Meta eyes big cuts to its metaverse budget in the AI era

Meta is reportedly preparing to trim metaverse spending — potentially by up to 30% in 2026 — and may initiate layoffs as it shifts capital and focus away from Reality Labs toward AI and core social apps. Reality Labs has accumulated over $60 billion in operating losses since 2021 (annual losses: $10.2B in 2021, $13.7B in 2022, $16.1B in 2023, $17.7B in 2024) and posted a $4.4B loss on ~$470M revenue in the most recent quarter; management is reallocating resources into $70–$72B of 2025 capex for data centers, custom chips and AI models and made a $14.3B investment for a 49% stake in Scale AI, a strategic move that lifted shares roughly 4% on the report.

Analysis

Market structure: Meta’s shift reallocates ~30% potential Reality Labs cuts (on top of $60bn cumulative losses) toward AI/data-center spending ($70–$72bn capex in 2025). Direct winners are AI-infrastructure suppliers (NVDA, AMD, cloud landlords) and ad-monetization franchises (Facebook/Instagram) that free cash flow supports; losers are niche AR/VR content and consumer headset suppliers with weak unit economics. GPU pricing power should remain elevated near-term as demand for training/infere­nce capacity absorbs reallocated spend.

Risk assessment: Tail risks include (1) regulatory limits on large-model monetization or ad-targeting (18–36 months), (2) chip supply shocks or passporting of critical export controls (0–12 months), and (3) an ad-revenue slowdown that negates margin gains (next 2 quarters). Immediate (days) reaction is sentiment-driven; short-term (weeks–months) risks center on announced layoffs and FY2026 guidance; long-term (years) depends on AR adoption curves and Meta’s AI monetization cadence. Hidden dependency: Meta’s compute needs are a material demand pillar for NVDA/AMD; cutting hardware R&D could paradoxically raise short-term chip demand for cloud training.

Trade implications: Favor conviction longs in NVDA (AI infra) and a tactical, hedged long in META to capture margin reallocation. Implement NVDA 3–6 month call spreads to express upside while capping cost; use protective put collars on META to monetize a sentiment re-rating with defined downside. Rotate away from pure-play AR/VR SMEs and shift 5–15% of that exposure into AI infrastructure and ad-revenue compounders within 2–6 weeks; size positions with 1–4% portfolio stakes and 10–12% stop-losses.

Contrarian angles: Markets underappreciate the negative network effect if Reality Labs cuts stall the AR hardware ecosystem—this could benefit AAPL’s spatial computing roadmap over 12–36 months. Conversely, the market may underprice faster ad/AI monetization at Meta; a successful enterprise AI product rollout in 6–12 months would be a positive asymmetric outcome. Historical parallels: scaled-back moonshots (e.g., Google X) often improve core margins and shareholder returns; the same could happen at Meta if capital is redeployed efficiently.

Key Decisions for Investors

• Establish a 2–3% long position in NVDA over the next 2–6 weeks using 3–6 month call debit spreads to express continued AI infra demand; take profits at +30% and tighten to a 12% trailing stop if NVDA misses revenue guidance by >5%.
• Initiate a 2–3% hedged long in META (buy shares + buy Jan 2026 puts ~12% OTM) to play margin recovery from Reality Labs cuts and higher ad/AI monetization; add on a pullback >5%, and trim to zero if management announces Reality Labs cuts >30% or guides FY25 free cash flow below consensus by >10%.
• Buy a 1–1.5% tactical position in AMD via 3–4 month call spreads to capture secular GPU/CPU demand, but cap exposure; exit if AMD reports sequential GPU server revenue decline >7% or if gross margin contracts by >200bps next quarter.
• Reduce exposure to pure-play AR/VR and headset content names by 50% within 30 days and redeploy proceeds into NVDA/NVDA call spreads and META hedged longs; reassess rotation after Meta’s next earnings (within 6–8 weeks) and material guidance on FY26 Reality Labs spend.