Six days after our note on Meta's layoffs, the FY2025 numbers confirmed the structural picture: $46B FCF, down from $54B, on capex that climbed to $61B. The Capital Desk view: this is the inflection from margin story to capex story.
Six days ago, we argued in 'Meta's 8,000 Layoffs Are a Margin Move, Not a Confidence Vote' that the headcount reductions were a structural operating-margin defence rather than a strategic-confidence signal. The full FY2025 numbers, now in the books, confirm that framing: operating margin held at 41.3%, the highest absolute level the company has ever printed. The headcount discipline worked.
What the layoff note did not anticipate fully was the speed of the capex acceleration. FY2025 capex came in at approximately $61 billion, up from $39 billion in FY2024 and $28 billion in FY2023. The combination of operating margin discipline and capex acceleration has produced an income statement that reads strong on the operating-margin line and a cash flow statement that reads more pressured on the FCF line. FY2025 FCF compressed to $46.1 billion from $54.1 billion in FY2024.
The Capital Desk view, updated, is that Meta's stock is now driven primarily by the capex line rather than the operating-margin line. The operating-margin defence has been won. The capex investment cycle is the next phase, and the capex returns will determine whether the multiple sustains, expands, or compresses over the FY2026-FY2028 window.
We maintain a constructive view at today's $677 print, target $830 within 12 months, with downside risk to $570 if the FY2026 capex line clears $80 billion without commensurate revenue acceleration. The catalyst path runs through three operational data points: Reels monetisation, Llama-derived AI-feature ad-revenue lift, and the cumulative AI-infrastructure operating cost flow-through. Each is currently trending favourably, but the trajectory is what we monitor.
This update walks through the FY2025 numbers, the capex-cycle dynamics, the comparison to other hyperscalers' capex profiles, and the updated capital allocation framework.
Our coverage on Meta over the past 30 days has been substantial. The Risk Desk's 'Five Reasons Meta's Layoffs Do Not Fix the Capex Problem' framed the layoff-driven margin story as incomplete because the capex line was running ahead of the cost discipline. The Valuation Desk's 'The Chart Showing Meta's Ad Revenue Crossover With Google' reframed the multiple debate around Meta's emerging position as the largest digital-ad franchise globally. The Capital Desk's earlier note positioned the layoffs as a margin move executed without confidence in the underlying labour productivity.
The synthesis across desks: Meta is structurally well-positioned in the digital-advertising market, the operating margin trajectory is defensible, the capex line is a real concern but the underlying demand absorption is strong, and the FY2026-FY2027 outlook hinges on AI-infrastructure return-on-capital that is not yet visible in disclosed segment data.
The stock has moved from $645 to $677 over the four weeks since the layoff note. The move has been modest given the underlying revenue-growth and operating-margin print. We attribute the relative under-performance versus the broader large-cap technology cluster primarily to the capex-line concern and secondarily to the broader cyclical-multiple compression across mega-cap names.
TickerXray Report
Get the complete Meta report with all 12 quantitative models, AI-generated investment thesis, and real-time data.
Meta's FY2025 capex of approximately $61 billion deployed against operating cash flow of $107 billion produced FCF of $46 billion. The capex absorbed approximately 57% of operating cash flow. That ratio has accelerated from 31% in FY2023 to 42% in FY2024 to 57% in FY2025. The trajectory is the cleanest single concern for the standalone Meta thesis.
The capex programme is concentrated in three lines: AI-compute infrastructure (estimated $35 billion of FY2025 capex), data centre and network infrastructure (estimated $18 billion), and Reality Labs metaverse infrastructure (estimated $8 billion). Each line has a different return-on-capital profile.
The AI-compute line generates returns through the ad-targeting and Reels-monetisation engines. Meta has disclosed, at multiple recent earnings calls, that AI-driven ad-targeting improvements drove a measurable lift to ad-pricing and ad-conversion in FY2025. The translation from compute investment to revenue is direct, even if the precise IRR is hard to attribute. The Capital Desk model points to AI-compute capex generating roughly 35-45% IRR through the FY2026-FY2027 window, comfortably above the 9% cost of capital.
The data centre infrastructure line generates returns through capacity-supply for the underlying ad-serving and product-feature platforms. The IRR is harder to compute but is structurally similar to the cloud-infrastructure economics at hyperscalers. Estimated IRR is 25-30%, again well above cost of capital.
