Meta Stock Seems to Be Bouncing Back. Is the Growth Stock Now Too Cheap to Ignore?

Meta reported Q3 2025 revenue of $51.2 billion, up 26% year-over-year, driven by advertising (≈98% of revenue) with ad impressions +14% and price per ad +10%; daily active users rose 8% to over 3.5 billion. Operating income was $20.5 billion (+18%), operating margin fell to 40% as costs and expenses rose 32% and free cash flow declined to $10.6 billion from $15.5 billion, while Q3 capex was $19.4 billion. Management raised 2025 capex guidance to $70–72 billion and warned 2026 capex and expense growth will be materially higher due to AI infrastructure investment; the stock trades at about 22x forward earnings and the company holds a net cash position (~$44.5B cash vs $29B long-term debt).

Analysis

Market structure: Meta’s heavy AI capex (Q3 capex $19.4B; 2025 guide $70–72B, 2026 “notably larger”) is a clear demand shock for GPUs, datacenter power and colocation — winners: NVIDIA, Equinix/AI cloud providers, power/transformer vendors; losers: smaller ad tech incumbents and legacy publishers as Meta increases impressions (up 14%) while CPMs rose 10%. Pricing power for ad inventory remains intact given 26% revenue growth, but margin compression is real as opex rises; expect upward pressure on GPU pricing and energy commodity demand over 6–24 months.

Risk assessment: Tail risks include a regulatory intervention (breakup/fines) or AI investment failure that leaves capex sunk and FCF depressed — both could shave >20–30% off EPS vs consensus. Time horizons: days—higher volatility around earnings/guidance; weeks–months—margin compression as 2025–26 capex flows through P&L; multi-year—monetization of AI could reaccelerate revenue >20% annually. Hidden dependencies: NVIDIA supply cadence, hyperscaler power availability, and global chip/export controls.

Trade implications: Direct play: asymmetric long in META if entry < $660 given ~22x forward EPS but set tight risk controls; complementary longs in NVDA and select cloud names for infrastructure exposure. Options: favors calendar/LEAP structures (buy Jan‑2027 calls or collars) to buy time for AI monetization while capping drawdown. Rotate: reduce cyclicals/exposed ad-tech and increase allocation to AI infra and energy‑grid capex names over 3–18 months.

Contrarian angles: Consensus is underpricing the upside of AI-driven ARPU improvements — if Meta converts 5–10% of new AI engagement to paid features over 24–36 months, upside re-rates >30%. Conversely, the market may be underestimating the execution and supply-chain risks; if 2026 capex > $80B or FCF falls >30% YoY, downside is underpriced. Historical parallel: Google/Facebook heavy investment cycles preceded durable ad share gains; outcome depends on execution cadence and NVDA supply.

Key Decisions for Investors

• Establish a 2–3% long position in META if price ≤ $660; target $900–920 within 12–18 months, set stop-loss at $560 (≈13–15% downside). Close or trim position if management signals 2026 capex > $80B or FCF decline >30% YoY.
• Allocate 0.5–1.5% to NVDA (direct equity) as a leveraged play on Meta’s capex; time horizon 6–12 months. Reduce exposure if NVIDIA reports material channel destocking or gross-margin compression >300bp next quarter.
• Implement a hedged equity collar when initiating or adding to META: buy shares, sell ~20% OTM calls (~$780 strike if entry $650) expiring 6–9 months to fund, and buy 12‑18 month puts ~15% OTM (~$550 strike) to cap tail risk while preserving upside over 12–18 months.
• Pair trade: long META (2% of portfolio) vs short SNAP (1%) or another small-cap ad name over 6–12 months to express consolidation in ad spend; unwind if SNAP outperforms META by >15% relative in 90 days.
• Trigger monitor: within 30–60 days, reassess on Q4 guidance and NVDA supply commentary; if Meta reiterates 2026 capex ‘notably larger’ but provides clear ARPU/monetization milestones, add incremental exposure up to +1%.