — the cash-versus-earnings gap is the debate
One price, six multiples (at $669)
The gap between a 15x cash-flow multiple and a 37x free-cash-flow multiple is the capital build sitting between operating and free cash.
- Cheapest of the cohort. On earnings lenses Meta trades near 20x forward and 17x EV/EBITDA — priced more conservatively than Alphabet or Amazon, not like a speculation.
- Dear on cash. At $46.1B of free cash flow the price-to-free-cash-flow multiple is 37x and the yield just 2.7%; the build suppresses the cash owners actually receive.
- Fairly priced, not a bargain. The clean 2022 discount — about 10x a depressed profit — is gone; today's cushion comes from the cash the ad engine throws off, not a low price.
The re-rating
A fallen star that already recovered — the 2022 bargain is gone
Daily close. From a 77% drawdown in 2022 the shares re-rated past their old high to $790, and trade near $669.
- The fall is recovered. The shares dropped 77% to an $89 close in November 2022, then made new highs at $790 by August 2025; near $669 they sit above their own five-year average multiple.
- The margin of safety moved. At $89 the stock traded near 10x a depressed profit; at $669 the company is far stronger, but the deep-pessimism discount a value buyer hunts for is no longer in the price.
What to watch
A durable, fairly-priced franchise whose payoff waits on a capital build no one can yet prove
- 01Capex as a share of revenue (35% in 2025, guided near 49% in 2026): turning toward the ~20–23% peer range is bullish; holding near 50% is the bear path.
- 02Free cash flow versus operating cash flow ($46.1B vs $115.8B): the gap narrowing signals the build is peaking; widening confirms the payoff is deferred.
- 03Family of Apps operating margin (52% in 2025, down from 54%): a reading below 50% would confirm the margin has peaked.
- 04Capital returned to owners ($31.6B in 2025): a cut while capex climbs would weaken the founder-alignment case.
This distills a guided study built chapter by chapter, from the financial statements through valuation and what to watch.
Compiled from the full report · 2026-07-11 · For information, not investment advice.