Data sourced exclusively from Meta Platforms 8-K Exhibit 99.1 earnings releases (SEC-filed), 10-Q filings, and the FY2025 Form 10-K (filed 2026-01-29) · All figures GAAP unless noted
Meta’s FY2025 results confirm a company running two businesses on entirely different trajectories under one roof: a Family of Apps advertising engine that generated $198.8 billion of revenue and $102.5 billion of operating income, and a Reality Labs division that has now absorbed roughly $77.0 billion in cumulative losses since 2021 with no disclosed path to breakeven. The more consequential shift is on the balance sheet — capex jumped to 34.7% of revenue in FY2025, gross debt more than doubled to $58.7 billion, and free cash flow fell 16% even as operating profit rose 20%. That trend has only accelerated into FY2026: Q2 2026 capex alone was $31.1 billion, quarterly free cash flow collapsed to $784 million, and full-year 2026 capex guidance now sits at $130–145 billion. Every figure below is drawn directly from primary SEC filings.
Annual figures are drawn from Meta’s FY2021–FY2025 8-K Exhibit 99.1 earnings releases and the FY2025 Form 10-K (filed 2026-01-29). Quarterly figures are drawn from the Q1 2026 and Q2 2026 8-K Exhibit 99.1 releases, the Q2 2026 Form 10-Q, and the corresponding earnings-call transcripts. Per this site’s zero-tolerance sourcing standard, any figure that could not be independently verified at the note level would be marked explicitly rather than estimated — no such figure remains flagged as of the latest update.
In Plain Terms
How Meta Actually Makes Money
Nearly every dollar Meta earns comes from one mechanic — showing ads inside Facebook, Instagram, Messenger, and WhatsApp, and charging advertisers per ad shown and how much they’re willing to bid in an auction. In FY2025, that engine alone generated $198.8 billion of the company’s $201.0 billion in total revenue and threw off $102.5 billion of profit. Everything else — the VR headsets, the AI glasses, the metaverse bet (Reality Labs) — is a side project that loses about $19 billion a year and is entirely paid for out of the advertising profits. Think of Family of Apps as a very profitable toll booth on people’s attention, and Reality Labs as a long-term, self-funded science project riding on top of it.
Why It Has a Competitive Advantage
Meta’s edge is scale feeding on itself. With 3.6 billion people using at least one of its apps daily, it has more behavioral data to target ads well, and more advertiser demand competing for that inventory — both of which push up what advertisers are willing to pay. That scale also means Meta can afford to spend nearly $70 billion a year on data centers and AI chips (and $130–145 billion planned for 2026) while still holding more cash than debt — something almost no other consumer-internet company can match. The risk sitting underneath that story is that regulators, especially in Europe, and courts, on youth-safety litigation, have real, scheduled power to interrupt it, and that the spending itself is now growing so fast it is starting to eat into the cash the business throws off — free cash flow was just $784 million in the latest quarter, down from $12.4 billion the quarter before.
Core Financial Data (FY2021–FY2025)
All figures in $ millions unless noted; per-share and margin figures as stated.
| Metric | FY2021 | FY2022 | FY2023 | FY2024 | FY2025 |
|---|---|---|---|---|---|
| Income Statement | |||||
| Revenue | 117,929 | 116,609 | 134,902 | 164,501 | 200,966 |
| — Family of Apps revenue | 115,655 | 114,450 | 133,006 | 162,355 | 198,759 |
| — Reality Labs revenue | 2,274 | 2,159 | 1,896 | 2,146 | 2,207 |
| Operating profit | 46,753 | 28,944 | 46,751 | 69,380 | 83,276 |
| — FoA operating income | 56,946 | 42,661 | 62,871 | 87,109 | 102,469 |
| — RL operating loss | (10,193) | (13,717) | (16,120) | (17,729) | (19,193) |
| Operating margin | 39.6% | 24.8% | 34.7% | 42.2% | 41.4% |
| Net income | 39,370 | 23,200 | 39,098 | 62,360 | 60,458 |
| Balance Sheet | |||||
