◆ What the Market Isn't Pricing In

Meta Platforms (META) — the variant view

The market still carries the 2022 money-pit fear while AI turned the ad engine into its biggest tailwind — though the June 2026 quarter showed what the spending costs.

📈 META valuation, revenue & earnings — P/E, P/S, revenue, EPS →

To see what the market may be missing in Meta, you have to remember where the fear came from. In late 2022 the stock collapsed to under eleven times earnings1 — a distressed multiple, the price of a company the market believed was being run into the ground: advertising stalling under Apple's privacy change, TikTok stealing the young, and tens of billions vanishing into a metaverse no one wanted. That frame — expensive social network, cash incinerated on a virtual fantasy — has never fully left the way Meta is discussed, even as the business underneath it changed. What follows is a note on where that frame and the facts have come apart, and it is not a price forecast.

Price to earnings (x)10.9x202230.6x202328.4x202527.8xSep 26Year-end multiples; stockanalysis.com, 22 September 2026
The fear multiple of 2022 is long gone, but the market pays less today than at the end of 2023.

Start with the multi-year record, because it is starker than the reputation. After the 2022 trough — revenue down one percent, net income halved to twenty-three billion2 — Meta did not wither; it roared. Revenue went from $117 billion to $135, then $165, then $201 billion by 20253; operating income nearly tripled from twenty-nine billion to eighty-three; and in the first half of 2026 advertising revenue grew 30%.4 This is not a declining franchise defending itself. It is one of the great advertising businesses on earth, reaccelerating.

And the engine of the reacceleration is the very thing the market files under 'risk': artificial intelligence. The same AI capex that made the stock wobble is already inside the results — better targeting, better ad ranking driven by Llama5, better content recommendation that lifted engagement — showing up in the June 2026 quarter as a 12% rise in the price per ad and a 14% rise in impressions at once.6 The market has been taught by a decade of headlines to read Meta's AI spending as a cost and a threat; the tape reads it, so far, as a tailwind to the most profitable machine the company owns.

Then there is the optionality the price barely credits. Meta reaches more than three billion people daily7, which makes it, almost by default, one of the best-positioned consumer-AI distributors on the planet — Meta AI is already among the most-used assistants simply by living inside apps everyone already opens. It owns Llama, a leading open model family, and its Ray-Ban AI glasses, whose daily users tripled in a year, may prove the form factor the clumsy VR headset never was. None of this — the largest built-in AI audience anywhere, a frontier model, a live hardware hit — is obviously in a multiple of about twenty-eight times trailing earnings.8

I will not pretend the bear has nothing, because it has plenty. The capital spending is staggering and its returns unproven — $130 to $145 billion in a single year, announced alongside eight thousand layoffs, and free cash flow of $784 million in the June 2026 quarter.9 Reality Labs still bleeds nineteen billion dollars a year into a metaverse that has not arrived. The recent earnings are noisy: the March quarter's profit included a one-time $8 billion tax benefit, the June quarter carried $2.4 billion of legal charges, and 2025's net income actually dipped below 2024's despite revenue climbing a fifth. The FTC is appealing the ruling that let Meta keep Instagram and WhatsApp.10 These are real, and a prudent owner keeps them in full view.

But the variant view does not require ignoring the risks; it requires only noticing that the market's mental model lags the machine. The frame is still '2022 Meta' — a threatened ad business torching money on a fantasy — while the reality is a reaccelerating advertising fortress that has turned AI from its feared disruptor into its sharpest tool, with a free option on consumer AI and glasses stapled on top. Why do I think there is a gap? Because the same market that paid eleven times earnings in the panic now pays about twenty-eight for a business growing revenue by more than a quarter and compounding its profits — a re-rating, yes, but a cautious one, still shadowed by the old fear. It is in the distance between that lingering fear and the accelerating fact that a patient owner earns their keep.