The Reality Labs line is the open question. FY2025 segment operating losses ran at approximately $20 billion, with revenue of approximately $2 billion. The Quest 3, Quest Pro, Ray-Ban Meta glasses, and the broader hardware-software roadmap are still in the investment phase. The IRR on Reality Labs capex is currently negative on a 5-year basis. The bull case requires the broader spatial-computing market to reach scale by FY2028-FY2029. The bear case views the spend as continuing without a credible monetisation path.
The Capital Desk view: AI-compute capex is value-accretive, data-centre capex is value-accretive, Reality Labs capex is value-questionable. The consolidated capex programme is therefore approximately 85% defensible and 15% open to challenge. That mix supports continued investment without breaking the value-creation case.
Total operating expenses grew at 19% in FY2025 against revenue growth of 22%, producing operating leverage that is, by Capital Desk standards, exceptional. Operating income reached $83 billion in FY2025, up from $68 billion in FY2024. The 41.3% operating margin is the structural ceiling under the current investment regime; further expansion would require either revenue acceleration or capex moderation.
Return on invested capital sits at 28% on the FY2025 base, up from 22% in FY2023. The trajectory is constructive even with the capex pressure on FCF. The Capital Desk reads ROIC at this level as confirmation that the capex programme is, in aggregate, value-creating despite the visible FCF compression.
Net cash position at end of FY2025 stood at approximately $48 billion against debt of $30 billion, producing a net-cash buffer of $18 billion. The buyback authorisation continues at scale, with FY2025 buyback of $30 billion. The combination of capex investment and buyback supports the per-share earnings trajectory through FY2027 even on flat FCF.
The analyst price target consensus has moved from $755 in mid-2024 to $855 today, reflecting both the FY2025 operational beats and the AI-narrative re-rating across the broader large-cap technology cluster. The target distribution shows a wide spread, with bull targets at $1,050 and bear targets at $590. The variance reflects the legitimate uncertainty around the capex-cycle return-on-capital.
Meta's $61 billion FY2025 capex compares to Microsoft's $50 billion, Amazon's $103 billion, and Google's $52 billion. As a percentage of revenue, Meta's capex ratio of 30% is the highest in the cluster, well above Microsoft's 18%, Amazon's 14%, or Google's 14%. The ratio reflects Meta's narrower revenue base relative to peers; the absolute capex is in line with the others, but the consolidated revenue is smaller.
The capex-to-FCF ratio is more informative. Meta's capex absorbed 57% of operating cash flow in FY2025. Microsoft's absorbed 41%. Amazon's absorbed 93% (driven by the elevated FY2025 spend). Google's absorbed approximately 35%. Meta's ratio sits in the middle of the cluster, which is structurally manageable for a franchise with 41% operating margins.
The AI-compute exposure differs across the cluster. Microsoft and Google have hyperscaler-cloud businesses where the AI-compute capex generates direct customer-facing revenue. Meta's AI-compute capex flows into the internal ad-platform and product-feature improvements. The translation from capex to revenue is more indirect for Meta, which adds uncertainty to the IRR computation. The Capital Desk view is that the indirect mechanism is no less effective at value creation; it is just harder to quantify externally.
The broader concern in the hyperscaler-capex debate is whether the cluster as a whole is over-investing relative to demand. The current operational data does not point that direction. AI-compute demand from enterprise customers continues to accelerate at Microsoft and Google. Meta's own ad-revenue acceleration in FY2025 confirms that the internal capex is producing measurable revenue lift. The cluster-wide demand absorption is the cleanest signal that the capex pace is justified.
If the cluster does begin to over-invest relative to demand, Meta would be more exposed than Microsoft or Google because of the indirect-monetisation mechanism. That is the structural risk we monitor.
Three engines drive Meta's growth profile through FY2028. The first is Reels monetisation, which has been the single cleanest revenue contributor over the past 18 months. The Reels ad-load now matches the broader Feed and Stories ad-load, and the per-impression CPM has converged toward the Feed level. The remaining monetisation runway is in the international markets where Reels engagement is high but advertiser penetration has lagged. We model Reels ad-revenue growing at 18-20% annually through FY2028 against a base of approximately $35 billion in FY2025.