| Total stockholders’ equity | 124,879 | 125,713 | 153,168 | 182,637 | 217,244 |
| Gross debt (long-term Notes) | 0 | 9,923 | 18,385 | 28,826 | 58,744 |
| Cash and cash equivalents | 16,601 | 14,681 | 41,862 | 43,889 | 35,873 |
| Short-term investments | 31,397 | 26,057 | 23,541 | 33,926 | 45,719 |
| Net debt | (47,998) | (30,815) | (47,018) | (48,989) | (22,848) |
| Net debt / EBIT | net cash | net cash | net cash | net cash | net cash |
| Cash Flow & Capex | |||||
| Capex | 18,567 | 31,431 | 27,266 | 37,256 | 69,691 |
| Capex / Revenue | 15.7% | 27.0% | 20.2% | 22.6% | 34.7% |
| Operating cash flow | 57,683 | 50,475 | 71,113 | 91,328 | 115,800 |
| Free cash flow (non-GAAP) | 38,439 | 18,439 | 43,010 | 52,103 | 43,585 |
| FCF conversion (/ Net income) | 97.6% | 79.5% | 110.0% | 83.6% | 72.1% |
| FCF conversion (/ Op. profit) | 82.2% | 63.7% | 92.0% | 75.1% | 52.3% |
| Returns | |||||
| Invested capital | 76,881 | 94,898 | 106,150 | 133,648 | 194,395 |
| Effective tax rate (disclosed) | 17% | 19% | 18% | 12% | 30%² |
| NOPAT | 38,805 | 23,445 | 38,336 | 61,054 | 58,293² |
| ROIC (NOPAT / Invested Capital) | 50.5% | 24.7% | 36.1% | 45.7% | 30.0%² |
¹ FY2025 total stockholders’ equity of $217,244M is derived from disclosed total assets ($366,022M) less total liabilities ($148,778M), both stated in the FY2025 10-K; FY2021–FY2024 figures are stated directly on the face of the respective 10-Ks / 8-K Ex-99.1 balance sheets.
² FY2025’s 30% effective tax rate includes a one-time valuation-allowance charge tied to the One Big Beautiful Bill Act enacted Q3 2025. Management states that absent this item, the FY2025 effective rate would have been 13%. On a normalized 13% rate, FY2025 NOPAT = $72,450M and ROIC = 37.3%. Both figures are shown so the reader can select the appropriate basis; this is a disclosed, verifiable adjustment, not an estimate.
Against the equity base above, FY2025 net income of $60,458M implies ROE of ~27.8% (net income / average or year-end equity, depending on convention used) — a useful cross-check alongside ROIC, since equity has nearly doubled since FY2021 ($124.9B → $217.2B) largely through retained earnings, even as the company began raising debt for the first time in 2022.
Capital Structure — Q2 2026 8-K (Period Ended June 30, 2026)
Fundamentals Visualized
The same FY2021–FY2025 figures from the table above, charted to make the trend lines and inflection points easier to read at a glance. Click any chart to expand it.
Growth reaccelerated every year since the FY2022 dip, compounding to +72% cumulative FY21→FY25.
Recovered sharply post-2022 “year of efficiency,” then plateaued just above 41% as AI infrastructure costs offset further leverage.
FY2025 FCF fell for the first time since FY2022, despite record revenue and operating profit — capex is now the swing factor.
FY2025’s 30.0% is GAAP-reported; normalized for the one-time tax charge, ROIC would be 37.3% — still the lowest of the five years either way.
FY2025’s 34.7% is the highest in the five-year window by a wide margin — the single most consequential line in this dataset.
Equity has funded most of the growth, but FY2025 is the first year debt grew faster than equity — gross debt is now 27% of equity, up from 0% in FY2021.
FY2025 is essentially flat vs. FY2024 despite +22% revenue growth — the first time net income hasn’t grown with the top line since FY2022.
Still net cash every year, but the FY2025 drop is the sharpest of the period — more than half the net cash cushion disappeared in one year.
FY2025’s 72.1% is the lowest of the five years — a widening gap between reported profit and the cash it actually converts to.
The ROIC denominator has grown 2.5x since FY2021, while the FY2025 numerator (NOPAT) grew far more slowly — the main mechanical driver of ROIC’s decline.
Latest Quarterly Results & Guidance (Q1–Q2 FY2026)
Figures below are drawn from the Q1 2026 and Q2 2026 8-K Exhibit 99.1 releases, the Q2 2026 Form 10-Q, and both quarters’ earnings-call transcripts (Q1 call: April 29, 2026; Q2 call: July 29, 2026). All figures in $ millions unless noted.