References
  1. Third-party estimateIn late 2022 the stock collapsed to under eleven times earnings — a distressed multiple, the price of a company the market believed was being run into the ground: advertising stalling under Apple's privacy change, TikTok stealing the young, and tens of billions vanishing into a metaverse no one wanted.
    Trailing P/E history (third-party market data) — ~11x at the late-2022 trough, ~24x by 2026 — 2022-2026 · source ↗
  2. ReportedAfter the 2022 trough — revenue down one percent, net income halved to twenty-three billion — Meta did not wither; it roared.
    Meta Platforms, Forms 10-K (FY2021-FY2025) — revenue $117B -> $201B; 2022 net income halved to $23B; Reality Labs losses ~$19.2B (2025), >$80B cumulative; dual-class structure — FY2021-FY2025 · publ. 2022-2026 · source ↗
  3. ReportedRevenue went from $117 billion to $135, then $165, then $201 billion by 2025; operating income nearly tripled from twenty-nine billion to eighty-three; and in the first half of 2026 advertising revenue grew 30%.
    Meta Platforms, Forms 10-K (FY2021-FY2025) — revenue $117B -> $201B; 2022 net income halved to $23B; Reality Labs losses ~$19.2B (2025), >$80B cumulative; dual-class structure — FY2021-FY2025 · publ. 2022-2026 · source ↗
  4. ReportedRevenue went from $117 billion to $135, then $165, then $201 billion by 2025; operating income nearly tripled from twenty-nine billion to eighty-three; and in the first half of 2026 advertising revenue grew 30%.
    Meta Q2 2026 results release (Form 8-K exhibit 99.1, 29 July 2026) - revenue $60,801M (+28%, +27% excluding currency); advertising $59,363M (+27%); Family of Apps income from operations $23,394M against $24,971M; Reality Labs revenue $431M and loss $4,619M; operating margin 31% against 43%; net income $15,848M (-14%); costs include $2.40B of legal charges and $1.18B of severance; ad impressions +14% and average price per ad +12%; DAP 3.60B (+3%); capital expenditures $31.08B; free cash flow $784M against $8,549M; long-term debt $83.66B; headcount 75,472; 2026 capex guided $130-145B (from $125-145B); H1 advertising $114,387M against $87,955M — Q2 2026 · publ. July 29, 2026 · source ↗
  5. ReportedThe same AI capex that made the stock wobble is already inside the results — better targeting, better ad ranking driven by Llama, better content recommendation that lifted engagement — showing up in the June 2026 quarter as a 12% rise in the price per ad and a 14% rise in impressions at once.
    Meta — Llama open-weight model family; AI-driven ranking and ads improvements credited on earnings calls — 2023-2026 · publ. 2023-2026 · source ↗
  6. ReportedThe same AI capex that made the stock wobble is already inside the results — better targeting, better ad ranking driven by Llama, better content recommendation that lifted engagement — showing up in the June 2026 quarter as a 12% rise in the price per ad and a 14% rise in impressions at once.
    Meta Q2 2026 results release (Form 8-K exhibit 99.1, 29 July 2026) - revenue $60,801M (+28%, +27% excluding currency); advertising $59,363M (+27%); Family of Apps income from operations $23,394M against $24,971M; Reality Labs revenue $431M and loss $4,619M; operating margin 31% against 43%; net income $15,848M (-14%); costs include $2.40B of legal charges and $1.18B of severance; ad impressions +14% and average price per ad +12%; DAP 3.60B (+3%); capital expenditures $31.08B; free cash flow $784M against $8,549M; long-term debt $83.66B; headcount 75,472; 2026 capex guided $130-145B (from $125-145B); H1 advertising $114,387M against $87,955M — Q2 2026 · publ. July 29, 2026 · source ↗
  7. ReportedMeta reaches more than three billion people daily, which makes it, almost by default, one of the best-positioned consumer-AI distributors on the planet — Meta AI is already among the most-used assistants simply by living inside apps everyone already opens.
    Meta, Q1 2026 earnings release + call (rev $56.3B +33%, fastest since 2021; FoA ads ~$55B at 41% op margin; NI $26.8B incl. an $8.03B one-time tax benefit; DAP 3.4B; CY2026 capex guide $125-145B) — Q1 2026 — quarter ended Mar 31, 2026 · publ. Apr 2026 · source ↗
  8. Third-party estimateNone of this — the largest built-in AI audience anywhere, a frontier model, a live hardware hit — is obviously in a multiple of about twenty-eight times trailing earnings.
    Market data (stockanalysis.com) - Meta closed at $736.60 on 22 September 2026, market capitalisation about $1.88 trillion, about 27.8 times trailing earnings and 22.9 times forward — 22 September 2026 · publ. September 23, 2026 · source ↗
  9. ReportedThe capital spending is staggering and its returns unproven — $130 to $145 billion in a single year, announced alongside eight thousand layoffs, and free cash flow of $784 million in the June 2026 quarter.
    Meta Q2 2026 results release (Form 8-K exhibit 99.1, 29 July 2026) - revenue $60,801M (+28%, +27% excluding currency); advertising $59,363M (+27%); Family of Apps income from operations $23,394M against $24,971M; Reality Labs revenue $431M and loss $4,619M; operating margin 31% against 43%; net income $15,848M (-14%); costs include $2.40B of legal charges and $1.18B of severance; ad impressions +14% and average price per ad +12%; DAP 3.60B (+3%); capital expenditures $31.08B; free cash flow $784M against $8,549M; long-term debt $83.66B; headcount 75,472; 2026 capex guided $130-145B (from $125-145B); H1 advertising $114,387M against $87,955M — Q2 2026 · publ. July 29, 2026 · source ↗
  10. ReportedThe FTC is appealing the ruling that let Meta keep Instagram and WhatsApp.
    Meta Form 10-Q, quarter ended 30 June 2026 - DAP 3.60B in June 2026 from 3.48B (+3%), the Q1 dip due to internet disruptions in Iran and restricted WhatsApp access in Russia; ARPP $16.86 (+24%); revenue by customer address US & Canada $23,863M, Europe $14,009M, Asia-Pacific $16,073M, Rest of World $6,856M; R&D $21,656M (+67%) including third-party AI token costs; FTC v. Meta: trial April-May 2025, judgment for Meta on 18 November 2025, FTC notice of appeal 20 January 2026; resellers serving China-based advertisers risk factor — Q2 2026 · publ. July 30, 2026 · source ↗
Sources
Generated September 23, 2026