The second is AI-driven ad-targeting improvements. The Andromeda and Lattice infrastructure (Meta's internal ad-AI engines) have been ramping through FY2024-FY2025, contributing measurable lift to ad-pricing and conversion rates. Each 5% lift in ad-pricing on Meta's $190 billion advertising revenue base contributes approximately $9.5 billion of incremental revenue at high incremental margins. The cumulative AI-driven lift over FY2026-FY2028 is the engine of our forecast revenue acceleration.
The third is the Threads and broader new-app monetisation that has been outside the core Family of Apps revenue base. Threads has scaled to over 350 million monthly active users by end of FY2025 with monetisation just beginning. WhatsApp Business and Click-to-WhatsApp ads continue to contribute incrementally. The aggregate contribution from non-Feed, non-Reels new monetisation should add 200-300 basis points to consolidated revenue growth annually through FY2028.
Layered together, the consolidated revenue growth trajectory through FY2028 sits at 18-22% annually, well above the broader large-cap technology cluster. EPS growth, supported by operating margin holding at 40%+ and continued buyback, lands at 18-22% annually. That is the compounding profile underlying our $830 12-month target.
Meta is positioned at the inflection point between the operating-margin defence (now substantially completed) and the capex-cycle execution (now the dominant variable). The capex programme is, in aggregate, value-creating, with AI-compute and data-centre lines clearly above cost of capital and Reality Labs an open question.
Fair value, anchored on a 25x forward multiple applied to FY2027 EPS of $33, lands at $825. Today's $677 print offers 22% upside on a 12-month view. The bull case to $950 requires Reels monetisation to accelerate further, AI-driven ad-targeting to lift consolidated ad-pricing by 8%+ in FY2026, and the FY2026 capex to come in below $75 billion. The bear case to $570 requires the FY2026 capex to clear $80 billion without revenue acceleration.
The trade is to own the franchise here. The compounding case is intact, the operating-margin defence is delivered, and the capex programme is, on balance, value-accretive. We are constructive at $677, target $830, trim only above $900.
The pattern across mega-cap technology capex cycles is consistent. The capex-driven FCF compression worries the market in the build phase, then the multiple recovers as the capex matures and the revenue line captures the benefit. Meta is in the build phase. The recovery is in front of investors. We position accordingly.
The updated thesis from this note is that capex now matters more than margin for the standalone Meta equity. The margin discipline has been delivered. The capex execution determines the next 24-36 months of return potential. We remain constructive on the execution path based on the disclosed operational data.
For portfolio managers running large-cap technology mandates, Meta remains a high-conviction overweight position within our framework. The combination of operating-margin discipline, AI-driven revenue acceleration, and disciplined-but-aggressive capex programme is rare across the mega-cap cluster. The capex line is the largest single source of near-term volatility, but it is also the largest source of long-term value creation in our model. Position the trade accordingly. The 41.3% operating margin Meta achieved in FY2025 sits among the highest absolute margins in any large-cap consumer-facing franchise globally. Sustaining that level while compounding revenue at 22% requires both pricing discipline and capacity-leverage execution that few peer-companies have produced. The capex programme is the next test of that execution. Our updated view is that the test continues to be passed and the position remains a constructive overweight.
Full forensic analysis of Meta
+ 6 more models included
150,000+ stocks covered
Global coverage across 60+ exchanges. Every report includes all 12 quantitative models and AI analysis.
View plansEvery report runs 12 quantitative models and generates an AI investment thesis. From Piotroski scores to manipulation detection -- get the full picture in seconds.
12 forensic models
Piotroski, Altman, Beneish, DuPont & more
AI investment thesis
Synthesized outlook on every stock
Manipulation detection
Spot red flags before they hit the news
150,000+ tickers
Global coverage across 60+ exchanges
Expected return
Forward return projections for every stock
Real-time data
Live prices, insider trades, news sentiment
Free accounts get 1 report per month. Pro gets unlimited.
Meta trades at 19.6x forward earnings with revenue growing 33% and operating margin near 50%. Five charts and the underlying analysis behind why the multiple makes sense.
Capex doubled to $69.7B while free cash flow compressed $8B. The layoffs protect operating margin. They do not protect the free cash flow yield investors are paying 22.6x forward earnings for.
Meta runs past Google Search in calendar 2026 at $243 billion. The capex cycle, the FCF plateau, and the margin expansion each produce different fair values.