META Platforms Intrinsic Value 2026: Revenue Hit $200B — So Why Is EPS Flat?
Meta Platforms grew revenue 22% in FY2025. Operating income grew 41%. The company crossed $200B in annual revenue for the first time in its history.
And earnings per share went down.
FY2024 EPS: $23.86. FY2025 EPS: $23.49. A year of record revenue and record margins, and shareholders earned slightly less per share.
If that sounds like a contradiction, it is — but only on the surface. Understanding what actually happened in FY2025 is the key to understanding whether META at ~$639 is cheap, fairly valued, or a value trap.
Not financial advice. All figures from Meta's FY2025 annual report and company filings.
The Numbers First
FY2025 — Meta's best operating year ever:
| Metric | FY2025 | FY2024 | YoY |
|---|---|---|---|
| Total revenue | $200.97B | $164.5B | +22% |
| Operating margin | 41.4% | 35.7% | Record high |
| Operating income | ~$83B | ~$58.7B | +41% |
| R&D spend | $57.4B | $43.9B | +31% |
| Full-year EPS (diluted) | $23.49 | $23.86 | -1.6% |
| Daily active users (Family) | 3.35–3.54B | ~3.27B | +8% |
| Reality Labs operating loss | -$19.2B | -$17.7B | Widening |
The headline: operating income grew 41% on 22% revenue growth. That's exceptional operating leverage — every dollar of new revenue dropped through to profit at a higher rate. Meta's 2025 was, by operating metrics, the best year in the company's history.
The anomaly: net income was $60.5B, actually down from $62.4B in FY2024 despite the significantly higher operating income.
The EPS Paradox: What Actually Happened in Q3 2025
This is the question that matters most for anyone modeling META's trajectory.
In Q3 2025, Meta reported:
- Revenue: $51.2B (+26% YoY) — excellent
- Operating income: $20.5B (40% op margin) — excellent
- Net income: $2.7B (5.3% net margin) — catastrophic
The gap between $20.5B in operating income and $2.7B in net income — roughly $17.8B — was driven by a large one-time non-operating charge in Q3. Based on the pattern of ongoing litigation (FTC antitrust proceedings, EU DMA enforcement actions, and the Meta Pixel class action settlements), the most likely source is a tax settlement or regulatory fine that hit entirely in the third quarter.
The critical point: this is not a structural earnings problem. In Q4 2025, Meta earned $22.8B in net income on $59.9B in revenue — its highest-ever quarterly net income. The Q3 anomaly is fully isolated.
What the annual EPS ($23.49) would look like without the Q3 charge: normalize Q3 to a typical net margin (~30%), and quarterly net income would have been ~$15.4B instead of $2.7B. Adjusted full-year EPS: approximately $28.50–29.00 — implying +19–21% growth versus FY2024.
This is the story the market is not pricing correctly. Surface EPS was flat. Normalized EPS grew ~20%. The underlying business is accelerating, not stalling.
Meta Is Growing Faster Than Apple — At a Lower Multiple
| Stock | Forward P/E | PEG | Revenue Growth | Operating Margin |
|---|---|---|---|---|
| META | ~20x | 0.91 | +22% | 41% |
| MSFT | ~22x | 1.61 | +17% | 47% |
| GOOGL | ~23–26x | 1.70 | +15% | 32% |
| AMZN | ~28x | 1.49 | +12% | 12% |
| AAPL | ~29x | 2.76 | +10% | 31% |
| TSLA | ~172x | 3.93 | -3% | 7% |
The PEG ratio divides P/E by earnings growth rate. Below 1.0 = potentially getting growth for free. META's PEG of 0.91 is the second-lowest in the Magnificent 7, trailing only NVDA.
Apple's PEG is 2.76. You are paying three times as much per unit of earnings growth for Apple as for Meta — and Apple is growing revenues at less than half Meta's rate.
The market is aware of this mismatch. The discount exists because of what Meta plans to do with its cash in 2026.
The AI Engine Nobody Talks About: Andromeda
Before the bear case, it's worth understanding why Meta's operating margins expanded from 35.7% to 41.4% in a single year.
The answer is Andromeda — Meta's AI-powered ad delivery engine, which completed global rollout in October 2025.
Andromeda doesn't show advertisers their pre-made ads. It dynamically generates and optimizes ad creative in real-time based on user context, then continuously refines based on response signals. The result: 20–35% higher ROAS (Return on Ad Spend) compared to traditional ad delivery.
Higher ROAS → advertisers pay higher CPMs → more revenue on the same 3.5B daily users → without proportionally higher costs. That's where the 41% margin comes from.
1 million+ advertisers created 15 million+ ads using Meta's AI generation tools in a single month. Andromeda is fully deployed and already generating the margin expansion visible in the FY2025 numbers.
The compounding dynamic: Andromeda improves with every ad impression across 3.5B daily users. Meta's data advantage makes it impossible for a competitor to replicate from scratch — the model improves faster the more data it sees, and no competitor has Meta's data volume.
This is the AI story inside Meta that doesn't get covered because everyone focuses on Reality Labs.
The Bear Case: $125 Billion
Meta's FY2026 capital expenditure guidance is ~$125 billion — nearly double 2025's $70–72B spend.
The bear argument: Meta already spent ~$60B on Reality Labs with minimal commercial traction. The $125B AI infrastructure bet echoes the same "invest now, figure out ROI later" logic. If this capex doesn't yield proportional revenue — from Llama licensing, Meta AI subscriptions, or continued ad performance gains — free cash flow gets permanently impaired.
EV/FCF at ~34x is already elevated because capex suppresses cash flow. Bears say the P/E looks cheap but the true cash economics don't.
