Is Meta Overvalued? The Case For and Against the Current Multiple

Meta's 2022 was a reset that looked, briefly, like a terminal decline. Revenue fell from $117.9 billion in 2021 to $116.6 billion. Operating income collapsed from $46.8 billion to $28.9 billion, compressing margins from 39.6% to 24.8%.

Mark Zuckerberg's response was the Year of Efficiency, which cut more than 20,000 jobs and restructured the cost base. The operational discipline was real, but the more important factor was that advertising demand recovered faster than expected through 2023.

The recovery was unusually fast. Revenue grew to $134.9 billion in 2023, $164.5 billion in 2024, and $201.0 billion in 2025. That is roughly $30 billion in new revenue added each year for three consecutive years, compounding at 22% annually.

By 2025, the narrative had shifted entirely. Meta was no longer cutting costs; it was spending aggressively on AI infrastructure. CapEx went from $27.3 billion in 2023 to $37.3 billion in 2024 to $69.7 billion in 2025, a 155% increase in two years. The efficiency era is over. The build-out era has begun.

Revenue of $201 billion in 2025 came almost entirely from advertising. Gross profit was $164.8 billion, a gross margin of 82.0%. That margin has held above 80% for five consecutive years, reflecting the near-zero marginal cost of serving additional advertisements to existing users.

Operating income reached $83.3 billion in 2025, a 41.4% margin. That compares to $28.9 billion and 24.8% at the 2022 trough. Revenue grew by $84 billion from 2022 to 2025, while operating income grew by $54 billion. The operating leverage is exceptional.

Net income was $60.5 billion in 2025, slightly below 2024's $62.4 billion. A higher effective tax rate and increased depreciation from new infrastructure offset the revenue growth. EBITDA expanded to $104.5 billion, up from $86.9 billion in 2024.

R&D spending hit $57.4 billion in 2025, or 28.5% of revenue. That is a high ratio for any advertising-driven business. Meta is simultaneously one of the most profitable and most aggressively reinvesting companies in the S&P 500. Those two facts are not usually found together.

A CapEx-to-revenue ratio of 34.6% is extraordinary for a software-driven advertising business. Amazon in peak AWS investment years ran 10 to 15%. Microsoft typically runs under 10%. Meta's capital intensity now rivals infrastructure companies, not software platforms.

Management has framed the spending as necessary AI infrastructure: training clusters, inference capacity for recommendation systems, and the data centers required to run both. Crucially, these investments directly power the products generating $201 billion in revenue today. They are not speculative in the way Reality Labs was.

Operating cash flow of $115.8 billion is the key number. Meta generated that from operations in 2025 and spent $69.7 billion on CapEx, leaving $46.1 billion in free cash flow. Of that, the company returned $26.2 billion via buybacks and $5.3 billion via dividends, totaling $31.5 billion, or 68% of free cash flow.

If CapEx stabilizes around $60 billion while revenue grows another 15% toward $231 billion, FCF expands toward $60 to $65 billion. That scenario supports meaningful shareholder return growth. If CapEx keeps rising proportionally to revenue, the FCF margin stays compressed.

The uncomfortable fact is that Meta's own guidance has underestimated capital spending for two consecutive years. Investors who model the bull case based on management CapEx guidance are using a number that has historically been too low.

A trailing PE of 22.4x and forward PE of 19.9x puts Meta roughly in line with the broader S&P 500. For a company with 22% revenue growth and 41% operating margins, that is not a demanding valuation. The PEG of 0.97 suggests the market has priced in the earnings growth rate without attaching a meaningful premium to it.

EV/EBITDA of 14.2x is well below peers. Microsoft trades above 25x EV/EBITDA. Alphabet trades around 15 to 17x. Meta's discount to both reflects the CapEx overhang, Reality Labs losses, and persistent regulatory uncertainty. Each of these is a known risk with a visible price.

Price-to-sales of 6.6x is near the low end of the mega-cap technology range. On a revenue multiple basis, Meta is among the cheapest in its peer group. That discount is either a structural mispricing or a justified reflection of capital intensity questions.

Analyst consensus is heavily bullish: 47 strong buy ratings, 13 buys, 6 holds, and 2 sells, with a median price target of $862.6. When analyst consensus is this one-sided, it typically signals that the downside scenario has been priced out. That rarely reflects reality accurately.

Meta's network density is the foundation of the entire investment case. Over 3 billion people use at least one Meta platform daily. The switching cost is social, not technical: users leave when their networks leave, and coordinated exits from social platforms at scale are historically rare.

The advertising business compounds the moat through first-party data. Meta knows more about its users' interests, behaviors, and social connections than any cookie-based or contextual advertising platform can replicate. When Apple's iOS 14 privacy changes disrupted attribution across digital advertising, Meta rebuilt its targeting stack using on-device signals and proprietary data from within its own ecosystem. The revenue recovery since that disruption has been substantial.

Advantage+ automated advertising is the current expression of that data advantage. Advertisers using the automated system consistently report higher returns on ad spend compared to manual campaign management. More automated spend means more data, which improves targeting accuracy, which attracts more spend. The feedback loop is self-reinforcing and difficult to replicate without an equivalent first-party data asset.

Reality Labs sits entirely outside this moat. The segment has consumed billions in annual losses and has no clear monetization timeline. The market appears to be pricing it close to zero. If spatial computing becomes a major platform, the current valuation would look deeply discounted. If it does not, the operating business can absorb the losses, but shareholder returns will be permanently reduced.

WhatsApp monetization is the largest underappreciated opportunity in Meta's portfolio. The platform has over 2 billion users globally but generates a fraction of the revenue per user that Facebook and Instagram produce. Meta has been carefully introducing business messaging tools, click-to-WhatsApp ads, and payment features in markets like India and Brazil. If WhatsApp generates even half the advertising revenue per user that Instagram does, the dollar impact on total company revenue would be measured in tens of billions annually.

Threads launched in 2023 and rapidly accumulated more than 300 million users. Monetization of Threads is still in early stages, but Meta has a clear template for how to layer advertising into a large social platform. Each new monetized surface adds incremental revenue at very high incremental margins, since the core infrastructure already exists and is largely already paid for.

AI-driven advertising tools are compressing the gap between creative production and campaign results. Meta's Advantage+ suite automates creative generation, audience selection, and budget allocation. The faster the feedback loop for advertisers, the more attractive Meta's platform becomes relative to alternatives that require more manual management. This flywheel directly rewards the AI infrastructure investment.

Shares outstanding have declined from roughly 2.89 billion in 2016 to 2.19 billion currently, a reduction of approximately 24%. That buyback program, combined with earnings growth, means EPS has compounded faster than net income. At $115.8 billion in annual operating cash flow, the capacity to continue buying back shares is substantial.