| Metric | Q1 FY2026 | Q2 FY2026 |
|---|---|---|
| Income Statement | ||
| Total revenue | 56,311 | 60,801 |
| — Family of Apps revenue | 55,909 | 60,370 |
| — FoA advertising revenue | 55,024 | 59,363 |
| — Reality Labs revenue | 402 | 431 |
| Revenue growth YoY | +33% | +28% |
| Total costs and expenses | 33,439 | 42,026 |
| Operating income | 22,872 | 18,775 |
| Operating margin | 41% | 31% |
| Effective tax rate | (23)%⁴ | 16% |
| Net income | 26,773⁴ | 15,848 |
| Diluted EPS | $10.44⁴ | $6.18 |
| Cash Flow & Capex | ||
| Capital expenditures (incl. finance leases) | 19,840 | 31,081 |
| Free cash flow | 12,386 | 784 |
| Balance Sheet | ||
| Cash + marketable securities (period-end) | 81,180 | 90,260 |
| Long-term debt (period-end) | 58,748 | 83,660 |
| Total stockholders’ equity (period-end) | 243,681 | 261,221 |
| Headcount (period-end) | 77,986 | 75,472 |
⁴ Q1 2026 net income and EPS include an $8.03 billion one-time income tax benefit tied to updated U.S. Treasury guidance (Notice 2026-7) on the tax treatment of previously capitalized R&D expenditures. Absent this item, Q1 2026 effective tax rate would have been 14 percentage points higher and diluted EPS would have been $3.13 lower (i.e., ~$7.31).
Q2 2026 Cost Detail
Headcount ended Q2 2026 at 75,472, down 3% from Q1 2026, with most of the May reduction expected to fall out of headcount by end of Q3 2026.
Guidance — Q3 2026 and Full-Year 2026
Update — Youth Litigation, Aug 26, 2026
Meta reached a settlement with 48 state attorneys general (plus a separate, related Cambridge Analytica settlement with a broad state coalition) on a bipartisan teen-safety framework, disclosed via a dedicated conference call two days before this quarter’s close. This materially updates, but does not close, the youth-litigation risk flagged above.
Key terms: a 2-hour/day combined Instagram+Facebook time limit for teens (above the ~1 hour/day actual average usage cited by the CFO), applying only to Instagram and Facebook — Meta AI is explicitly excluded as “not a social media app.” Most obligations run 10 years; the time limit specifically runs 5 years, extending if TikTok and YouTube adopt matching terms. Oversight is via an independent auditor, which management distinguishes from a DOJ-style “monitor” with broad remedial powers. Q3 2026 revenue guidance is unchanged and management states it does not expect a material engagement/monetization impact, though incremental compliance opex and headcount have not been disclosed.
What remains open: school district litigation and individual user litigation are explicitly not covered by this settlement and continue — Meta argues plaintiffs there bear a causation burden the state AGs did not, and cites Section 230/First Amendment defenses it expects to litigate further. Florida’s teen-safety claim is unresolved with no near-term trial date. Roughly 30% of the total potential payment remains contingent on a competitor decision (TikTok/YouTube joining) outside Meta’s control.
Separately, the European Commission’s April 2025 decision found Meta’s “subscription for no ads” model non-compliant with the EU Digital Markets Act; further forced modifications “could result in a materially worse user experience for European users and a significant impact to our European business and revenue.” This matter is unaffected by the August 2026 AG settlement and remains open.
Segment Analysis — The Cash Machine
Meta reports two segments: Family of Apps (FoA) and Reality Labs (RL) (10-K, Note 15 / Item 1 Business).
Family of Apps
Facebook, Instagram, Messenger, WhatsApp, Threads, and Meta AI — the company’s sole cash machine. Momentum has continued into FY2026, with Q2 ad revenue growth driven by 14% impression growth and a 12% increase in average price per ad.
Monetization mechanics: “Currently, we generate substantially all of our revenue from selling advertising placements on our family of apps to marketers” (10-K Item 1). Revenue is a function of ad impressions delivered and average price per ad. In FY2025, ad impressions grew 12% y/y and average price per ad grew 9% y/y — both levers contributing to growth, a healthier mix than years such as FY2023, where impression growth (+28%) was offset by falling price per ad (−9%) as Reels monetization diluted blended pricing. Family daily active people (DAP) reached 3.58 billion on average for December 2025, and 3.6 billion as of the Q2 2026 call.
Reality Labs
VR/AR hardware (Meta Quest), software (Horizon), AI/smart glasses (Ray-Ban Meta, Oakley Meta, Meta Ray-Ban Display), and long-term AR research (Orion prototype). Q2 2026 RL revenue was $431 million, up 16% YoY on strong AI-glasses sales partially offset by weaker Quest headset sales.
Management is explicit that this is not self-funding: “our ability to support our efforts to build the next computing platform is dependent on generating sufficient profits from other areas of our business” (10-K Item 1), and RL is expected to “continue to operate at a loss for the foreseeable future,” with 2026 losses “similar to 2025.”