Why the bull case answers this directly:
CFO Susan Li stated explicitly that 2026 operating income is expected to exceed 2025 levels despite the near-double capex. The core Family of Apps business generates $83B in operating income — more than enough to fund $125B in infrastructure investment while still growing profit. This isn't speculative; the math works.
The $125B breaks down roughly:
- AI compute for ad targeting (Andromeda infrastructure — direct, proven revenue impact)
- Meta AI / Llama 4 infrastructure (future monetization potential)
- Reality Labs hardware (the contentious piece)
Two of those three categories have clear or emerging revenue justification. The Reality Labs drag ($19.2B in annual losses) is fully visible in every analyst model — it's a known cost, not a hidden risk.
DCF: What Is META Actually Worth?
Inputs
- Normalized FY2025 EPS: ~$28.50 (adjusting out the Q3 one-time charge)
- FY2026E EPS consensus: ~$28–32 (based on operating income guidance and capex headwinds)
- Discount rate: 10% (ad-dependent revenue, advertising cyclicality risk)
- Terminal growth rate: 3%
Scenarios
| Scenario | Driver | Implied Fair Value |
|---|---|---|
| Bear | Ad recession, PEG compresses to 15–18x forward EPS | $450–520 |
| Base | Andromeda drives sustained 20%+ EPS growth, 22–25x P/E | $680–780 |
| Bull | Capex ROI shows up in FCF, PEG re-rates to 1.5x | $850–950 |
At ~$639, the stock is trading below base case.
Independent DCF models converge around $709 intrinsic value. Analyst consensus (67 analysts) averages $838–935 — implying 31–46% upside from current levels.
What the Market Is Pricing In
At ~$639 on ~$31 forward EPS, the market is applying roughly 20x forward earnings. For a business growing EPS at 20%+, with a PEG of 0.91 and 41% operating margins, this implies the market is pricing in either an ad cycle recession or FCF compression from capex.
Neither is irrational. But if Meta sustains 20%+ earnings growth for five years — which normalized FY2025 trajectory supports — the current price represents a significant discount to intrinsic value.
The Three Risks Worth Taking Seriously
1. Capex ROI timeline $125B is a bold commitment. If H1 2026 revenue growth decelerates while FCF stays compressed, multiple compression continues regardless of strong EBITDA. The market's patience for "invest now, monetize later" has limits that show up in stock price before they show up in fundamentals.
2. Privacy regulation and data targeting erosion Andromeda's advantage depends on Meta's ability to use behavioral data at scale. EU DMA enforcement, US federal privacy legislation, or new iOS/Android restrictions on data collection would directly impair targeting performance — the most structural risk to the bull case that doesn't appear in quarterly numbers until it suddenly does.
3. Advertising cyclicality ~97% of Meta's revenue is digital advertising — the most cyclical revenue stream in tech. In a recession, ad budgets are cut before headcount. Meta's 22% growth could decelerate sharply and quickly. The 41% operating margin provides cushion on the earnings line, but revenue growth would slow.
Run Your Own META Valuation
The scenarios above use our assumptions on EPS growth and discount rate. Your view on Andromeda's durability, the capex ROI timeline, and ad market cyclicality will differ from ours.
Run a free META valuation on EvidInvest →
Frequently Asked Questions
What is Meta's intrinsic value in 2026?
Base-case DCF models converge at approximately $680–780. Analyst consensus (67 analysts) averages $838–935. At ~$639, the stock trades at a discount to base case — unusual for a company with 41% operating margins and 22% revenue growth. The key input uncertainty is whether the $125B FY2026 capex generates adequate returns.
What is Meta's PEG ratio?
0.91 — below 1.0. Peter Lynch's rule: PEG below 1.0 suggests the market may be underpricing earnings growth relative to the multiple. For context, Apple's PEG is 2.76. Meta is growing revenues at more than twice Apple's rate and trading at a cheaper forward P/E.
Why was Meta's FY2025 EPS flat despite record revenue and margins?
A large one-time non-operating charge in Q3 2025 — likely a tax settlement or regulatory fine — drove Q3 net income down to $2.7B despite $20.5B in operating income. Remove that charge and normalized FY2025 EPS would have been approximately $28.50–29.00, implying ~20% growth vs. FY2024. The market is pricing the surface EPS number; the normalized trajectory is significantly more positive.
Is Meta's $19.2B Reality Labs loss a reason to avoid the stock?
It's a known drag, not a hidden risk. Every analyst model includes it; management has guided for similar levels in 2026. The bear case on Meta comes from capex ROI uncertainty and ad cyclicality — not Reality Labs specifically.
Is META stock a buy at $639?
At a PEG of 0.91 and a forward P/E of ~20x, the quantitative case is compelling versus peers. The practical case requires a view on: (1) whether the $125B capex generates adequate returns, (2) whether the ad cycle holds through 2026-27, and (3) whether privacy regulation materially impairs Andromeda's data targeting advantage. If your answers are cautiously optimistic, the base case implies meaningful upside. If you expect a recession or major privacy shock in 2026, the bear case at $450–520 is plausible.
Bottom Line
META grew operating income 41% in FY2025 while EPS appeared flat — the gap explained by a single quarter's one-time charge that has nothing to do with the underlying business. Normalized EPS grew ~20%.
The company is growing faster than Apple, Microsoft, or Google while trading at a cheaper forward multiple than all three. The market's discount reflects the $125B capex commitment and its uncertain ROI. That's a legitimate concern — but a business generating $83B in operating income and explicitly guiding for 2026 operating income to exceed 2025 has the financial capacity to fund the investment.
At a PEG of 0.91 and base-case intrinsic value of $680–780, the gap between what META earns and what the market is pricing for it remains unusually wide.
Fair Value Weekly
Get DCF breakdowns, fair value updates, and portfolio ideas for serious investors. No spam, no paywalled teasers.