Family of Apps advertising is the entirety of Meta’s economic engine — a two-sided network-effects business monetized almost purely through auction-based ad pricing. Reality Labs is a fully cross-subsidized, multi-year bet: its ~$77B of cumulative losses have been entirely absorbed by FoA’s operating profit, which itself has compounded from $56.9B (2021) to $102.5B (2025).
Competitive Advantage
Scale of the network / data flywheel
Family DAP grew from 2.82 billion (Dec 2021) to 3.6 billion (mid-2026) — a base larger than any single competitor’s, which mechanically improves both ad targeting (more behavioral data per advertiser dollar) and ad auction liquidity (more advertisers bidding against more inventory, raising average price per ad, which rose 12% y/y in Q2 2026 alone).
Operating leverage in the core business
FoA operating margin expanded from an implied ~49% (2021) to ~52% (2025) even as the company absorbed AI infrastructure cost growth — evidence that incremental ad revenue drops through at high marginal profitability once the infrastructure and audience base exist.
Balance-sheet capacity to out-invest rivals
Meta funded $69.7 billion of FY2025 capex and a cumulative ~$77 billion of Reality Labs losses while remaining in a net cash position every year from 2021–2025 — even after a $30 billion bond raise in November 2025, net debt was still negative $22.8 billion at YE2025. Few advertising or consumer-internet competitors can self-fund AI infrastructure at this scale without diluting shareholders or stressing the balance sheet.
Multi-app optionality
Four products (Facebook, Instagram, Messenger, WhatsApp) each independently exceed one billion users, giving Meta several monetization surfaces (Feed, Stories, Reels) to shift ad load and engagement into as user behavior evolves — management explicitly discloses it is “promoting Reels” despite Reels monetizing at a lower rate, i.e., managing a portfolio of surfaces rather than depending on one.
These are structural, network-effect and scale-economics advantages, not durable regulatory or IP moats — the 10-K’s own Risk Factors section is candid that Meta “faces significant competition in every aspect of our business,” including from TikTok for engagement and from Apple/Google as gatekeepers of the mobile ad-targeting stack.
Red Flag Log
StructuralReality Labs is a structural, non-shrinking cash drain. Cumulative operating losses ≈ $77.0 billion since 2021, with FY2025 losses ($19.2B) the highest yet, and management guiding 2026 RL losses “similar to 2025” with no path to segment profitability disclosed. (10-K Item 1; Item 1A Risk Factors — “investments in Reality Labs reduced our 2025 overall operating profit by approximately $19.19 billion.”) This is a permanent tax on FoA’s profits with no disclosed break-even date; it depresses consolidated operating margin, ROIC, and FCF conversion every year.
Capital intensityCapex and debt are both scaling sharply faster than revenue, and the trend has accelerated into FY2026. Capex/Revenue jumped from 22.6% (FY2024) to 34.7% (FY2025); gross debt more than doubled in one year ($28.8B → $58.7B) via a new $29.9B note issuance; FCF fell from $52.1B to $43.6B even as FY2025 operating profit rose 20%. Into FY2026, the pattern has intensified rather than stabilized: Q2 2026 capex alone was $31.1 billion, funded in part by a second debt raise ($24.9B net proceeds in Q2 2026 alone, per the 8-K), taking debt to $83.7 billion by quarter-end — two separate debt issuances within roughly eight months. Quarterly free cash flow collapsed to just $784 million (from $12.4 billion in Q1 2026), even as FY2026 capex guidance has now been revised upward three times in a row, most recently to $130–145 billion.
Regulatory / legalYouth litigation partly resolved, but with a real cash cost and real residual exposure. On August 26, 2026, Meta settled with 48 state attorneys general (plus a related Cambridge Analytica settlement) on a bipartisan teen-safety framework, recognizing a ~$10 billion GAAP legal accrual in Q3 2026 (NPV of $12.7 billion in guaranteed payments over 10 years), with up to $18.0 billion total if TikTok and YouTube also join. Florida (trial not expected before 2028) and New Mexico (already tried, judgment on appeal) remain unsettled, and school-district and individual-user litigation are explicitly untouched by this agreement, with Meta stating it expects to keep litigating those on Section 230/First Amendment grounds. Separately, the EU Digital Markets Act non-compliance finding on Meta’s “subscription for no ads” model (April 2025) remains open and unaffected by the AG settlement, with management continuing to flag potential “significant impact to our European business and revenue.”